1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
25 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
26 files in the spool were present and unlocked. A queue-runner could spot
27 them, resulting in a duplicate delivery. Fix that by doing the unlock
28 after the unlink. Investigation by Tim Stewart. Take the opportunity to
29 add more error-checking on spoolfile handling while that code is being
32 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
33 No known attacks, no CVE, this is defensive hardening.
35 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
36 a queue-runner could start a delivery while other operations were ongoing.
37 Cutthrough delivery was a common victim, resulting in duplicate delivery.
38 Found and investigated by Tim Stewart. Fix by using the open message data
39 file handle rather than opening another, and not locally closing it (which
40 releases a lock) for that case, while creating the temporary .eml format
41 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
43 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
44 $sender_verify_failure/$recipient_verify_failure to "random".
46 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
49 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
50 Previously this would segfault.
52 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
55 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
56 like zero, since the resolver should be doing this for us, But we need one
57 as a CNAME but no MX presence gets the CNAME returned; we need to check
58 that doesn't point to an MX to declare it "no MX returned" rather than
59 "error, loop". A new main option is added so the older capability of
60 following some limited number of chain links is maintained.
62 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
64 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
65 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
66 now-deprecated earlier definitions used only specified the range up to TLS
67 1.2 (in the older-version library docs).
69 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
71 JH/15 Rework TLS client-side context management. Stop using a global, and
72 explicitly pass a context around. This enables future use of TLS for
73 connections to service-daemons (eg. malware scanning) while a client smtp
74 connection is using TLS; with cutthrough connections this is quite likely.
76 JH/16 Fix ARC verification to do AS checks in reverse order.
78 JH/17 Support a "tls" option on the ${readsocket } expansion item.
80 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
81 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
82 Previously the "utf8" would be re-prepended for every additional message.
84 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
85 Previously thery were accepted, resulting in issues when attempting to
86 forward messages to a non-supporting MTA.
88 PP/02 Let -n work with printing macros too, not just options.
90 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
91 one parent address was copied, and bogus data was used at delivery-logging
92 time. Either a crash (after delivery) or bogus log data could result.
93 Discovery and analysis by Tim Stewart.
95 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
96 Previously if the string ended mid-character, we did not insert the
97 promised '?' replacement.
99 PP/04 Documentation: current string operators work on bytes, not codepoints.
101 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
102 should pack well giving a smaller memory footprint so better caching and
103 therefore performance. Group the declarations where this can't be done so
104 that the byte-sized flag variables are not interspersed among pointer
105 variables, giving a better chance of good packing by the compiler.
107 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
108 non-null, to avoid issues with sites running BATV. Previously reports were
109 sent with an empty envelope sender so looked like bounces.
111 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
112 The ignore_error flag wasn't being returned from the filter subprocess so
113 was not set for later routers. Investigation and fix by Matthias Kurz.
115 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
116 and a msg:complete for the whole, when a message is manually reoved using
117 -Mrm. Developement by Matthias Kurz, hacked on by JH.
123 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
124 When redis_servers is set to a list of > 1 element, and the Redis servers
125 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
126 case of MOVED into a DEFER case instead, thus moving the query onto the
127 next server in the list. For a cluster of N elements, all N servers must
128 be defined in redis_servers.
130 GF/02 Catch and remove uninitialized value warning in exiqsumm
131 Check for existence of @ARGV before looking at $ARGV[0]
133 JH/01 Replace the store_release() internal interface with store_newblock(),
134 which internalises the check required to safely use the old one, plus
135 the allocate and data copy operations duplicated in both (!) of the
136 extant use locations.
138 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
139 modifier. This matches the restriction on the commandline.
141 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
142 Previously only the last row was returned.
144 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
145 we assumed that tags in the header were well-formed, and parsed the
146 element content after inspecting only the first char of the tag.
147 Assumptions at that stage could crash the receive process on malformed
150 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
151 While running the DKIM ACL we operate on the Permanent memory pool so that
152 variables created with "set" persist to the DATA ACL. Also (at any time)
153 DNS lookups that fail create cache records using the Permanent pool. But
154 expansions release any allocations made on the current pool - so a dnsdb
155 lookup expansion done in the DKIM ACL releases the memory used for the
156 DNS negative-cache, and bad things result. Solution is to switch to the
157 Main pool for expansions.
158 While we're in that code, add checks on the DNS cache during store_reset,
159 active in the testsuite.
160 Problem spotted, and debugging aided, by Wolfgang Breyha.
162 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
163 When none of the hosts presented to a transport match an already-open
164 connection, close it and proceed with the list. Previously we would
165 queue the message. Spotted by Lena with Yahoo, probably involving
168 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
169 Previously a spurious "250 OK id=" response was appended to the proper
172 JH/08 The "support for" informational output now, which built with Content
173 Scanning support, has a line for the malware scanner interfaces compiled
174 in. Interface can be individually included or not at build time.
176 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
177 by the template makefile "src/EDITME". The "STREAM" support for an older
178 ClamAV interface method is removed.
180 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
181 rows affected is given instead).
183 JH/11 The runtime Berkeley DB library version is now additionally output by
184 "exim -d -bV". Previously only the compile-time version was shown.
186 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
187 SMTP connection. Previously, when one had more receipients than the
188 first, an abortive onward connection was made. Move to full support for
189 multiple onward connections in sequence, handling cutthrough connection
190 for all multi-message initiating connections.
192 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
193 routers. Previously, a multi-recipient message would fail to match the
194 onward-connection opened for the first recipient, and cause its closure.
196 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
197 a timeout on read on a GnuTLS initiating connection, resulting in the
198 initiating connection being dropped. This mattered most when the callout
199 was marked defer_ok. Fix to keep the two timeout-detection methods
202 JH/15 Relax results from ACL control request to enable cutthrough, in
203 unsupported situations, from error to silently (except under debug)
204 ignoring. This covers use with PRDR, frozen messages, queue-only and
207 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
209 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
210 metadata, resulting in a crash in free().
212 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
213 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
214 Broken also in d185889f4, with init system revamp.
216 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
217 Previously we abruptly closed the connection after reading a malware-
218 found indication; now we go on to read the "scan ok" response line,
221 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
222 ACL. Previously, a crash would result.
224 JH/19 Speed up macro lookups during configuration file read, by skipping non-
225 macro text after a replacement (previously it was only once per line) and
226 by skipping builtin macros when searching for an uppercase lead character.
228 JH/20 DANE support moved from Experimental to mainline. The Makefile control
229 for the build is renamed.
231 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
232 was allocated for every new TLS startup, meaning one per message. Fix
233 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
235 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
236 reported the original. Fix to report (as far as possible) the ACL
237 result replacing the original.
239 JH/23 Fix memory leak during multi-message connections using STARTTLS under
240 OpenSSL. Certificate information is loaded for every new TLS startup,
241 and the resources needed to be freed.
243 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
245 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
248 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
249 DATA response info to the (existing) per-recipient response info for
250 the "C=" log element. It can have useful tracking info from the
251 destination system. Patch from Simon Arlott.
253 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
254 length value. Previously this would segfault.
256 HS/02 Support Avast multiline protoocol, this allows passing flags to
257 newer versions of the scanner.
259 JH/28 Ensure that variables possibly set during message acceptance are marked
260 dead before release of memory in the daemon loop. This stops complaints
261 about them when the debug_store option is enabled. Discovered specifically
262 for sender_rate_period, but applies to a whole set of variables.
263 Do the same for the queue-runner and queue-list loops, for variables set
264 from spool message files. Do the same for the SMTP per-message loop, for
265 certain variables indirectly set in ACL operations.
267 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
268 as a multi-recipient message from a mailinglist manager). The coding had
269 an arbitrary cutoff number of characters while checking for more input;
270 enforced by writing a NUL into the buffer. This corrupted long / fast
271 input. The problem was exposed more widely when more pipelineing of SMTP
272 responses was introduced, and one Exim system was feeding another.
273 The symptom is log complaints of SMTP syntax error (NUL chars) on the
274 receiving system, and refused recipients seen by the sending system
275 (propating to people being dropped from mailing lists).
276 Discovered and pinpointed by David Carter.
278 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
279 replaced by the ${authresults } expansion.
281 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
283 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
284 allows proper process termination in container environments.
286 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
287 Previously the "final dot" had a newline after it; ensure it is CR,LF.
289 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
290 and "err_perm", deprecated since 4.83 when the RFC-defined words
291 "temperror" and "permerror" were introduced.
293 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
294 transport-filters or DKIM-signing. The restriction was lost in the
295 consolidation of verify-callout and delivery SMTP handling.
296 Extend the restriction to also cover ARC-signing.
298 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
299 in defer=pass mode supply a 450 to the initiator. Previously the message
302 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
303 tls_require_ciphers is used as before.
305 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
306 "pass_unscanned". Only tmpfails from the scanner are written to
307 the paniclog, as they may require admin intervention (permission
308 denied, license issues). Other scanner errors (like decompression
309 bombs) do not cause a paniclog entry.
311 JH/36 Fix reinitialisation of DKIM logging variable between messages.
312 Previously it was possible to log spurious information in receive log
315 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
316 triggered odd behaviour from Outlook Express clients.
318 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
321 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
322 since the IETF WG has not yet settled on that versus the original
323 "bare" representation.
325 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
326 Previously the millisecond value corrupted the output.
327 Fix also for syslog_pid=no and log_selector +pid, for which the pid
328 corrupted the output.
334 JH/01 Rework error string handling in TLS interface so that the caller in
335 more cases is responsible for logging. This permits library-sourced
336 string to be attached to addresses during delivery, and collapses
337 pairs of long lines into single ones.
339 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
340 during configuration. Wildcards are allowed and expanded.
342 JH/02 Rework error string handling in DKIM to pass more info back to callers.
343 This permits better logging.
345 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
346 do not close it down and have the child transport start it up again on
347 the passed-on TCP connection. Instead, proxy the child (and any
348 subsequent ones) for TLS via a unix-domain socket channel. Logging is
349 affected: the continued delivery log lines do not have any DNSSEC, TLS
350 Certificate or OCSP information. TLS cipher information is still logged.
352 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
353 identical IP addresses on different listening ports. Will also affect
356 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
357 add noisy ifdef guards to special-case this sillyness.
358 Patch from Bernd Kuhls.
360 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
361 than 255 are no longer allowed.
363 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
364 Disable the session-cache too, which might reduce our load. Since we
365 currrectly use a new context for every connection, both as server and
366 client, there is no benefit for these.
367 GnuTLS appears to not support tickets server-side by default (we don't
368 call gnutls_session_ticket_enable_server()) but client side is enabled
369 by default on recent versions (3.1.3 +) unless the PFS priority string
372 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
373 <https://reproducible-builds.org/specs/source-date-epoch/>.
375 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
376 the check for any unsuccessful recipients did not notice the limit, and
377 erroneously found still-pending ones.
379 JH/08 Pipeline CHUNKING command and data together, on kernels that support
380 MSG_MORE. Only in-clear (not on TLS connections).
382 JH/09 Avoid using a temporary file during transport using dkim. Unless a
383 transport-filter is involved we can buffer the headers in memory for
384 creating the signature, and read the spool data file once for the
385 signature and again for transmission.
387 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
388 4.77 as the kernel support then wasn't solid, having issues in 64bit
389 mode. Now, it's been long enough. Add support for FreeBSD also.
391 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
392 case where the routing stage had gathered several addresses to send to
393 a host before calling the transport for the first, we previously failed
394 to close down TLS in the old transport process before passing the TCP
395 connection to the new process. The new one sent a STARTTLS command
396 which naturally failed, giving a failed delivery and bloating the retry
397 database. Investigation and fix prototype from Wolfgang Breyha.
399 JH/12 Fix check on SMTP command input synchronisation. Previously there were
400 false-negatives in the check that the sender had not preempted a response
401 or prompt from Exim (running as a server), due to that code's lack of
402 awareness of the SMTP input buffering.
404 PP/04 Add commandline_checks_require_admin option.
405 Exim drops privileges sanely, various checks such as -be aren't a
406 security problem, as long as you trust local users with access to their
407 own account. When invoked by services which pass untrusted data to
408 Exim, this might be an issue. Set this option in main configuration
409 AND make fixes to the calling application, such as using `--` to stop
412 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
413 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
414 responses to those, into a single TLS record each way (this usually means
415 a single packet). As a side issue, smtp_enforce_sync now works on TLS
418 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
419 affects you only if you're dancing at the edge of the param size limits.
420 If you are, and this message makes sense to you, then: raise the
421 configured limit or use OpenSSL 1.1. Nothing we can do for older
424 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
425 cmdline element to get the documented default one. Previously it was
428 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
429 are unsuable for non-message-specific reasons - eg. connection timeout,
430 banner-time rejection.
432 JH/16 Fix logging of delivery remote port, when specified by router, under
435 PP/06 Repair manualroute's ability to take options in any order, even if one
436 is the name of a transport.
439 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
441 JH/17 Change the list-building routines interface to use the expanding-string
442 triplet model, for better allocation and copying behaviour.
444 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
445 Previously it was constructed the first time a possibly-matching string
446 was met in the configuration file input during startup; now it is done
449 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
450 compatible one, to avoid the (poorly documented) possibility of a config
451 file in the working directory redirecting the DB files, possibly correpting
452 some existing file. CVE-2017-10140 assigned for BDB.
454 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
455 cache-hot. Previously, although the result was properly cached, the
456 initial verify call returned a defer.
458 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
459 the main verify for receipient in uncached-mode.
461 JH/22 Retire historical build files to an "unsupported" subdir. These are
462 defined as "ones for which we have no current evidence of testing".
464 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
465 if present. Previously it was ignored.
467 JH/24 Start using specified-initialisers in C structure init coding. This is
468 a C99 feature (it's 2017, so now considered safe).
470 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
471 if was a fixed-sized field and bitmask ops via macros; it is now more
474 PP/07 GitHub PR 56: Apply MariaDB build fix.
475 Patch provided by Jaroslav Å karvada.
477 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
478 during Coverity cleanups [4.87 JH/47]
479 Diagnosis and fix provided by Michael Fischer v. Mollard.
481 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
482 the right size to place the terminating semicolon on its own folded
483 line, the header hash was calculated to an incorrect value thanks to
484 the (relaxed) space the fold became.
486 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
487 and confused the parent.
489 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
490 which could crash as a result. This could lead to undeliverable messages.
492 JH/28 Logging: "next input sent too soon" now shows where input was truncated
495 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
496 matters on fast-turnover and PID-randomising systems, which were getting
497 out-of-order delivery.
499 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
500 a possibly-overlapping copy. The symptom was that "Remote host closed
501 connection in response to HELO" was logged instead of the actual 4xx
504 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
505 Previously only that bufferd was discarded, resulting in SYMTP command
508 JH/32 DKIM: when a message has multiple signatures matching an identity given
509 in dkim_verify_signers, run the dkim acl once for each. Previously only
510 one run was done. Bug 2189.
512 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
513 "panic the current process" to "deliberately defer". The panic log is
514 still written with the problem list name; the mail and reject logs now
515 get a temp-reject line for the message that was being handled, saying
516 something like "domains check lookup or other defer". The SMTP 451
517 message is still "Temporary local problem".
519 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
520 A crafted sequence of BDAT commands could result in in-use memory beeing
521 freed. CVE-2017-16943.
523 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
524 from SMTP input. Previously it was always done; now only done for DATA
525 and not BDAT commands. CVE-2017-16944.
527 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
528 to the message (such as an overlong header line). Previously this was
529 not done and we did not exit BDAT mode. Followon from the previous item
530 though a different problem.
536 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
537 than -2003 did; needs libidn2 in addition to libidn.
539 JH/02 The path option on a pipe transport is now expanded before use.
541 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
542 Patch provided by "Björn", documentation fix added too.
544 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
545 missing a wire-to-host endian conversion.
547 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
548 close after a BDAT command line could be taken as a following command,
549 giving a synch failure. Fix by only checking for synch immediately
550 before acknowledging the chunk.
552 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
553 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
554 macro. Patches provided by Josh Soref.
556 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
557 Previously we did not; the RFC seems ambiguous and VRFY is not listed
558 by IANA as a service extension. However, John Klensin suggests that we
561 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
562 the dkim code may be unix-mode line endings rather than smtp wire-format
563 CRLF, so prepend a CR to any bare LF.
565 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
566 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
568 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
569 the first were themselves being wrongly included in the feed into dkim
570 processing; with most chunk sizes in use this resulted in an incorrect
571 body hash calculated value.
573 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
574 DKIM signature block, for verification. Although advised against by
575 standards it is specifically not ruled illegal.
577 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
579 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
580 missing a body hash (the bh= tag).
582 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
583 It seems that HAProxy sends the Proxy Protocol information in clear and
584 only then does a TLS startup, so do the same.
586 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
587 TCP connections (such as for Spamd) unless the daemon successfully set
588 Fast Open mode on its listening sockets. This fixes breakage seen on
589 too-old kernels or those not configured for Fast Open, at the cost of
590 requiring both directions being enabled for TFO, and TFO never being used
591 by non-daemon-related Exim processes.
593 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
594 endings, at least on the first header line. Try to canonify any that get
595 past that check, despite the cost.
597 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
598 now limited to an arbitrary five deep, while parsing addresses with the
599 strip_excess_angle_brackets option enabled.
601 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
602 instead leave the unprompted TLS handshake in socket buffer for the
603 TLS library to consume.
605 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
607 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
609 JH/16 Drop variables when they go out of scope. Memory management drops a whole
610 region in one operation, for speed, and this leaves assigned pointers
611 dangling. Add checks run only under the testsuite which checks all
612 variables at a store-reset and panics on a dangling pointer; add code
613 explicitly nulling out all the variables discovered. Fixes one known
614 bug: a transport crash, where a dangling pointer for $sending_ip_address
615 originally assigned in a verify callout, is re-used.
617 PP/06 Drop '.' from @INC in various Perl scripts.
619 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
621 PP/08 Reduce a number of compilation warnings under clang; building with
622 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
623 should be warning-free.
625 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
627 HS/01 Fix portability problems introduced by PP/08 for platforms where
628 realloc(NULL) is not equivalent to malloc() [SunOS et al].
630 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
631 chunk. This allows us to accept broken chunked messages. We need a more
632 general solution here.
634 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
635 already-broken messages in the queue.
637 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
639 JH/19 Fix reference counting bug in routing-generated-address tracking.
645 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
646 supports it and a size is available (ie. the sending peer gave us one).
648 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
649 years of being deprecated). The replacements are the ACLs
650 acl_smtp_mime and acl_not_smtp_mime.
652 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
653 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
654 or even an in-clear connection were permitted. Now, if the host lookup
655 was dnssec and dane was requested then the host is only used if the
656 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
657 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
658 if one fails this test.
659 This means that a poorly-configured remote DNS will make it incommunicado;
660 but it protects against a DNS-interception attack on it.
662 JH/04 Bug 1810: make continued-use of an open smtp transport connection
663 non-noisy when a race steals the message being considered.
665 JH/05 If main configuration option tls_certificate is unset, generate a
666 self-signed certificate for inbound TLS connections.
668 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
669 in rewrites and routers.
671 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
672 and logged a warning sing 4.83; now they are a configuration file error.
674 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
675 (lacking @domain). Apply the same qualification processing as RCPT.
677 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
679 JH/10 Support ${sha256:} applied to a string (as well as the previous
682 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
683 a cutthrough deliver is pending, as we always want to make a connection.
684 This also avoids re-routing the message when later placing the cutthrough
685 connection after a verify cache hit.
686 Do not update it with the verify result either.
688 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
689 when routing results in more than one destination address.
691 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
692 signing (which inhibits the cutthrough capability). Previously only
693 the presence of an option was tested; now an expansion evaluating as
694 empty is permissible (obviously it should depend only on data available
695 when the cutthrough connection is made).
697 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
698 the relevant preceding SMTP command did not note the pipelining mode.
700 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
701 Previously they were not counted.
703 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
704 as one having no matching records. Previously we deferred the message
705 that needed the lookup.
707 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
708 distinguished as "(=".
710 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
711 for missing MX records. Previously it only worked for missing A records.
713 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
715 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
716 after the data-go-ahead and data-ack. Patch from Jason Betts.
718 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
719 even for a "none" policy. Patch from Tony Meyer.
721 JH/22 Fix continued use of a connection for further deliveries. If a port was
722 specified by a router, it must also match for the delivery to be
725 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
726 When one of the recipients of a message was unsuitable for the connection
727 (has no matching addresses), we lost track of needing to mark it
728 deferred. As a result mail would be lost.
730 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
732 JH/25 Decoding ACL controls is now done using a binary search; the source code
733 takes up less space and should be simpler to maintain. Merge the ACL
734 condition decode tables also, with similar effect.
736 JH/26 Fix problem with one_time used on a redirect router which returned the
737 parent address unchanged. A retry would see the parent address marked as
738 delivered, so not attempt the (identical) child. As a result mail would
741 JH/27 Fix a possible security hole, wherein a process operating with the Exim
742 UID can gain a root shell. Credit to http://www.halfdog.net/ for
743 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
746 JH/28 Enable {spool,log} filesystem space and inode checks as default.
747 Main config options check_{log,spool}_{inodes,space} are now
748 100 inodes, 10MB unless set otherwise in the configuration.
750 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
751 Previously it only applied to the main-section connection policy
754 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
756 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
757 by me. Added RFC7919 DH primes as an alternative.
759 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
760 are not in the system include path.
762 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
763 GnuTLS, when a session startup failed (eg because the client disconnected)
764 Exim did stdio operations after fclose. This was exposed by a recent
765 change which nulled out the file handle after the fclose.
767 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
768 signed directly by the cert-signing cert, rather than an intermediate
769 OCSP-signing cert. This is the model used by LetsEncrypt.
771 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
773 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
774 an incoming connection.
776 HS/02 Bug 1802: Do not half-close the connection after sending a request
779 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
780 fallback to "prime256v1".
782 JH/34 SECURITY: Use proper copy of DATA command in error message.
783 Could leak key material. Remotely exploitable. CVE-2016-9963.
789 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
790 and 3.4.4 - once the server is enabled to respond to an OCSP request
791 it does even when not requested, resulting in a stapling non-aware
792 client dropping the TLS connection.
794 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
795 support variable-length bit vectors. No functional change.
797 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
798 The I= interface field on outgoing lines is now after the H= remote
799 host field, same as incoming lines. There is a separate
800 outgoing_interface log selector which allows you to disable the
803 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
804 If not running log_selector +smtp_connection the mainlog would be held
805 open indefinitely after a "too many connections" event, including to a
806 deleted file after a log rotate. Leave the per net connection logging
807 leaving it open for efficiency as that will be quickly detected by the
808 check on the next write.
810 HS/01 Bug 1671: Fix post transport crash.
811 Processing the wait-<transport> messages could crash the delivery
812 process if the message IDs didn't exist for some reason. When
813 using 'split_spool_directory=yes' the construction of the spool
814 file name failed already, exposing the same netto behaviour.
816 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
817 mime_regex ACL conditions.
819 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
820 to DSN fail messages (bounces): remote IP, remote greeting, remote response
821 to HELO, local diagnostic string.
823 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
824 log line to debug. Even when configured with a tls authenticator many
825 client connections are expected to not authenticate in this way, so
826 an authenticate fail is not an error.
828 HS/02 Add the Exim version string to the process info. This way exiwhat
829 gives some more detail about the running daemon.
831 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
832 matter for fast-change records such as DNSBLs.
834 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
835 retry record, even if constant. There may be multiple transports with
836 different interface settings and the retry behaviour needs to be kept
839 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
841 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
843 JH/10 Bug 840: fix log_defer_output option of pipe transport
845 JH/11 Bug 830: use same host for all RCPTS of a message, even under
846 hosts_randomize. This matters a lot when combined with mua_wrapper.
848 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
849 ${quote_pgsql:<string>} operator.
851 JH/13 Bug 1708: avoid misaligned access in cached lookup.
853 JH/14 Change header file name for freeradius-client. Relevant if compiling
854 with Radius support; from the Gentoo tree and checked under Fedora.
856 JH/15 Bug 1712: Introduce $prdr_requested flag variable
858 JH/16 Bug 1714: Permit an empty string as expansion result for transport
859 option transport_filter, meaning no filtering.
861 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
863 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
864 defaults to "*" (all hosts). The variable is now available when not built
865 with TLS, default unset, mainly to enable keeping the testsuite sane.
866 If a server certificate is not supplied (via tls_certificate) an error is
867 logged, and clients will find TLS connections fail on startup. Presumably
868 they will retry in-clear.
869 Packagers of Exim are strongly encouraged to create a server certificate
870 at installation time.
872 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
873 with the $config_file variable.
875 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
876 in transport context, after the attempt, and per-recipient. The latter type
877 is per host attempted. The event data is the error message, and the errno
878 information encodes the lookup type (A vs. MX) used for the (first) host,
879 and the trailing two digits of the smtp 4xx response.
881 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
882 to write to mainlog (or rejectlog, paniclog) in the window between file
883 creation and permissions/ownership being changed. Particularly affects
884 installations where exicyclog is run as root, rather than exim user;
885 result is that the running daemon panics and dies.
887 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
889 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
890 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
891 "pri" and "weight". Note that the previous implicit priority given by the
892 list order is no longer honoured.
894 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
897 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
898 by defining SUPPORT_SOCKS.
900 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
901 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
902 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
903 variables are renamed to proxy_{local,external}_{address,port}.
905 JH/27 Move Internationalisation support from Experimental to mainline, enabled
906 for a build by defining SUPPORT_I18N
908 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
909 of the query string, and make ${quote_redis:} do that quoting.
911 JH/29 Move Events support from Experimental to mainline, enabled by default
912 and removable for a build by defining DISABLE_EVENT.
914 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
916 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
917 cached by the daemon.
919 JH/32 Move Redis support from Experimental to mainline, enabled for a build
920 by defining LOOKUP_REDIS. The libhiredis library is required.
922 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
923 keys are given for lookup.
925 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
926 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
927 only supported when built with TLS support. The PolarSSL SHA routines
928 are still used when the TLS library is too old for convenient support.
930 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
931 openssl_options), for security. OpenSSL forces this from version 1.1.0
932 server-side so match that on older versions.
934 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
935 allocation for $value could be released as the expansion processing
936 concluded, but leaving the global pointer active for it.
938 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
939 and to use the domains and local_parts ACL conditions.
941 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
942 incorrectly not doubled on cutthrough transmission, hence seen as a
943 body-termination at the receiving system - resulting in truncated mails.
944 Commonly the sender saw a TCP-level error, and retransmitted the message
945 via the normal store-and-forward channel. This could result in duplicates
946 received - but deduplicating mailstores were liable to retain only the
947 initial truncated version.
949 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
951 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
953 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
954 we're in there, support oversigning also; bug 1309.
956 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
958 HS/04 Add support for keep_environment and add_environment options.
960 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
961 either intentional arithmetic overflow during PRNG, or testing config-
964 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
965 delivery resulted in actual delivery. Cancel cutthrough before DATA
968 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
969 rejects a recipient: pass the reject to the originator.
971 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
972 Many were false-positives and ignorable, but it's worth fixing the
975 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
976 for the new environment-manipulation done at startup. Move the routines
977 from being local to tls.c to being global via the os.c file.
979 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
980 an extract embedded as result-arg for a map, the first arg for extract
981 is unavailable so we cannot tell if this is a numbered or keyed
982 extraction. Accept either.
988 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
991 JH/02 The smtp transport option "multi_domain" is now expanded.
993 JH/03 The smtp transport now requests PRDR by default, if the server offers
996 JH/04 Certificate name checking on server certificates, when exim is a client,
997 is now done by default. The transport option tls_verify_cert_hostnames
998 can be used to disable this per-host. The build option
999 EXPERIMENTAL_CERTNAMES is withdrawn.
1001 JH/05 The value of the tls_verify_certificates smtp transport and main options
1002 default to the word "system" to access the system default CA bundle.
1003 For GnuTLS, only version 3.0.20 or later.
1005 JH/06 Verification of the server certificate for a TLS connection is now tried
1006 (but not required) by default. The verification status is now logged by
1007 default, for both outbound TLS and client-certificate supplying inbound
1010 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1013 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1014 Status Notification (bounce) messages are now MIME format per RFC 3464.
1015 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1016 under the control of the dsn_advertise_hosts option, and routers may
1017 have a dsn_lasthop option.
1019 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1020 default, modifiable by a malware= option. The list separator for
1021 the options can now be changed in the usual way. Bug 68.
1023 JH/10 The smtp_receive_timeout main option is now expanded before use.
1025 JH/11 The incoming_interface log option now also enables logging of the
1026 local interface on delivery outgoing connections.
1028 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1029 if the interface and destination host and port all match.
1031 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1034 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1035 Patch from Andrew Lewis.
1037 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1038 now supports optional time-restrictions, weighting, and priority
1039 modifiers per server. Patch originally by <rommer@active.by>.
1041 JH/16 The spamd_address main option now supports a mixed list of local
1042 and remote servers. Remote servers can be IPv6 addresses, and
1043 specify a port-range.
1045 JH/17 Bug 68: The spamd_address main option now supports an optional
1046 timeout value per server.
1048 JH/18 Bug 1581: Router and transport options headers_add/remove can
1049 now have the list separator specified.
1051 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1054 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1057 JH/21 Support for the A6 type of dns record is withdrawn.
1059 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1060 rather than the verbs used.
1062 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1063 from 255 to 1024 chars.
1065 JH/24 Verification callouts now attempt to use TLS by default.
1067 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1068 are generic router options now. The defaults didn't change.
1070 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1071 Original patch from Alexander Shikoff, worked over by JH.
1073 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1074 files if ZCAT_COMMAND is not executable.
1076 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1078 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1080 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1081 Normally benign, it bites when the pair was led to by a CNAME;
1082 modern usage is to not canonicalize the domain to a CNAME target
1083 (and we were inconsistent anyway for A-only vs AAAA+A).
1085 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1087 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1088 when evaluating $sender_host_dnssec.
1090 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1091 $sender_helo_dnssec variable.
1093 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1095 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1097 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1099 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1100 documented as working, but never had. Support all but $spam_report.
1102 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1103 added for tls authenticator.
1105 HS/03 Add perl_taintmode main config option
1111 TL/01 When running the test suite, the README says that variables such as
1112 no_msglog_check are global and can be placed anywhere in a specific
1113 test's script, however it was observed that placement needed to be near
1114 the beginning for it to behave that way. Changed the runtest perl
1115 script to read through the entire script once to detect and set these
1116 variables, reset to the beginning of the script, and then run through
1117 the script parsing/test process like normal.
1119 TL/02 The BSD's have an arc4random API. One of the functions to induce
1120 adding randomness was arc4random_stir(), but it has been removed in
1121 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1122 function when detected.
1124 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1125 cause callback expansion.
1127 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1128 syntax errors in an expansion can be treated as a string instead of
1129 logging or causing an error, due to the internal use of bool_lax
1130 instead of bool when processing it.
1132 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1133 server certificates when making smtp deliveries.
1135 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1137 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1139 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1141 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1142 Merged patch from Sebastian Wiedenroth.
1144 JH/05 Fix results-pipe from transport process. Several recipients, combined
1145 with certificate use, exposed issues where response data items split
1146 over buffer boundaries were not parsed properly. This eventually
1147 resulted in duplicates being sent. This issue only became common enough
1148 to notice due to the introduction of connection certificate information,
1149 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1151 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1152 size buffer was used, resulting in syntax errors when an expansion
1155 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1156 version 3.3.6 or later.
1158 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1159 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1160 both become "event_action", the variables become $event_name, $event_data
1161 and $event_defer_errno. There is a new variable $verify_mode, usable in
1162 routers, transports and related events. The tls:cert event is now also
1163 raised for inbound connections, if the main configuration event_action
1166 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1167 early OCSP support, but no stapling (appears to be less than 1.0.0).
1169 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1170 server certificate names available under the smtp transport option
1171 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1174 JH/10 Time-related extraction expansions from certificates now use the main
1175 option "timezone" setting for output formatting, and are consistent
1176 between OpenSSL and GnuTLS compilations. Bug 1541.
1178 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1179 encoded parameter in the incoming message. Bug 1558.
1181 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1182 include certificate info, eximon was claiming there were spoolfile
1185 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1187 JH/14 Log delivery-related information more consistently, using the sequence
1188 "H=<name> [<ip>]" wherever possible.
1190 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1191 are problematic for Debian distribution, omit them from the release
1194 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1196 JH/16 Fix string representation of time values on 64bit time_t architectures.
1199 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1200 output list separator was used.
1205 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1206 checkers that were complaining about end of non-void function with no
1209 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1210 This was a regression introduced in 4.83 by another bugfix.
1212 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1214 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1215 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1221 TF/01 Correctly close the server side of TLS when forking for delivery.
1223 When a message was received over SMTP with TLS, Exim failed to clear up
1224 the incoming connection properly after forking off the child process to
1225 deliver the message. In some situations the subsequent outgoing
1226 delivery connection happened to have the same fd number as the incoming
1227 connection previously had. Exim would try to use TLS and fail, logging
1228 a "Bad file descriptor" error.
1230 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1231 utilities have not been installed.
1233 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1234 temporary space as the ACL may create new global variables.
1236 TL/01 LDAP support uses per connection or global context settings, depending
1237 upon the detected version of the libraries at build time.
1239 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1240 to extract and use the src ip:port in logging and expansions as if it
1241 were a direct connection from the outside internet. PPv2 support was
1242 updated based on HAProxy spec change in May 2014.
1244 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1246 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1247 Properly escape header and check for NULL return.
1249 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1252 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1254 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1255 characters in header names, implemented as a verify condition.
1256 Contributed by Michael Fischer v. Mollard.
1258 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1259 results permerror and temperror. Previous values are deprecated but
1260 still accepted. In a future release, err_perm and err_temp will be
1261 completely removed, which will be a backward incompatibility if the
1262 ACL tests for either of these two old results. Patch contributed by
1263 user bes-internal on the mailing list.
1265 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1267 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1268 selectors, in both main and reject logs.
1270 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1273 JH/07 Add malware type "sock" for talking to simple daemon.
1275 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1277 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1278 routers/transports under cutthrough routing.
1280 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1281 numbers. Touch up "bool" conditional to keep the same definition.
1283 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1285 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1287 JH/12 Expand items in router/transport headers_add or headers_remove lists
1288 individually rather than the list as a whole. Bug 1452.
1290 Required for reasonable handling of multiple headers_ options when
1291 they may be empty; requires that headers_remove items with embedded
1292 colons must have them doubled (or the list-separator changed).
1294 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1295 view the policy declared in the DMARC record. Currently, $dmarc_status
1296 is a combined value of both the record presence and the result of the
1299 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1301 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1302 dnslookup router and the smtp transport (applying to the forward
1305 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1306 of ldap servers used for a specific lookup. Patch provided by Heiko
1309 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1310 New variable $lookup_dnssec_authenticated for observability.
1312 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1313 Patch submitted by Lars Timman.
1315 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1317 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1318 Requires trusted mode and valid format message id, aborts otherwise.
1319 Patch contributed by Heiko Schlichting.
1321 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1322 certextract with support for various fields. Bug 1358.
1324 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1325 is requested by default, modifiable by smtp transport option
1328 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1329 operate on certificate variables to give certificate fingerprints
1330 Also new ${sha256:cert_variable}.
1332 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1334 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1335 Christian Aistleitner.
1337 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1339 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1340 file. Patch from Wolfgang Breyha.
1342 JH/25 Expand the coverage of the delivery $host and $host_address to
1343 client authenticators run in verify callout. Bug 1476.
1345 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1346 align with daemon_smtp_ports. Bug 72.
1348 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1349 support and error reporting did not work properly.
1351 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1352 and is readable. Patch from Andrew Colin Kissa.
1354 TL/14 Enhance documentation of ${run expansion and how it parses the
1355 commandline after expansion, particularly in the case when an
1356 unquoted variable expansion results in an empty value.
1358 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1360 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1361 by renaming away from it. Addresses GH issue 15, reported by
1364 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1366 TL/15 SECURITY: prevent double expansion in math comparison functions
1367 (can expand unsanitized data). Not remotely exploitable.
1374 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1376 PP/02 Make -n do something, by making it not do something.
1377 When combined with -bP, the name of an option is not output.
1379 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1382 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1383 $sender_host_name and config options to manage this, and basic check
1386 PP/05 DSCP support for outbound connections and control modifier for inbound.
1388 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1389 (Only plugin which currently uses this is kerberos4, which nobody should
1390 be using, but we should make it available and other future plugins might
1391 conceivably use it, even though it would break NAT; stuff *should* be
1392 using channel bindings instead).
1394 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1395 name; added for Sendmail compatibility; requires admin caller.
1396 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1397 just ignore it); requires trusted caller.
1398 Also parse but ignore: -Ac -Am -X<logfile>
1401 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1403 TL/02 Add +smtp_confirmation as a default logging option.
1405 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1406 Patch by Magnus Holmgren from 2007-02-20.
1408 TL/04 Bugzilla 1281 - Spec typo.
1409 Bugzilla 1283 - Spec typo.
1410 Bugzilla 1290 - Spec grammar fixes.
1412 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1414 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1416 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1417 mailing list by by Dmitry Isaikin.
1419 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1421 JH/02 Support "G" suffix to numbers in ${if comparisons.
1423 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1425 NM/01 Bugzilla 1197 - Spec typo
1426 Bugzilla 1196 - Spec examples corrections
1428 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1430 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1431 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1434 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1435 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1437 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1438 "acl {{name}{arg}...}", and optional args on acl condition
1441 JH/05 Permit multiple router/transport headers_add/remove lines.
1443 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1445 JH/07 Avoid using a waiting database for a single-message-only transport.
1446 Performance patch from Paul Fisher. Bugzilla 1262.
1448 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1451 JH/09 Add $headers_added variable, with content from use of ACL modifier
1452 add_header (but not yet added to the message). Bugzilla 199.
1454 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1455 Pulled from Bugzilla 817 by Wolfgang Breyha.
1457 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1459 (nb: this is the same fix as in Exim 4.80.1)
1461 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1464 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1467 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1469 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1470 advertise SMTP AUTH mechanism to us, instead of a generic
1471 protocol violation error. Also, make Exim more robust to bad
1472 data from the Dovecot auth socket.
1474 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1476 When a queue runner is handling a message, Exim first routes the
1477 recipient addresses, during which it prunes them based on the retry
1478 hints database. After that it attempts to deliver the message to
1479 any remaining recipients. It then updates the hints database using
1482 So if a recipient address works intermittently, it can get repeatedly
1483 deferred at routing time. The retry hints record remains fresh so the
1484 address never reaches the final cutoff time.
1486 This is a fairly common occurrence when a user is bumping up against
1487 their storage quota. Exim had some logic in its local delivery code
1488 to deal with this. However it did not apply to per-recipient defers
1489 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1491 This change adds a proper retry rule check during routing so that the
1492 final cutoff time is checked against the message's age. We only do
1493 this check if there is an address retry record and there is not a
1494 domain retry record; this implies that previous attempts to handle
1495 the address had the retry_use_local_parts option turned on. We use
1496 this as an approximation for the destination being like a local
1497 delivery, as in LMTP.
1499 I suspect this new check makes the old local delivery cutoff check
1500 redundant, but I have not verified this so I left the code in place.
1502 TF/02 Correct gecos expansion when From: is a prefix of the username.
1504 Test 0254 submits a message to Exim with the header
1508 When I ran the test suite under the user fanf2, Exim expanded
1509 the header to contain my full name, whereas it should have added
1510 a Resent-Sender: header. It erroneously treats any prefix of the
1511 username as equal to the username.
1513 This change corrects that bug.
1515 GF/01 DCC debug and logging tidyup
1516 Error conditions log to paniclog rather than rejectlog.
1517 Debug lines prefixed by "DCC: " to remove any ambiguity.
1519 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1521 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1522 Bug spotted by Jeremy Harris; was flawed since initial commit.
1523 Would have resulted in OCSP responses post-SNI triggering an Exim
1524 NULL dereference and crash.
1526 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1528 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1529 Bug detection, analysis and fix by Samuel Thibault.
1530 Bugzilla 1331, Debian bug #698092.
1532 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1534 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1535 Server implementation by Todd Lyons, client by JH.
1536 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1537 config variable "prdr_enable" controls whether the server
1538 advertises the facility. If the client requests PRDR a new
1539 acl_data_smtp_prdr ACL is called once for each recipient, after
1540 the body content is received and before the acl_smtp_data ACL.
1541 The client is controlled by both of: a hosts_try_prdr option
1542 on the smtp transport, and the server advertisement.
1543 Default client logging of deliveries and rejections involving
1544 PRDR are flagged with the string "PRDR".
1546 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1547 fclose(). Diagnosis by Todd Lyons.
1549 PP/17 Update configure.default to handle IPv6 localhost better.
1550 Patch by Alain Williams (plus minor tweaks).
1553 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1554 This is now consistent with GnuTLS, and is now documented: the
1555 previous undocumented portable approach to treating the option as
1556 unset was to force an expansion failure. That still works, and
1557 an empty string is now equivalent.
1559 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1560 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1561 not performing validation itself.
1563 PP/20 Added force_command boolean option to pipe transport.
1564 Patch from Nick Koston, of cPanel Inc.
1566 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1569 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1571 PP/21 Fix eximon continuous updating with timestamped log-files.
1572 Broken in a format-string cleanup in 4.80, missed when I repaired the
1573 other false fix of the same issue.
1574 Report and fix from Heiko Schlichting.
1577 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1578 Report from Prashanth Katuri.
1580 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1581 It's SecureTransport, so affects any MacOS clients which use the
1582 system-integrated TLS libraries, including email clients.
1584 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1585 using a MIME ACL for non-SMTP local injection.
1586 Report and assistance in diagnosis by Warren Baker.
1588 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1590 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1592 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1593 last id that failed so it may be referenced in subsequent ACL's.
1595 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1598 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1599 ldap library initialization, allowing self-signed CA's to be
1600 used. Also properly sets require_cert option later in code by
1601 using NULL (global ldap config) instead of ldap handle (per
1602 session). Bug diagnosis and testing by alxgomz.
1604 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1605 the src/util/ subdirectory.
1607 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1608 renamed to Transport Post Delivery Action by Jeremy Harris, as
1611 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1612 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1613 redis_servers = needs to be configured which will be used by the redis
1614 lookup. Patch from Warren Baker, of The Packet Hub.
1616 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1618 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1619 hostname or reverse DNS when processing a host list. Used suggestions
1620 from multiple comments on this bug.
1622 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1624 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1625 Missed a few lines, added it to make the runtest require no keyboard
1628 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1629 contains upper case chars. Make router use caseful_local_part.
1631 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1632 support when GnuTLS has been built with p11-kit.
1638 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1640 This, or similar/improved, will also be change PP/11 of 4.82.
1646 PP/01 Handle short writes when writing local log-files.
1647 In practice, only affects FreeBSD (8 onwards).
1648 Bugzilla 1053, with thanks to Dmitry Isaikin.
1650 NM/01 Bugzilla 949 - Documentation tweak
1652 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1655 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1657 PP/02 Implemented gsasl authenticator.
1659 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1661 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1662 `pkg-config foo` for cflags/libs.
1664 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1665 with rest of GSASL and with heimdal_gssapi.
1667 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1668 `pkg-config foo` for cflags/libs for the TLS implementation.
1670 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1671 properties get this fed in as external SSF. A number of robustness
1672 and debugging improvements to the cyrus_sasl authenticator.
1674 PP/08 cyrus_sasl server now expands the server_realm option.
1676 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1677 Patch by Jeremy Harris.
1679 PP/10 Added dbmjz lookup type.
1681 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1683 PP/12 MAIL args handles TAB as well as SP, for better interop with
1684 non-compliant senders.
1685 Analysis and variant patch by Todd Lyons.
1687 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1688 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1689 Patch from Dirk Mueller <dmueller@suse.com>
1691 PP/13 tls_peerdn now print-escaped for spool files.
1692 Observed some $tls_peerdn in wild which contained \n, which resulted
1693 in spool file corruption.
1695 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1696 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1697 or write after TLS renegotiation, which otherwise led to messages
1700 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1701 as a tracking header (ie: a signed header comes before the signature).
1702 Patch from Wolfgang Breyha.
1704 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1705 comma-sep list; embedded commas doubled.
1707 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1709 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1711 Report and patch from Dmitry Banschikov.
1713 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1714 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1715 failures. We appear to now support TLS1.1+ with Exim.
1717 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1718 lets Exim select keys and certificates based upon TLS SNI from client.
1719 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1720 before an outbound SMTP session. New log_selector, +tls_sni.
1722 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1723 NULL dereference. Report and patch from Alun Jones.
1725 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1726 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1727 Not seeing resolver debug output on NetBSD, but suspect this is a
1728 resolver implementation change.
1730 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1731 Left warnings. Added "eximon gdb" invocation mode.
1733 PP/21 Defaulting "accept_8bitmime" to true, not false.
1735 PP/22 Added -bw for inetd wait mode support.
1737 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1738 locate the relevant includes and libraries. Made this the default.
1740 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1741 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1743 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1744 This may cause build issues on older platforms.
1746 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1747 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1748 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1749 Added SNI support via GnuTLS too.
1750 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1752 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1754 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1755 Applied second patch from Janne, implementing suggestion to default
1756 multiple-strings-in-record handling to match SPF spec.
1758 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1760 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1761 read-only, out of scope).
1762 Patch from Wolfgang Breyha, report from Stuart Northfield.
1764 PP/29 Fix three issues highlighted by clang analyser static analysis.
1765 Only crash-plausible issue would require the Cambridge-specific
1766 iplookup router and a misconfiguration.
1767 Report from Marcin Mirosław.
1769 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1771 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1772 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1773 As part of this, removing so much warning spew let me fix some minor
1774 real issues in debug logging.
1776 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1777 assignment on my part. Fixed.
1779 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1780 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1781 Janne Snabb (who went above and beyond: thank you).
1783 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1784 string otherwise requires a connection and a bunch more work and it's
1785 relatively easy to get wrong. Should also expose TLS library linkage
1788 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1789 64-bit ${eval} (JH/03).
1791 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1792 GNU libc to support some of the 64-bit stuff, should not lead to
1793 conflicts. Defined before os.h is pulled in, so if a given platform
1794 needs to override this, it can.
1796 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1797 protection layer was required, which is not implemented.
1798 Bugzilla 1254, patch from Wolfgang Breyha.
1800 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1801 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1802 tls_dhparam take prime identifiers. Also unbreak combination of
1803 OpenSSL+DH_params+TLSSNI.
1805 PP/39 Disable SSLv2 by default in OpenSSL support.
1811 PP/01 Solaris build fix for Oracle's LDAP libraries.
1812 Bugzilla 1109, patch from Stephen Usher.
1814 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1816 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1819 TF/02 Fix a couple more cases where we did not log the error message
1820 when unlink() failed. See also change 4.74-TF/03.
1822 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1823 lock up or crash if it happened to be inside a call to libc when it
1824 got a SIGUSR1 from exiwhat.
1826 The SIGUSR1 handler appends the current process status to the process
1827 log which is later printed by exiwhat. It used to use the general
1828 purpose logging code to do this, but several functions it calls are
1829 not safe for signals.
1831 The new output code in the SIGUSR1 handler is specific to the process
1832 log, and simple enough that it's easy to inspect for signal safety.
1833 Removing some special cases also simplifies the general logging code.
1834 Removing the spurious timestamps from the process log simplifies
1837 TF/04 Improved ratelimit ACL condition.
1839 The /noupdate option has been deprecated in favour of /readonly which
1840 has clearer semantics. The /leaky, /strict, and /readonly update modes
1841 are mutually exclusive. The update mode is no longer included in the
1842 database key; it just determines when the database is updated. (This
1843 means that when you upgrade Exim will forget old rate measurements.)
1845 Exim now checks that the per_* options are used with an update mode that
1846 makes sense for the current ACL. For example, when Exim is processing a
1847 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1848 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1849 must specify per_mail/readonly. If you omit the update mode it defaults to
1850 /leaky where that makes sense (as before) or /readonly where required.
1852 The /noupdate option is now undocumented but still supported for
1853 backwards compatibility. It is equivalent to /readonly except that in
1854 ACLs where /readonly is required you may specify /leaky/noupdate or
1855 /strict/noupdate which are treated the same as /readonly.
1857 A useful new feature is the /count= option. This is a generalization
1858 of the per_byte option, so that you can measure the throughput of other
1859 aggregate values. For example, the per_byte option is now equivalent
1860 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1862 The per_rcpt option has been generalized using the /count= mechanism
1863 (though it's more complicated than the per_byte equivalence). When it is
1864 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1865 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1866 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1867 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1868 Note that using per_rcpt with a non-readonly update mode in more than
1869 one ACL will cause the recipients to be double-counted. (The per_mail
1870 and per_byte options don't have this problem.)
1872 The handling of very low rates has changed slightly. If the computed rate
1873 is less than the event's count (usually one) then this event is the first
1874 after a long gap. In this case the rate is set to the same as this event's
1875 count, so that the first message of a spam run is counted properly.
1877 The major new feature is a mechanism for counting the rate of unique
1878 events. The new per_addr option counts the number of different
1879 recipients that someone has sent messages to in the last time period. It
1880 behaves like per_rcpt if all the recipient addresses are different, but
1881 duplicate recipient addresses do not increase the measured rate. Like
1882 the /count= option this is a general mechanism, so the per_addr option
1883 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1884 example, measure the rate that a client uses different sender addresses
1885 with the options per_mail/unique=$sender_address. There are further
1886 details in the main documentation.
1888 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1890 TF/06 Removed a few PCRE remnants.
1892 TF/07 Automatically extract Exim's version number from tags in the git
1893 repository when doing development or release builds.
1895 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1896 Bugzilla 879. Patch from Paul Fisher.
1898 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1899 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1902 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1904 PP/05 Variable $av_failed, true if the AV scanner deferred.
1905 Bugzilla 1078. Patch from John Horne.
1907 PP/06 Stop make process more reliably on build failure.
1908 Bugzilla 1087. Patch from Heiko Schlittermann.
1910 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1911 Bugzilla 1089. Patch from Heiko Schlittermann.
1913 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1914 Bugzilla 1131. Patch from Holger Weiß.
1916 PP/09 Handle IPv6 addresses with SPF.
1917 Bugzilla 860. Patch from Wolfgang Breyha.
1919 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1921 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1924 PP/11 match_* no longer expand right-hand-side by default.
1925 New compile-time build option, EXPAND_LISTMATCH_RHS.
1926 New expansion conditions, "inlist", "inlisti".
1928 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1930 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1932 PP/14 fix log_write() format string regression from TF/03.
1933 Bugzilla 1152. Patch from Dmitry Isaikin.
1939 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1941 PP/02 Harmonised TLS library version reporting; only show if debugging.
1942 Layout now matches that introduced for other libraries in 4.74 PP/03.
1944 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1946 PP/04 New "dns_use_edns0" global option.
1948 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1951 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1952 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1954 TK/01 Updated PolarSSL code to 0.14.2.
1955 Bugzilla 1097. Patch from Andreas Metzler.
1957 PP/07 Catch divide-by-zero in ${eval:...}.
1958 Fixes bugzilla 1102.
1960 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1963 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1964 format-string attack -- SECURITY: remote arbitrary code execution.
1966 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1967 time unintentionally subject to list matching rules, letting the header
1968 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1969 arbitrary string expansion). This allowed for information disclosure.
1971 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1972 INT_MIN/-1 -- value coerced to INT_MAX.
1978 NM/01 Workaround for PCRE version dependency in version reporting
1981 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1982 This fixes portability to compilers other than gcc, notably
1983 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1985 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1986 makefiles for portability to HP-UX and POSIX correctness.
1988 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1989 Also via indented variable definition in the Makefile.
1990 (Debugging by Oliver Heesakkers).
1992 PP/02 Restore caching of spamd results with expanded spamd_address.
1993 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1995 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1996 Improves build reliability. Fix from: Frank Elsner
1998 NM/02 Fix wide character breakage in the rfc2047 coding
1999 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2001 NM/03 Allow underscore in dnslist lookups
2002 Fixes bug 1026. Patch from Graeme Fowler
2004 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2005 Code patches from Adam Ciarcinski of NetBSD.
2007 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2010 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2011 is logged, to avoid truncation. Patch from John Horne.
2013 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2014 Patch from Jakob Hirsch.
2016 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2017 SQL string expansion failure details.
2018 Patch from Andrey Oktyabrski.
2020 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2021 Patch from Simon Arlott.
2023 PP/09 New lookups functionality failed to compile on old gcc which rejects
2024 extern declarations in function scope.
2025 Patch from Oliver Fleischmann
2027 PP/10 Use sig_atomic_t for flags set from signal handlers.
2028 Check getgroups() return and improve debugging.
2029 Fixed developed for diagnosis in bug 927 (which turned out to be
2032 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2033 Patch from Mark Zealey.
2035 PP/12 Bugzilla 1056: Improved spamd server selection.
2036 Patch from Mark Zealey.
2038 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2039 Based on patch from Heiko Schlittermann.
2041 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2042 Patch from Uwe Doering, sign-off by Michael Haardt.
2044 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2045 variable declaration deep within a block. Bug and patch from
2048 PP/15 lookups-Makefile IRIX compatibility coercion.
2050 PP/16 Make DISABLE_DKIM build knob functional.
2052 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2053 Patch by Simon Arlott
2055 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2056 variable argument macros. Our copy now differs from upstream.
2062 TF/01 Failure to get a lock on a hints database can have serious
2063 consequences so log it to the panic log.
2065 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2066 controlled using the smtp_confirmation log selector.
2068 TF/03 Include the error message when we fail to unlink a spool file.
2070 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2071 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2072 for maintaining out-of-tree patches for some time.
2074 PP/01 Bugzilla 139: Documentation and portability issues.
2075 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2076 Handle per-OS dynamic-module compilation flags.
2078 PP/02 Let /dev/null have normal permissions.
2079 The 4.73 fixes were a little too stringent and complained about the
2080 permissions on /dev/null. Exempt it from some checks.
2081 Reported by Andreas M. Kirchwitz.
2083 PP/03 Report version information for many libraries, including
2084 Exim version information for dynamically loaded libraries. Created
2085 version.h, now support a version extension string for distributors
2086 who patch heavily. Dynamic module ABI change.
2088 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2089 privilege escalation vulnerability whereby the Exim run-time user
2090 can cause root to append content of the attacker's choosing to
2093 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2096 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2097 If dropping privileges for untrusted macros, we disabled normal logging
2098 on the basis that it would fail; for the Exim run-time user, this is not
2099 the case, and it resulted in successful deliveries going unlogged.
2100 Fixed. Reported by Andreas Metzler.
2106 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2107 only prepend for the Resent-* case. Fixes regression introduced in
2108 Exim 4.70 by NM/22 for Bugzilla 607.
2110 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2111 increasing numbers of administrators be bitten by this.
2113 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2115 PP/03 Bugzilla 994: added openssl_options main configuration option.
2117 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2119 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2121 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2123 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2124 without a peer certificate, leading to a segfault because of an
2125 assumption that peers always have certificates. Be a little more
2126 paranoid. Problem reported by Martin Tscholak.
2128 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2129 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2130 NB: ClamAV planning to remove STREAM in "middle of 2010".
2131 CL also introduces -bmalware, various -d+acl logging additions and
2132 more caution in buffer sizes.
2134 PP/09 Implemented reverse_ip expansion operator.
2136 PP/10 Bugzilla 937: provide a "debug" ACL control.
2138 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2140 PP/12 Bugzilla 973: Implement --version.
2142 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2144 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2146 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2148 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2149 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2150 ignore trailing whitespace.
2152 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2154 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2157 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2158 Notification from Dr Andrew Aitchison.
2160 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2161 ExtendedDetectionInfo response format.
2162 Notification from John Horne.
2164 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2167 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2168 XSL and documented dependency on system catalogs, with examples of how
2171 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2174 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2175 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2176 configuration file which is writeable by the Exim user or group.
2178 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2179 of configuration files to cover files specified with the -C option if
2180 they are going to be used with root privileges, not just the default
2183 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2184 option (effectively making it always true).
2186 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2187 files to be used while preserving root privileges.
2189 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2190 that rogue child processes cannot use them.
2192 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2193 run-time user, instead of root.
2195 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2196 Exim run-time user without dropping privileges.
2198 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2199 result string, instead of calling string_vformat() twice with the same
2202 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2203 for other users. Others should always drop root privileges if they use
2204 -C on the command line, even for a whitelisted configure file.
2206 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2208 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2214 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2215 $data_path, and $header_path variables; fixed documentation bugs and
2218 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2219 exipick to access non-standard spools, including the "frozen" queue
2222 NM/01 Bugzilla 965: Support mysql stored procedures.
2223 Patch from Alain Williams
2225 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2227 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2228 Patch from Andreas Metzler
2230 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2231 Patch from Kirill Miazine
2233 NM/05 Bugzilla 671: Added umask to procmail example.
2235 JJ/03 installed exipick 20100323.0, fixing doc bug
2237 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2238 directory. Notification and patch from Dan Rosenberg.
2240 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2242 TK/02 Improve log output when DKIM signing operation fails.
2244 MH/01 Treat the transport option dkim_domain as a colon separated
2245 list, not as a single string, and sign the message with each element,
2246 omitting multiple occurences of the same signer.
2248 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2249 Bugzilla 985, 986. Patch by Simon Arlott
2251 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2252 Patch by Simon Arlott
2254 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2255 MBX locking. Notification from Dan Rosenberg.
2261 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2263 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2265 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2267 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2269 NM/04 Bugzilla 915: Flag broken perl installation during build.
2275 TK/01 Added patch by Johannes Berg that expands the main option
2276 "spamd_address" if it starts with a dollar sign.
2278 TK/02 Write list of recipients to X-Envelope-Sender header when building
2279 the mbox-format spool file for content scanning (suggested by Jakob
2282 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2283 (http://www.dcc-servers.net/) support via dccifd. Activated by
2284 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2286 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2287 by Mark Daniel Reidel <mr@df.eu>.
2289 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2290 When building exim an external PCRE library is now needed -
2291 PCRE is a system library on the majority of modern systems.
2292 See entry on PCRE_LIBS in EDITME file.
2294 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2295 conversation. Added nologin parameter to request.
2296 Patch contributed by Kirill Miazine.
2298 TF/01 Do not log submission mode rewrites if they do not change the address.
2300 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2302 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2303 log files in place. Contributed by Roberto Lima.
2305 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2307 TF/03 Bugzilla 615: When checking the local_parts router precondition
2308 after a local_part_suffix or local_part_prefix option, Exim now
2309 does not use the address's named list lookup cache, since this
2310 contains cached lookups for the whole local part.
2312 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2313 Robert Millan. Documentation is in experimental-spec.txt.
2315 TF/04 Bugzilla 668: Fix parallel build (make -j).
2317 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2319 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2320 Patch provided by Jan Srzednicki.
2322 TF/05 Leading white space used to be stripped from $spam_report which
2323 wrecked the formatting. Now it is preserved.
2325 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2326 that they are available at delivery time.
2328 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2330 TF/08 TLS error reporting now respects the incoming_interface and
2331 incoming_port log selectors.
2333 TF/09 Produce a more useful error message if an SMTP transport's hosts
2334 setting expands to an empty string.
2336 NM/06 Bugzilla 744: EXPN did not work under TLS.
2337 Patch provided by Phil Pennock.
2339 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2340 Patch provided by Richard Godbee.
2342 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2343 acl_smtp_notquit, added index entry.
2345 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2346 Patch provided by Eugene Bujak.
2348 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2349 accept(). Patch provided by Maxim Dounin.
2351 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2352 Patch provided by Phil Pennock.
2354 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2356 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2357 Patch provided by Brad "anomie" Jorsch.
2359 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2360 Patch provided by Dean Brooks.
2362 TK/05 Add native DKIM support (does not depend on external libraries).
2364 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2365 Patch provided by Graeme Fowler.
2367 NM/16 Bugzilla 851: Documentation example syntax fix.
2369 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2371 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2374 NM/19 Bugzilla 745: TLS version reporting.
2375 Patch provided by Phil Pennock.
2377 NM/20 Bugzilla 167: bool: condition support.
2378 Patch provided by Phil Pennock.
2380 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2381 clients. Patch provided by Phil Pennock.
2383 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2384 Patch provided by Brad "anomie" Jorsch.
2386 NM/23 Bugzilla 687: Fix misparses in eximstats.
2387 Patch provided by Heiko Schlittermann.
2389 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2390 Patch provided by Heiko Schlittermann.
2392 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2393 plus update to original patch.
2395 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2397 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2398 Patch provided by David Brownlee.
2400 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2402 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2404 NM/30 Bugzilla 888: TLS documentation bugfixes.
2406 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2408 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2409 Unlike the original bugzilla I have changed all shell scripts in src tree.
2411 NM/33 Bugzilla 898: Transport filter timeout fix.
2412 Patch by Todd Rinaldo.
2414 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2415 Patch by Serge Demonchaux.
2417 NM/35 Bugzilla 39: Base64 decode bug fixes.
2418 Patch by Jakob Hirsch.
2420 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2422 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2424 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2426 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2432 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2433 ALT-N's libdkim that I have put here:
2434 http://duncanthrax.net/exim-experimental/
2436 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2437 were called 'true' and it seems that C99 defines that as a reserved
2438 keyword to be used with 'bool' variable types. That means you could
2439 not include C99-style headers which use bools without triggering
2440 build errors in sieve.c.
2442 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2443 as mailq or other aliases. Changed the --help handling significantly
2444 to do whats expected. exim_usage() emits usage/help information.
2446 SC/01 Added the -bylocaldomain option to eximstats.
2448 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2450 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2452 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2458 PH/01 Another patch from the Sieve maintainer.
2460 PH/02 When an IPv6 address is converted to a string for single-key lookup
2461 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2462 dots are used instead of colons so that keys in lsearch files need not
2463 contain colons. This was done some time before quoting was made available
2464 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2465 (notated using the quote facility) so as to distinguish them from IPv4
2466 keys. This meant that lookups for IP addresses in host lists did not work
2467 for iplsearch lookups.
2469 This has been fixed by arranging for IPv6 addresses to be expressed with
2470 colons if the lookup type is iplsearch. This is not incompatible, because
2471 previously such lookups could never work.
2473 The situation is now rather anomalous, since one *can* have colons in
2474 ordinary lsearch keys. However, making the change in all cases is
2475 incompatible and would probably break a number of configurations.
2477 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2480 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2481 conversion specification without a maximum field width, thereby enabling
2482 a rogue spamd server to cause a buffer overflow. While nobody in their
2483 right mind would setup Exim to query an untrusted spamd server, an
2484 attacker that gains access to a server running spamd could potentially
2485 exploit this vulnerability to run arbitrary code as the Exim user.
2487 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2488 $primary_hostname instead of what libspf2 thinks the hosts name is.
2490 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2491 a directory entry by the name of the lookup key. Previously, if a
2492 symlink pointed to a non-existing file or a file in a directory that
2493 Exim lacked permissions to read, a lookup for a key matching that
2494 symlink would fail. Now it is enough that a matching directory entry
2495 exists, symlink or not. (Bugzilla 503.)
2497 PH/03 The body_linecount and body_zerocount variables are now exported in the
2500 PH/04 Added the $dnslist_matched variable.
2502 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2503 This means they are set thereafter only if the connection becomes
2506 PH/06 Added the client_condition to authenticators so that some can be skipped
2507 by clients under certain conditions.
2509 PH/07 The error message for a badly-placed control=no_multiline_responses left
2510 "_responses" off the end of the name.
2512 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2514 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2515 (without spaces) instead of just copying the configuration text.
2517 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2519 PH/11 Added $max_received_linelength.
2521 PH/12 Added +ignore_defer and +include_defer to host lists.
2523 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2524 way in which PCRE > 7.0 is built.
2526 PH/14 Implemented queue_only_load_latch.
2528 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2529 MAIL command. The effect was to mangle the value on 64-bit systems.
2531 PH/16 Another patch from the Sieve maintainer.
2533 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2535 PH/18 If a system quota error occurred while trying to create the file for
2536 a maildir delivery, the message "Mailbox is full" was not appended to the
2537 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2538 only to a quota excession during the actual writing of the file.
2540 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2541 characters?) which causes problems in log lines. The DN values are now
2542 passed through string_printing() before being added to log lines.
2544 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2545 and InterBase are left for another time.)
2547 PH/21 Added message_body_newlines option.
2549 PH/22 Guard against possible overflow in moan_check_errorcopy().
2551 PH/23 POSIX allows open() to be a macro; guard against that.
2553 PH/24 If the recipient of an error message contained an @ in the local part
2554 (suitably quoted, of course), incorrect values were put in $domain and
2555 $local_part during the evaluation of errors_copy.
2561 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2562 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2565 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2566 issue a MAIL command.
2568 PH/02 In an ACL statement such as
2570 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2572 if a client was not listed at all, or was listed with a value other than
2573 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2574 the condition was not true (as it should be), so access was not denied.
2575 The bug was that the ! inversion was incorrectly passed on to the second
2576 item. This has been fixed.
2578 PH/03 Added additional dnslists conditions == and =& which are different from
2579 = and & when the dns lookup returns more than one IP address.
2581 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2582 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2584 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2585 FSYNC, which compiles an option called disable_fsync that allows for
2586 bypassing fsync(). The documentation is heavily laced with warnings.
2588 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2590 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2591 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2592 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2593 including adding "make clean"; (3) Added -fPIC when compiling the test
2594 dynamically loaded module, to get rid of a warning.
2596 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2597 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2598 = 0s. The bug is otherwise harmless.
2600 PH/07 There was a bug in the dovecot authenticator such that the value of
2601 $auth1 could be overwritten, and so not correctly preserved, after a
2602 successful authentication. This usually meant that the value preserved by
2603 the server_setid option was incorrect.
2605 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2607 PH/09 Installed PCRE release 7.0.
2609 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2610 run for batched SMTP input. It is now run at the start of every message
2611 in the batch. While fixing this I discovered that the process information
2612 (output by running exiwhat) was not always getting set for -bs and -bS
2613 input. This is fixed, and it now also says "batched" for BSMTP.
2615 PH/11 Added control=no_pipelining.
2617 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2618 patch, slightly modified), and move the expansion of helo_data till after
2619 the connection is made in the smtp transport (so it can use these
2622 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2624 PH/14 Added log_selector = +pid.
2626 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2628 PH/16 Add ${if forany and ${if forall.
2630 PH/17 Added dsn_from option to vary the From: line in DSNs.
2632 PH/18 Flush SMTP output before performing a callout, unless control =
2633 no_callout_flush is set.
2635 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2636 was true (the default) a successful delivery failed to delete the retry
2637 item, thus causing premature timeout of the address. The bug is now
2640 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2642 PH/21 Long custom messages for fakedefer and fakereject are now split up
2643 into multiline responses in the same way that messages for "deny" and
2644 other ACL rejections are.
2646 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2647 with slight modification.
2649 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2650 draft, changing the syntax and factoring some duplicate code".
2652 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2653 for deliveries of the second and subsequent messages over the same SMTP
2656 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2657 ${reduce, with only minor "tidies".
2659 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2661 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2662 expansion side effects.
2664 PH/27 When a message times out after an over-quota error from an Exim-imposed
2665 quota, the bounce message says "mailbox is full". This message was not
2666 being given when it was a system quota that was exceeded. It now should
2669 MH/03 Made $recipients available in local_scan(). local_scan() already has
2670 better access to the recipient list through recipients_list[], but
2671 $recipients can be useful in postmaster-provided expansion strings.
2673 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2674 in the case of a MAIL command with additional options following the
2675 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2676 were accidentally chopped off.
2678 PH/29 SMTP synchronization checks are implemented when a command is read -
2679 there is a check that no more input is waiting when there shouldn't be
2680 any. However, for some commands, a delay in an ACL can mean that it is
2681 some time before the response is written. In this time, more input might
2682 arrive, invalidly. So now there are extra checks after an ACL has run for
2683 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2684 pipelining has not been advertised.
2686 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2688 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2689 returned in the SMTP response when smtp_return_error_details was false.
2690 This has been fixed.
2692 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2693 instead of the C I/O that was originally supplied, because problems were
2694 reported on Solaris.
2696 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2697 Exim which did not show up earlier: it was assuming that a call to
2698 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2699 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2700 was a macro that became an assignment, so it seemed to work. This has
2701 changed to a proper function call with a void return, hence the compile
2702 error. Exim's code has been fixed.
2704 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2707 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2709 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2711 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2712 because it turns out to be too big or there is a timeout) there is no
2713 "Completed" line in the log. When some messages of this type were
2714 selected by exigrep, they were listed as "not completed". Others were
2715 picked up by some special patterns. I have improved the selection
2716 criteria to be more general.
2718 PH/38 The host_find_failed option in the manualroute router can now be set
2719 to "ignore", to completely ignore a host whose IP address cannot be
2720 found. If all hosts are ignored, the behaviour is controlled by the new
2721 host_all_ignored option.
2723 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2724 homing or because of multiple MX records with /mx) generated more than
2725 one IP address, and the following item turned out to be the local host,
2726 all the secondary addresses of the first item were incorrectly removed
2727 from the list, along with the local host and any following hosts (which
2728 is what is supposed to happen).
2730 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2731 whoever called Exim into the -H file. In the case of the daemon it was
2732 behaving confusingly. When first started, it used values for whoever
2733 started the daemon, but after a SIGHUP it used the Exim user (because it
2734 calls itself on a restart). I have changed the code so that it now always
2737 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2738 message are rejected with the same error (e.g. no authentication or bad
2739 sender address), and a DATA command is nevertheless sent (as can happen
2740 with PIPELINING or a stupid MUA), the error message that was given to the
2741 RCPT commands is included in the rejection of the DATA command. This is
2742 intended to be helpful for MUAs that show only the final error to their
2745 PH/42 Another patch from the Sieve maintainer.
2747 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2748 Eximstats - Fixed some broken HTML links and added missing column headers
2750 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2751 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2753 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2755 PH/43 Yet another patch from the Sieve maintainer.
2757 PH/44 I found a way to check for a TCP/IP connection going away before sending
2758 the response to the final '.' that terminates a message, but only in the
2759 case where the client has not sent further data following the '.'
2760 (unfortunately, this is allowed). However, in many cases there won't be
2761 any further data because there won't be any more messages to send. A call
2762 to select() can be used: if it shows that the input is "ready", there is
2763 either input waiting, or the socket has been closed. An attempt to read
2764 the next input character can distinguish the two cases. Previously, Exim
2765 would have sent an OK response which the client would never have see.
2766 This could lead to message repetition. This fix should cure that, at
2767 least in a lot of common cases.
2769 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2770 advertised in response to EHLO.
2776 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2777 fixed by 4.65/MH/01 (is this a record?) are fixed:
2779 (i) An empty string was always treated as zero by the numeric comparison
2780 operators. This behaviour has been restored.
2782 (ii) It is documented that the numeric comparison operators always treat
2783 their arguments as decimal numbers. This was broken in that numbers
2784 starting with 0 were being interpreted as octal.
2786 While fixing these problems I realized that there was another issue that
2787 hadn't been noticed. Values of message_size_limit (both the global option
2788 and the transport option) were treated as octal if they started with 0.
2789 The documentation was vague. These values are now always treated as
2790 decimal, and I will make that clear in the documentation.
2796 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2797 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2800 MH/01 Don't check that the operands of numeric comparison operators are
2801 integers when their expansion is in "skipping" mode (fixes bug
2802 introduced by 4.64-PH/07).
2804 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2805 child addresses, Exim now panics and dies. Previously, because the count
2806 is held in a short int, deliveries were likely to be lost. As such a
2807 large number of recipients for a single message is ridiculous
2808 (performance will be very, very poor), I have chosen to impose a limit
2809 rather than extend the field.
2815 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2816 leftover -K file (the existence of which was triggered by #402).
2817 While we were at it, introduced process PID as part of the -K
2818 filename. This should rule out race conditions when creating
2821 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2822 processing considerably. Previous code took too long for large mails,
2823 triggering a timeout which in turn triggers #401.
2825 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2826 in the DK code in transports.c. sendfile() is not really portable,
2827 hence the _LINUX specificness.
2829 TF/01 In the add_headers option to the mail command in an Exim filter,
2830 there was a bug that Exim would claim a syntax error in any
2831 header after the first one which had an odd number of characters
2834 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2835 callout verification, Exim cached a "reject" for the entire domain. This
2836 is correct for most verifications, but it is not correct for a recipient
2837 verification with use_sender or use_postmaster set, because in that case
2838 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2839 case of MAIL FROM:<> rejection from other early rejections (e.g.
2840 rejection of HELO). When verifying a recipient using a non-null MAIL
2841 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2842 Whatever the result of the callout, the value of the domain cache is
2843 left unchanged (for any other kind of callout, getting as far as trying
2844 RCPT means that the domain itself is ok).
2846 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2849 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2850 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2851 session, the daemon ignores ECONNECT errors and logs others; it now
2852 ignores EPIPE as well.
2854 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2855 (quoted-printable decoding).
2857 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2858 later the small subsequent patch to fix an introduced bug.
2860 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2862 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2864 PH/08 An error is now given if message_size_limit is specified negative.
2866 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2867 to be given (somewhat) arbitrary names.
2869 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2872 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2873 miscellaneous code fixes
2875 PH/10 Added the log_reject_target ACL modifier to specify where to log
2878 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2879 hostname. This is wrong, because it relates to the incoming message (and
2880 probably the interface on which it is arriving) and not to the outgoing
2881 callout (which could be using a different interface). This has been
2882 changed to use the value of the helo_data option from the smtp transport
2883 instead - this is what is used when a message is actually being sent. If
2884 there is no remote transport (possible with a router that sets up host
2885 addresses), $smtp_active_hostname is used.
2887 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2888 tweaks were necessary in order to get it to work (see also 21 below):
2889 (a) The code assumed that strncpy() returns a negative number on buffer
2890 overflow, which isn't the case. Replaced with Exim's string_format()
2892 (b) There were several signed/unsigned issues. I just did the minimum
2893 hacking in of casts. There is scope for a larger refactoring.
2894 (c) The code used strcasecmp() which is not a standard C function.
2895 Replaced with Exim's strcmpic() function.
2896 (d) The code set only $1; it now sets $auth1 as well.
2897 (e) A simple test gave the error "authentication client didn't specify
2898 service in request". It would seem that Dovecot has changed its
2899 interface. Fortunately there's a specification; I followed it and
2900 changed what the client sends and it appears to be working now.
2902 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2905 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2906 address A is aliased to B and C, where B exists and C does not. Without
2907 -v the output is "A verified" because verification stops after a
2908 successful redirection if more than one address is generated. However,
2909 with -v the child addresses are also verified. Exim was outputting "A
2910 failed to verify" and then showing the successful verification for C,
2911 with its parentage. It now outputs "B failed to verify", showing B's
2912 parentage before showing the successful verification of C.
2914 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2915 look up a TXT record in a specific list after matching in a combined
2918 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2919 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2920 they consult the DNS. I had assumed they would set it the way they
2921 wanted; and indeed my experiments on Linux seem to show that in some
2922 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2923 To be on the safe side, however, I have now made the interface to
2924 host_find_byname() similar to host_find_bydns(), with an argument
2925 containing the DNS resolver options. The host_find_byname() function now
2926 sets these options at its start, just as host_find_bydns() does. The smtp
2927 transport options dns_qualify_single and dns_search_parents are passed to
2928 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2929 of host_find_byname() use the default settings of RES_DEFNAMES
2930 (qualify_single) but not RES_DNSRCH (search_parents).
2932 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2933 spool_read_header() do less string testing, by means of a preliminary
2934 switch on the second character of optional "-foo" lines. (This is
2935 overdue, caused by the large number of possibilities that now exist.
2936 Originally there were few.) While I was there, I also converted the
2937 str(n)cmp tests so they don't re-test the leading "-" and the first
2938 character, in the hope this might squeeze out yet more improvement.
2940 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2941 flag allowing group syntax was set by the header_syntax check but not
2942 turned off, possible causing trouble later; (2) The flag was not being
2943 set at all for the header_verify test, causing "group"-style headers to
2944 be rejected. I have now set it in this case, and also caused header_
2945 verify to ignore an empty address taken from a group. While doing this, I
2946 came across some other cases where the code for allowing group syntax
2947 while scanning a header line wasn't quite right (mostly, not resetting
2948 the flag correctly in the right place). These bugs could have caused
2949 trouble for malformed header lines. I hope it is now all correct.
2951 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2952 with the "reply" argument non-NULL. The code, however (which originally
2953 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2954 but it didn't always do it. This confused somebody who was copying the
2955 code for some other use. I have removed all the tests.
2957 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2958 feature that was used to support insecure browsers during the U.S. crypto
2959 embargo. It requires special client support, and Exim is probably the
2960 only MTA that supported it -- and would never use it because real RSA is
2961 always available. This code has been removed, because it had the bad
2962 effect of slowing Exim down by computing (never used) parameters for the
2963 RSA_EXPORT functionality.
2965 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2966 authenticator to fail if there's a tab character in the incoming data
2967 (there should never be unless someone is messing about, as it's supposed
2968 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2969 if the connection is using TLS or if the remote IP is the same as the
2970 local IP, and the "valid-client-cert option" if a client certificate has
2973 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2974 authenticators. This can be used for authorization after authentication
2975 succeeds. (In the case of plaintext, it servers for both authentication
2978 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2979 if any retry times were supplied.
2981 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2982 connection, where there is no client IP address to check. In this
2983 situation, the verify now always succeeds.
2985 PH/25 Applied John Jetmore's -Mset patch.
2987 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2989 PH/27 In a string expansion for a processed (not raw) header when multiple
2990 headers of the same name were present, leading whitespace was being
2991 removed from all of them, but trailing whitespace was being removed only
2992 from the last one. Now trailing whitespace is removed from each header
2993 before concatenation. Completely empty headers in a concatenation (as
2994 before) are ignored.
2996 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2997 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2999 PH/29 [Removed. This was a change that I later backed out, and forgot to
3000 correct the ChangeLog entry (that I had efficiently created) before
3001 committing the later change.]
3003 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3004 address errors (4xx response to RCPT) over the same connection as other
3005 messages routed to the same hosts. Such deliveries are always "forced",
3006 so retry times are not inspected. This resulted in far too many retries
3007 for the affected addresses. The effect occurred only when there were more
3008 hosts than the hosts_max_try setting in the smtp transport when it had
3009 the 4xx errors. Those hosts that it had tried were not added to the list
3010 of hosts for which the message was waiting, so if all were tried, there
3011 was no problem. Two fixes have been applied:
3013 (i) If there are any address or message errors in an SMTP delivery, none
3014 of the hosts (tried or untried) are now added to the list of hosts
3015 for which the message is waiting, so the message should not be a
3016 candidate for sending over the same connection that was used for a
3017 successful delivery of some other message. This seems entirely
3018 reasonable: after all the message is NOT "waiting for some host".
3019 This is so "obvious" that I'm not sure why it wasn't done
3020 previously. Hope I haven't missed anything, but it can't do any
3021 harm, as the worst effect is to miss an optimization.
3023 (ii) If, despite (i), such a delivery is accidentally attempted, the
3024 routing retry time is respected, so at least it doesn't keep
3025 hammering the server.
3027 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3028 in ${readsocket because some servers need this prod.
3030 PH/32 Added some extra debug output when updating a wait-xxx database.
3032 PH/33 The hint "could be header name not terminated by colon", which has been
3033 given for certain expansion errors for a long time, was not being given
3034 for the ${if def:h_colon_omitted{... case.
3036 PH/34 The spec says: "With one important exception, whenever a domain list is
3037 being scanned, $domain contains the subject domain." There was at least
3038 one case where this was not true.
3040 PH/35 The error "getsockname() failed: connection reset by peer" was being
3041 written to the panic log as well as the main log, but it isn't really
3042 panic-worthy as it just means the connection died rather early on. I have
3043 removed the panic log writing for the ECONNRESET error when getsockname()
3046 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3047 runs only) independently of the message's sender address. This meant
3048 that, if the 4xx error was in fact related to the sender, a different
3049 message to the same recipient with a different sender could confuse
3050 things. In particular, this can happen when sending to a greylisting
3051 server, but other circumstances could also provoke similar problems.
3052 I have changed the default so that the retry time for these errors is now
3053 based a combination of the sender and recipient addresses. This change
3054 can be overridden by setting address_retry_include_sender=false in the
3057 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3058 remote server are returned as part of bounce messages. This was not
3059 happening for LMTP over a pipe (the lmtp transport), but now it is the
3060 same for both kinds of LMTP.
3062 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3063 in header lines that were in fact CNAMEs. This is no longer the case.
3065 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3066 and queue runs started by the daemon processed all messages. This has
3067 been fixed so that -R and -S can now usefully be given with -q<time>.
3069 PH/40 Import PCRE release 6.7 (fixes some bugs).
3071 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3073 PH/42 Give an error if -q is specified more than once.
3075 PH/43 Renamed the variables $interface_address and $interface_port as
3076 $received_ip_address and $received_port, to make it clear that these
3077 values apply to message reception, and not to the outgoing interface when
3078 a message is delivered. (The old names remain recognized, of course.)
3080 PH/44 There was no timeout on the connect() call when using a Unix domain
3081 socket in the ${readsocket expansion. There now is.
3083 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3084 be meaningful with "accept".
3086 SC/01 Eximstats V1.43
3087 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3089 SC/02 Eximstats V1.44
3090 Use a glob alias rather than an array ref in the generated
3091 parser. This improves both readability and performance.
3093 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3094 Collect SpamAssassin and rejection statistics.
3095 Don't display local sender or destination tables unless
3096 there is data to show.
3097 Added average volumes into the top table text output.
3099 SC/04 Eximstats V1.46
3100 Collect data on the number of addresses (recipients)
3101 as well as the number of messages.
3103 SC/05 Eximstats V1.47
3104 Added 'Message too big' to the list of mail rejection
3105 reasons (thanks to Marco Gaiarin).
3107 SC/06 Eximstats V1.48
3108 Mainlog lines which have GMT offsets and are too short to
3109 have a flag are now skipped.
3111 SC/07 Eximstats V1.49 (Alain Williams)
3112 Added the -emptyok flag.
3114 SC/08 Eximstats V1.50
3115 Fixes for obtaining the IP address from reject messages.
3117 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3118 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3119 whitespace changes from 4.64-PH/27
3121 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3124 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3125 are found, allow negative numbers in numeric criteria)
3127 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3129 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3130 to match changes made in 4.64-PH/43
3132 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3134 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3135 "domain" is actually an IP address, and force a failure. This locks out
3136 those revolvers/nameservers that support "A-for-A" lookups, in
3137 contravention of the specifications.
3139 PH/48 When a host name was looked up from an IP address, and the subsequent
3140 forward lookup of the name timed out, the host name was left in
3141 $sender_host_name, contrary to the specification.
3143 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3144 restricted to single-key lookups, Exim was not diagnosing an error if
3145 * or *@ was used with a query-style lookup.
3147 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3149 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3150 long ago, but noone interested enough thought of it. Let's just say that
3151 the "1.1" means that there are some new functions that weren't there at
3152 some point in the past.
3154 PH/51 Error processing for expansion failure of helo_data from an smtp
3155 transport during callout processing was broken.
3157 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3158 tested/used via the -bh/-bhc/-bs options.
3160 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3161 bug, fixed in subsequent PCRE releases).
3163 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3164 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3166 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3172 SC/01 Use a glob alias rather than an array ref in eximstats generated
3173 parser. This improves both readability and performance.
3175 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3176 Don't display local sender or destination tables in eximstats unless
3177 there is data to show.
3178 Added average volumes into the eximstats top table text output.
3180 SC/03 Collect data on the number of addresses (recipients) as well
3181 as the number of messages in eximstats.
3183 TF/01 Correct an error in the documentation for the redirect router. Exim
3184 does (usually) call initgroups() when daemonizing.
3186 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3187 with consistent privilege compared to when running as a daemon.
3189 TF/03 Note in the spec that $authenticated_id is not set for local
3190 submissions from trusted users.
3192 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3193 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3195 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3196 by adding some example configuration directives to the default
3197 configuration file. A little bit of work is required to uncomment the
3198 directives and define how usernames and passwords are checked, but
3199 there is now a framework to start from.
3201 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3202 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3203 without this. I don't know how relevant this is to other LDAP libraries.
3205 PH/02 Add the verb name to the "unknown ACL verb" error.
3207 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3209 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3211 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3212 directory not expanded when it should be if an expanded home directory
3213 was set for the address (which is overridden by the transport).
3215 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3218 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3219 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3220 because it is too late at that time, and has no effect.
3222 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3223 security issue with \' (bugzilla #107). I could not use the
3224 PQescapeStringConn() function, because it needs a PGconn value as one of
3227 PH/08 When testing addresses using -bt, indicate those final addresses that
3228 are duplicates that would not cause an additional delivery. At least one
3229 person was confused, thinking that -bt output corresponded to deliveries.
3230 (Suppressing duplicates isn't a good idea as you lose the information
3231 about possibly different redirections that led to the duplicates.)
3233 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3234 systems where poll() doesn't work, in particular OS X.
3236 PH/10 Added more information to debugging output for retry time not reached.
3238 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3239 operations in malware.c.
3241 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3244 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3245 syslog_duplication set false, log lines that would normally be written
3246 both the the main log and to the reject log were not written to syslog at
3249 PH/14 In the default configuration, change the use of "message" in ACL warn
3250 statements to "add_header".
3252 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3253 not followed by a command (e.g. "seen endif").
3255 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3256 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3259 PH/17 Added extra conditions to the default value of delay_warning_condition
3263 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3264 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3265 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3268 The Auto-Submitted: and various List- headers are standardised, whereas I
3269 don't think Precedence: ever was.
3271 PH/18 Refactored debugging code in route_finduser() to show more information,
3272 in particular, the error code if getpwnam() issues one.
3274 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3275 This is apparently needed in addition to the PH/07 change above to avoid
3276 any possible encoding problems.
3278 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3279 but not after initializing Perl.
3281 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3282 output them only if debugging. By default they are written stderr,
3283 apparently, which is not desirable.
3285 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3288 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3291 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3293 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3294 authenticated or an ident call has been made. Suppress the default
3295 values for $authenticated_id and $authenticated_sender (but permit -oMai
3296 and -oMas) when testing with -bh.
3298 PH/24 Re-jigged the order of the tests in the default configuration so that the
3299 tests for valid domains and recipients precede the DNS black list and CSA
3300 tests, on the grounds that those ones are more expensive.
3302 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3303 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3304 This bug exists in every version of Exim that I still have, right back to
3307 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3308 However, an attempt to turn on case-sensitivity in a regex key by
3309 including (?-i) didn't work because the subject string was already
3310 lowercased, and the effects were non-intuitive. It turns out that a
3311 one-line patch can be used to allow (?-i) to work as expected.
3317 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3318 other effects) broke the use of negated acl sub-conditions.
3320 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3323 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3324 "Deny" causes Exim to reject the incoming connection with a 554 error.
3325 Unfortunately, if there is a major crisis, such as a disk failure,
3326 tcp-wrappers gives "deny", whereas what one would like would be some
3327 kind of temporary error. A kludge has been added to help with this.
3328 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3329 554 error is used if errno is still zero or contains ENOENT (which occurs
3330 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3333 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3335 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3336 errors. Otherwise a message that provokes a temporary error (when other
3337 messages do not) can cause a whole host to time out.
3339 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3340 addresses were routed directly to files or pipes from a redirect router.
3341 File deliveries just didn't batch; pipe deliveries might have suffered
3344 PH/06 A failure to get a lock for a hints database would erroneously always say
3345 "Failed to get write lock", even when it was really a read lock.
3347 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3348 of 0600. This has been changed to use the value of the lockfile_mode
3349 option (which defaults to 0600).
3351 PH/08 Applied small patch from the Sieve maintainer.
3353 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3354 folder from quota calculations, a direct delivery into this folder messed
3355 up the contents of the maildirsize file. This was because the regex was
3356 used only to exclude .Trash (or whatever) when the size of the mailbox
3357 was calculated. There was no check that a delivery was happening into an
3358 excluded directory. This bug has been fixed by ignoring all quota
3359 processing for deliveries into excluded directories.
3361 PH/10 Added the maildirfolder_create_regex option to appendfile.
3367 PH/01 The code for finding all the local interface addresses on a FreeBSD
3368 system running IPv6 was broken. This may well have applied to all BSD
3369 systems, as well as to others that have similar system calls. The broken
3370 code found IPv4 interfaces correctly, but gave incorrect values for the
3371 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3372 that it would not match correctly against @[] and not recognize the IPv6
3375 PH/02 The ipliteral router was not recognizing addresses of the form user@
3376 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3378 PH/03 Added disable_ipv6.
3380 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3381 decoded form, because it is most often used to construct To: headers
3382 lines in autoreplies, and the decoded form may well be syntactically
3383 invalid. However, $reply_address has leading white space removed, and all
3384 newlines turned into spaces so that the autoreply transport does not
3387 PH/05 If group was specified without a user on a router, and no group or user
3388 was specified on a transport, the group from the router was ignored.
3390 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3391 for visible compile-time settings that can be used to change these
3392 numbers, for those that want even more. Backwards compatibility with old
3393 spool files has been maintained. However, going back to a previous Exim
3394 release will lost any variables that are in spool files.
3396 PH/07 Two small changes when running in the test harness: increase delay when
3397 passing a TCP/IP connection to a new process, in case the original
3398 process has to generate a bounce, and remove special handling of
3399 127.0.0.2 (sic), which is no longer necessary.
3401 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3402 be the same on different OS.
3404 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3407 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3408 whether --show-vars was specified or not
3410 JJ/02 exipick: Added support for new ACL variable spool format introduced
3413 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3414 syntactically invalid From: or Reply-to: line, and a filter used this to
3415 generate an autoreply, and therefore failed to obtain an address for the
3416 autoreply, Exim could try to deliver to a non-existent relative file
3417 name, causing unrelated and misleading errors. What now happens is that
3418 it logs this as a hard delivery error, but does not attempt to create a
3421 PH/11 The exinext utility has a -C option for testing purposes, but although
3422 the given file was scanned by exinext itself; it wasn't being passed on
3423 when Exim was called.
3425 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3426 an end-of-file indication when reading a command response.
3428 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3429 compiled. In many other places in Exim, IPv6 addresses are always
3430 recognized, so I have changed this. It also means that IPv4 domain
3431 literals of the form [IPV4:n.n.n.n] are now always recognized.
3433 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3434 used if the router is not running as root, for example, when verifying at
3435 ACL time, or when using -bh. The debugging output from this situation was
3436 non-existent - all you got was a failure to exec. I have made two
3439 (a) Failures to set uid/gid, the current directory, or a process leader
3440 in a subprocess such as that created by queryprogram now generate
3441 suitable debugging output when -d is set.
3443 (b) The queryprogram router detects when it is not running as root,
3444 outputs suitable debugging information if -d is set, and then runs
3445 the subprocess without attempting to change uid/gid.
3447 PH/15 Minor change to Makefile for building test_host (undocumented testing
3450 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3451 additional section of a DNS packet that returns MX or SRV records.
3452 Instead, it always explicitly searches for A/AAAA records. This avoids
3453 major problems that occur when a DNS server includes only records of one
3454 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3455 fixed another bug: if SRV records were looked up and the corresponding
3456 address records were *not* found in the additional section, the port
3457 values from the SRV records were lost.
3459 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3460 using the correct key (the original address) when searching the retry
3461 rules in order to find which one to use for generating the retry hint.
3463 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3464 adding the default one. Similarly, if it contains a Reply-To: header, the
3465 errors_reply_to option, if set, is not used.
3467 PH/19 When calculating a retry time, Exim used to measure the "time since
3468 failure" by looking at the "first failed" field in the retry record. Now
3469 it does not use this if it is later than than the arrival time of the
3470 message. Instead it uses the arrival time. This makes for better
3471 behaviour in cases where some deliveries succeed, thus re-setting the
3472 "first failed" field. An example is a quota failure for a huge message
3473 when small messages continue to be delivered. Without this change, the
3474 "time since failure" will always be short, possible causing more frequent
3475 delivery attempts for the huge message than are intended.
3476 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3478 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3479 $1, $2, $3) because the numerical variables can be reset during some
3480 expansion items (e.g. "match"), thereby losing the authentication data.
3482 PH/21 Make -bV show the size of off_t variables so that the test suite can
3483 decide whether to run tests for quotas > 2G.
3485 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3486 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3487 filecount value is greater than 2G or if a quota value is greater than 2G
3488 on a system where the size of off_t is not greater than 4, a panic error
3491 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3492 never match. The debug and -bh output now contains an explicit error
3493 message indicating a malformed IPv4 address or mask.
3495 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3496 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3497 PH/23 above applies.
3499 PH/25 Do not write to syslog when running in the test harness. The only
3500 occasion when this arises is a failure to open the main or panic logs
3501 (for which there is an explicit test).
3503 PH/26 Added the /no_tell option to "control=freeze".
3505 PH/27 If a host name lookup failed very early in a connection, for example, if
3506 the IP address matched host_lookup and the reverse lookup yielded a name
3507 that did not have a forward lookup, an error message of the form "no IP
3508 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3509 could be logged. Now it outputs the IP address instead of "NULL".
3511 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3512 allows the sender and the authenticated sender to be set when
3513 submitting a message from within Exim. Since child_open_exim() is
3514 documented for local_scan(), the new function should be too.
3516 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3517 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3518 results in an empty string is now treated as unset.
3520 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3522 PH/31 Added sender_verify_fail logging option.
3524 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3525 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3526 tidied the source and removed it altogether.
3528 PH/33 When a queue run was abandoned because the load average was too high, a
3529 log line was always written; now it is written only if the queue_run log
3530 selector is set. In addition, the log line for abandonment now contains
3531 information about the queue run such as the pid. This is always present
3532 in "start" and "stop" lines but was omitted from the "abandon" line.
3534 PH/34 Omit spaces between a header name and the colon in the error message that
3535 is given when verify = headers_syntax fails (if there are lots of them,
3536 the message gets confusing).
3538 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3539 names, as there are now some PTR records that contain slashes. This check
3540 is only to protect against broken name servers that fall over on strange
3541 characters, so the fact that it applies to all lookups doesn't matter.
3543 PH/36 Now that the new test suite is complete, we can remove some of the
3544 special code in Exim that was needed for the old test suite. For example,
3545 sorting DNS records because real resolvers return them in an arbitrary
3546 order. The new test suite's fake resolver always returns records in the
3549 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3550 bounces) except when queue_only is set, to avoid logging races between
3551 the different processes.
3553 PH/38 Panic-die if .include specifies a non-absolute path.
3555 PH/39 A tweak to the "H" retry rule from its user.
3557 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3558 a label. They prevented compilation on older perls.
3560 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3561 a warning to be raised on newish perls.
3563 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3564 on queue. Changes to match documented behaviour of showing count of
3565 messages matching specified criteria.
3567 PH/40 Changed the default ident timeout from 30s to 5s.
3569 PH/41 Added support for the use of login_cap features, on those BSD systems
3570 that have them, for controlling the resources used by pipe deliveries.
3572 PH/42 The content-scanning code uses fopen() to create files in which to put
3573 message data. Previously it was not paying any attention to the mode of
3574 the files. Exim runs with umask(0) because the rest of the code creates
3575 files with open(), and sets the required mode explicitly. Thus, these
3576 files were ending up world-writeable. This was not a big issue, because,
3577 being within the spool directory, they were not world-accessible. I have
3578 created a function called modefopen, which takes an additional mode
3579 argument. It sets umask(777), creates the file, chmods it to the required
3580 mode, then resets the umask. All the relevant calls to fopen() in the
3581 content scanning code have been changed to use this function.
3583 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3584 to 24 hours. This avoids potential overflow problems when processing G
3585 and H retry rules. I suspect nobody ever tinkers with this value.
3587 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3589 PH/45 When the plaintext authenticator is running as a client, the server's
3590 challenges are checked to ensure they are valid base64 strings. By
3591 default, the authentication attempt is cancelled if an invalid string is
3592 received. Setting client_ignore_invalid_base64 true ignores these errors.
3593 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3594 they are received. Thus, the responses can be made to depend on the
3595 challenges. If an invalid string is ignored, an empty string is placed in
3598 PH/46 Messages that are created by the autoreply transport now contains a
3599 References: header, in accordance with RFCs 2822 and 3834.
3601 PH/47 Added authenticated_sender_force to the smtp transport.
3603 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3605 PH/49 Installed latest patch from the Sieve maintainer.
3607 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3608 was also set, the appendfile transport was unnecessarily scanning a
3609 directory of message files (e.g. for maildir delivery) to find the count
3610 of files (along with the size), even though it did not need this
3611 information. It now does the scan only if it needs to find either the
3612 size of the count of files.
3614 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3616 PH/52 Two bugs concerned with error handling when the smtp transport is
3619 (i) Exim was not creating retry information for temporary errors given
3620 for individual recipients after the DATA command when the smtp transport
3621 was used in LMTP mode. This meant that they could be retried too
3622 frequently, and not timed out correctly.
3624 (ii) Exim was setting the flag that allows error details to be returned
3625 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3626 recipients that were returned after the DATA command.
3628 PH/53 This is related to PH/52, but is more general: for any failing address,
3629 when detailed error information was permitted to be returned to the
3630 sender, but the error was temporary, then after the final timeout, only
3631 "retry timeout exceeded" was returned. Now it returns the full error as
3632 well as "retry timeout exceeded".
3634 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3635 do this, and (what is worse) MTAs that accept it.
3637 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3638 will now be deprecated.
3640 PH/56 New os.c-cygwin from the Cygwin maintainer.
3642 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3643 formats (previously only available in exim formats via -bpr, -bpru,
3644 and -bpra. Now also available in native and exiqgrep formats)
3646 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3647 with very large, slow to parse queues
3649 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3651 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3653 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3654 responses to authentication challenges, though it was showing the
3655 challenges; (ii) I've removed the CR characters from the debug output for
3658 PH/58 Allow for the insertion of a newline as well as a space when a string
3659 is turned into more than one encoded-word during RFC 2047 encoding. The
3660 Sieve code now uses this.
3662 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3663 data_4xx, lost_connection, tls_required.
3665 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3666 message was being sent as an SMTP response.
3668 PH/61 Add -l and -k options to exicyclog.
3670 PH/62 When verifying, if an address was redirected to one new address, so that
3671 verification continued, and the new address failed or deferred after
3672 having set something in $address_data, the value of $address_data was not
3673 passed back to the ACL. This was different to the case when no
3674 redirection occurred. The value is now passed back in both cases.
3676 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3677 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3678 use login_cap.h, so on its own it isn't the distinguishing feature. The
3679 new name refers directly to the setclassresources() function.
3681 PH/65 Added configuration files for NetBSD3.
3683 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3685 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3686 is preferred over IPv4.
3688 PH/68 The bounce_return_message and bounce_return_body options were not being
3689 honoured for bounces generated during the reception of non-SMTP messages.
3690 In particular, this applied to messages rejected by the ACL. This bug has
3691 been fixed. However, if bounce_return_message is true and bounce_return_
3692 body is false, the headers that are returned for a non-SMTP message
3693 include only those that have been read before the error was detected.
3694 (In the case of an ACL rejection, they have all been read.)
3696 PH/69 The HTML version of the specification is now built in a directory called
3697 spec_html instead of spec.html, because the latter looks like a path with
3698 a MIME-type, and this confuses some software.
3700 PH/70 Catch two compiler warnings in sieve.c.
3702 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3703 function verify_get_ident() calls ip_connect() to connect a socket, but
3704 if the "connect()" function timed out, ip_connect() used to close the
3705 socket. However, verify_get_ident() also closes the socket later, and in
3706 between Exim writes to the log, which may get opened at this point. When
3707 the socket was closed in ip_connect(), the log could get the same file
3708 descriptor number as the socket. This naturally causes chaos. The fix is
3709 not to close the socket in ip_connect(); the socket should be closed by
3710 the function that creates it. There was only one place in the code where
3711 this was missing, in the iplookup router, which I don't think anybody now
3712 uses, but I've fixed it anyway.
3714 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3715 well as to direct DNS lookups. Otherwise the handling of names in host
3716 lists is inconsistent and therefore confusing.
3722 PH/01 Two changes to the default runtime configuration:
3724 (1) Move the checks for relay_from_hosts and authenticated clients from
3725 after to before the (commented out) DNS black list checks.
3727 (2) Add control=submission to the relay_from_hosts and authenticated
3728 clients checks, on the grounds that messages accepted by these
3729 statements are most likely to be submissions.
3731 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3733 (1) Generate an error if the third argument for the ${prvs expansion is
3736 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3739 (3) Reset the variables that are obtained from the first argument of
3740 ${prvscheck and used in the second argument before leaving the code,
3741 because their memory is reclaimed, so using them afterwards may do
3744 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3745 one (it's much easier than Tom thought :-).
3747 (5) Because of (4), we can now allow for the use of $prvscheck_result
3748 inside the third argument.
3750 PH/03 For some reason, the default setting of PATH when running a command from
3751 a pipe transport was just "/usr/bin". I have changed it to
3754 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3755 anything to be listed in the output from -bV.
3757 PH/05 When a filter generated an autoreply, the entire To: header line was
3758 quoted in the delivery log line, like this:
3760 => >A.N.Other <ano@some.domain> <original@ddress> ...
3762 This has been changed so that it extracts the operative address. There
3763 may be more than one such address. If so, they are comma-separated, like
3766 => >ano@some.domain,ona@other.domain <original@ddress> ...
3768 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3769 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3770 not being looked up in the rDNS to get a host name, Exim was showing the
3771 IP address twice in Received: lines, even though the IP addresses were
3772 identical. For example:
3774 Received: from [1.2.3.4] (helo=[1.2.3.4])
3776 However, if the real host name was known, it was omitting the HELO data
3777 if it matched the actual IP address. This has been tidied up so that it
3778 doesn't show the same IP address twice.
3780 PH/07 When both +timestamp and +memory debugging was on, the value given by
3781 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3782 called by the debug printing, thereby overwriting the timestamp buffer.
3783 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3785 PH/08 When the original message was included in an autoreply transport, it
3786 always said "this is a copy of the message, including all the headers",
3787 even if body_only or headers_only was set. It now gives an appropriate
3790 PH/09 Applied a patch from the Sieve maintainer which:
3792 o fixes some comments
3793 o adds the (disabled) notify extension core
3794 o adds some debug output for the result of if/elsif tests
3795 o points to the current vacation draft in the documentation
3796 and documents the missing references header update
3800 o fixes a bug in processing the envelope test (when testing
3801 multiple envelope elements, the last element determined the
3804 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3805 Electronic Mail") by including:
3807 Auto-submitted: auto-generated
3809 in the messages that it generates (bounce messages and others, such as
3810 warnings). In the case of bounce messages for non-SMTP messages, there was
3811 also a typo: it was using "Auto_submitted" (underscore instead of
3812 hyphen). Since every message generated by Exim is necessarily in response
3813 to another message, thes have all been changed to:
3815 Auto-Submitted: auto-replied
3817 in accordance with these statements in the RFC:
3819 The auto-replied keyword:
3821 - SHOULD be used on messages sent in direct response to another
3822 message by an automatic process,
3824 - MUST NOT be used on manually-generated messages,
3826 - MAY be used on Delivery Status Notifications (DSNs) and Message
3827 Disposition Notifications (MDNs),
3829 - MUST NOT be used on messages generated by automatic or periodic
3830 processes, except for messages which are automatic responses to
3833 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3834 to the default Received: header definition.
3836 PH/12 Added log selector acl_warn_skipped (default on).
3838 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3839 variables because (a) they are in the wrong storage pool and (b) even if
3840 they were copied, it wouldn't work properly because of the caching.
3842 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3843 checking when decoding. Apparently there are clients that generate
3844 overlong encoded strings. Why am I not surprised?
3846 PH/15 If the first argument of "${if match_address" was not empty, but did not
3847 contain an "@" character, Exim crashed. Now it writes a panic log message
3848 and treats the condition as false.
3850 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3852 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3853 extension "envelope-auth". The code is finished and in agreement with
3854 other implementations, but there is no documentation so far and in fact,
3855 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3856 not changing the active code.
3858 Print executed "if" and "elsif" statements when debugging is used. This
3859 helps a great deal to understand what a filter does.
3861 Document more things not specified clearly in RFC3028. I had all this
3862 sorted out, when out of a sudden new issues came to my mind. Oops."
3864 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3867 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3868 canonical form (as documented). However, after a host name lookup from
3869 the IP address, check_host() was doing a simple string comparison with
3870 addresses acquired from the DNS when checking that the found name did
3871 have the original IP as one of its addresses. Since any found IPv6
3872 addresses are likely to be in abbreviated form, the comparison could
3873 fail. Luckily, there already exists a function for doing the comparison
3874 by converting both addresses to binary, so now that is used instead of
3875 the text comparison.
3877 PH/20 There was another similar case to PH/19, when a complete host name was
3878 given in a host list; looking up its IP address could give an abbreviated
3879 form, whereas the current host's name might or might not be abbreviated.
3880 The same fix has been applied.
3886 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3887 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3890 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3891 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3893 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3895 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3896 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3897 Also, for "Auto-Submitted", treat anything other than "no" as
3898 non-personal, in accordance with RFC 3834. (Previously it treated
3899 anything starting "auto-" as non-personal.)
3901 TF/01 The control=submission/name=... option had a problem with syntax
3902 errors if the name included a slash character. The /name= option
3903 now slurps the rest of the string, so it can include any characters
3904 but it must come last in the list of options (after /sender_retain
3907 PH/05 Some modifications to the interface to the fake nameserver for the new
3915 TK/01 Added the "success_on_redirect" address verification option. See
3916 NewStuff for rationale and an example.
3918 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3920 PH/02 Patch to exigrep to allow it to work on syslog lines.
3922 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3923 fread() to skip over the body file's header line, because in Cygwin the
3924 header line is locked and is inaccessible.
3926 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3927 co-exist for some time) to make it clear that it is the Exim ID that is
3928 referenced, not the Message-ID: header line.
3930 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3931 string_format() function, because snprintf() does not exist on all
3934 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3935 ${stat: expansion item.
3937 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3938 protocol synchronization error", to keep the pedants happy.
3940 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3941 well as for IRIX systems, when gcc is being used. See the host.c source
3944 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3946 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3949 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3950 transport and to the smtp transport in LMTP mode.
3952 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3954 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3955 run a filter in a subprocess. This could lead to confusion in subsequent
3956 lookups in the parent process. There should also be a search_tidyup() at
3957 the end of the subprocess.
3959 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3960 only if the host matched helo_try_verify_hosts, which caused the
3961 verification to occur when the EHLO/HELO command was issued. The ACL just
3962 tested the remembered result. Now, if a previous verification attempt has
3963 not happened, "verify = helo" does it there and then.
3965 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3967 TK/03 Fix log output including CR from clamd.
3969 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3970 exist provoked a memory error which could cause a segfault.
3972 PH/15 Installed PCRE 6.2
3974 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3976 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3977 of the problem. Specifically, suggested +O2 rather than +O1 for the
3980 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3982 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3983 record, the port was not correctly logged when the outgoing_port log
3984 selector was set (it logged the transort's default port).
3986 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3987 fallback_hosts, and "hosts" in the smtp transport.
3989 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3990 host errors such as "Connection refused".
3992 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3993 authentication with radiusclient 0.4.9:
3995 - Error returned from rc_read_config was caught wrongly
3996 - Username/password not passed on to radius server due to wrong length.
3998 The presumption is that some radiusclient API changes for 4.51/PH/17
3999 were not taken care of correctly. The code is still untested by me (my
4000 Linux distribution still has 0.3.2 of radiusclient), but it was
4001 contributed by a Radius user.
4003 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4004 expanding the "port" option of the smtp transport.
4006 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4007 while reading a MIME header. Thanks to Tom Hughes for a patch.
4009 PH/24 Include config.h inside local_scan.h so that configuration settings are
4012 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4013 that in an ACL for RCPT (for example), you can examine exactly what was
4016 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4017 commands, but it was not correctly comparing the address with the actual
4018 client host address. Thus, it would show the EHLO address in Received:
4019 header lines when this was not necessary.
4021 PH/27 Added the % operator to ${eval:}.
4023 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4024 it should be ignoring failures (because with -C, for example, it has lost
4025 privilege). It wasn't ignoring creation failures other than "already
4028 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4031 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4032 because an input line was too long, either on its own, or by virtue of
4033 too many continuations, the temporary file was not being removed, and the
4034 return code was incorrect.
4036 PH/31 Missing "BOOL" in function definition in filtertest.c.
4038 PH/32 Applied Sieve patches from the maintainer.
4040 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4042 PH/33 Added "verify = not_blind".
4044 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4045 Local/Makefile (with some defaults set). These are used in built scripts
4046 such as exicyclog, but they have never been used in the exim_install
4047 script (though there are many overriding facilities there). I have
4048 arranged that the exim_install script now takes note of these two
4051 PH/35 Installed configuration files for Dragonfly.
4053 PH/36 When a locally submitted message by a trusted user did not contain a
4054 From: header, and the sender address was obtained from -f or from an SMTP
4055 MAIL command, and the trusted user did not use -F to supply a sender
4056 name, $originator_name was incorrectly used when constructing a From:
4057 header. Furthermore, $originator_name was used for submission mode
4058 messages from external hosts without From: headers in a similar way,
4059 which is clearly wrong.
4061 PH/37 Added control=suppress_local_fixups.
4063 PH/38 When log_selector = +received_sender was set, and the addition of the
4064 sender made the log line's construction buffer exactly full, or one byte
4065 less than full, an overflow happened when the terminating "\n" was
4068 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4069 when the result of a list match is failure because a DNS lookup failed.
4071 PH/40 RM_COMMAND is now used in the building process.
4073 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4074 the "build-* directories that it finds.
4076 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4077 address in a domain literal was a prefix of an interface address.
4079 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4080 when verifying a sender address, unless rewrite_headers is false.
4082 PH/44 Wrote a long comment about why errors_to addresses are verified as
4083 recipients, not senders.
4085 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4086 the ratelimit ACL was added.
4088 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4090 PH/46 Added extra information about PostgreSQL errors to the error string.
4092 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4093 avoiding the need to install special zones in a real server. This is
4094 backwards compatible; if it can't find the fake resolver, it drops back.
4095 Thus, both old and new test suites can be run.
4097 TF/02 Added util/ratelimit.pl
4099 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4100 clock is set back in time.
4102 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4103 Candler <B.Candler@pobox.com>.
4105 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4106 OK for addresses that are the result of redirections.
4108 PH/48 A number of further additions for the benefit of the new test suite,
4109 including a fake gethostbyname() that interfaces to the fake DNS resolver
4112 TF/06 The fix for widen_domains has also been applied to qualify_single and
4113 search_parents which are the other dnslookup options that can cause
4116 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4119 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4121 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4122 service" if it encountered an MX record with an empty target hostname.
4123 The message is now "an MX or SRV record indicated no SMTP service".
4125 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4126 if the DNS of the sending site is misconfigured. This is quite a
4127 common situation. This change restores the behaviour of treating a
4128 helo verification defer as a failure.
4130 PH/51 If self=fail was set on a router, the bounce message did not include the
4131 actual error message.
4137 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4139 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4140 command itself ended in error, the underlying message about the transport
4141 filter timeout was being overwritten with the pipe command error. Now the
4142 underlying error message should be appended to the second error message.
4144 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4146 PH/02 Reduce the amount of output that "make" produces by default. Full output
4147 can still be requested.
4149 PH/03 The warning log line about a condition test deferring for a "warn" verb
4150 was being output only once per connection, rather than after each
4151 occurrence (because it was using the same function as for successful
4152 "warn" verbs). This seems wrong, so I have changed it.
4154 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4155 it should not have, which might have caused a crash in the right
4156 circumstances, but probably never did.
4158 PH/04 Installed a modified version of Tony Finch's patch to make submission
4159 mode fix the return path as well as the Sender: header line, and to
4160 add a /name= option so that you can make the user's friendly name appear
4163 TF/03 Added the control = fakedefer ACL modifier.
4165 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4166 Mark Lowes for thorough testing.
4168 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4170 TK/03 Merged latest SRS patch from Miles Wilton.
4172 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4173 with the definition in sysexits.h (which is #included earlier).
4174 Fortunately, Exim does not actually use EX_OK. The code used to try to
4175 preserve the sysexits.h value, by assuming that macro definitions were
4176 scanned for macro replacements. I have been disabused of this notion,
4177 so now the code just undefines EX_OK before #including unistd.h.
4179 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4180 in the smtp transport. When a block could not be written in a single
4181 write() function, the timeout was being re-applied to each part-write.
4182 This seems wrong - if the receiver was accepting one byte at a time it
4183 would take for ever. The timeout is now adjusted when this happens. It
4184 doesn't have to be particularly precise.
4186 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4187 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4189 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4190 without a domain if the check to <postmaster@domain> fails.
4192 SC/01 Eximstats: added -xls and the ability to specify output files
4193 (patch written by Frank Heydlauf).
4195 SC/02 Eximstats: use FileHandles for outputting results.
4197 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4199 SC/04 Eximstats: fixed display of large numbers with -nvr option
4201 SC/05 Eximstats: fixed merging of reports with empty tables.
4203 SC/06 Eximstats: added the -include_original_destination flag
4205 SC/07 Eximstats: removed tabs and trailing whitespace.
4207 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4209 TK/06 MBOX spool code: Add real "From " MBOX separator line
4210 so the .eml file is really in mbox format (even though
4211 most programs do not really care). Patch from Alex Miller.
4213 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4214 The latter is generated from $received_to and is only set if the
4215 message has one envelope recipient. SA can use these headers,
4216 obviously out-of-the-box. Patch from Alex Miller.
4218 PH/08 The ${def test on a variable was returning false if the variable's
4219 value was "0", contrary to what the specification has always said!
4220 The result should be true unless the variable is empty.
4222 PH/09 The syntax error of a character other than { following "${if
4223 def:variable_name" (after optional whitespace) was not being diagnosed.
4224 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4225 accidental colon was present, for example, could give incorrect results.
4227 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4228 result is used (not including appendfile, where other changes are about
4231 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4232 This involved changing a lot of size variables from int to off_t. It
4233 should work with maildirs and everything.
4235 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4236 spamd dying while we are connected to it.
4238 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4241 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4242 function for BDB 4.3.
4244 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4246 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4247 the sources for PCRE have been re-arranged and more files are now
4250 PH/15 The code I had for printing potentially long long variables in PH/11
4251 above was not the best (it lost precision). The length of off_t variables
4252 is now inspected at build time, and an appropriate printing format (%ld
4253 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4254 to be "long long int" or "long int". This is needed for the internal
4255 formatting function string_vformat().
4257 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4258 the configuration file to be ":syslog", then the script "guesses" where
4259 the logs files are, rather than using the compiled in default. In our
4260 case the guess is not the same as the compiled default, so the script
4261 suddenly stopped working when I started to use syslog. The patch checks
4262 to see if log_file_path is "". If so, it attempts to read it from exim
4263 with no configuration file to get the compiled in version, before it
4264 falls back to the previous guessing code."
4266 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4267 implementing BATV in an Exim configuration. See NewStuff for the gory
4270 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4271 Makefile that are specific to HP-UX.
4273 PH/18 If the "use_postmaster" option was set for a recipient callout together
4274 with the "random" option, the postmaster address was used as the MAIL
4275 FROM address for the random test, but not for the subsequent recipient
4276 test. It is now used for both.
4278 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4279 patch removes a few documentation additions to RFC 3028, because the
4280 latest draft now contains them. It adds the new en;ascii-case comparator
4281 and a new error check for 8bit text in MIME parts. Comparator and
4282 require names are now matched exactly. I enabled the subaddress
4283 extension, but it is not well tested yet (read: it works for me)."
4285 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4286 rework some of the code of TK/09 above to avoid the hardwired use of
4287 "%lld" and "long long". Replaced the call to snprintf() with a call to
4290 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4291 records point to non-existent hosts", "retry timeout exceeded", and
4292 "retry time not reached for any host after a long failure period".
4294 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4295 experimental DomainKeys support:
4297 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4298 (2) On an error such as an illegally used "control", the wrong name for
4299 the control was given.
4301 These problems did NOT occur unless DomainKeys support was compiled.
4303 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4305 PH/24 Added ${if match_ip condition.
4307 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4308 databases so that it will be absolutely obvious if a crash occurs in the
4309 DB library. This is a regular occurrence (often caused by mis-matched
4312 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4313 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4314 picked up on a user's system that detects such things. There doesn't seem
4315 to be a gcc warning option for this - only an attribute that has to be
4316 put on the function's prototype. It seems that in Fedora Core 4 they have
4317 set this on a number of new functions. No doubt there will be more in due
4320 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4321 specify a transport. However, if an address that was verified by such a
4322 router was the subject of a callout, Exim crashed because it tried to
4323 read the rcpt_include_affixes from the non-existent transport. Now it
4324 just assumes that the setting of that option is false. This bug was
4325 introduced by 4.51/PH/31.
4327 PH/28 Changed -d+all to exclude +memory, because that information is very
4328 rarely of interest, but it makes the output a lot bigger. People tend to
4329 do -d+all out of habit.
4331 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4332 code in os-type was giving problems when libc.so lives in lib64, like on
4335 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4336 aren't the modern standard, and it seems that some systems' include files
4337 don't always have them. Exim was already checking for some of the newer
4338 ones like T_AAAA, and defining it itself. I've added checks for all the
4339 record types that Exim uses.
4341 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4342 not automatically generating a new one, as it is supposed to. This
4343 prevented TLS from working. If the file did exist, but contained invalid
4344 data, a new version was generated, as expected. It was only the case of a
4345 non-existent file that was broken.
4347 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4348 with a change in libdomainkeys > 0.64.
4350 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4351 from DNS. If the selector record carries the flag, it now has
4352 precedence over the domain-wide flag.
4354 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4356 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4357 the use of a port name that isn't defined in /etc/services) occurred, the
4358 message was deferred as in a normal delivery, and thus remained on the
4359 spool, instead of being failed because of the mua_wrapper setting. This
4360 is now fixed, and I tidied up some of the mua_wrapper messages at the
4363 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4364 the messages in a hash of arrays rather than using individual hashes.
4365 This is a bit cleaner and results in dramatic memory savings, albeit
4366 at a slight CPU cost.
4368 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4369 as requested by Marc Sherman.
4371 SC/10 Eximstats: added histograms for user specified patterns as requested
4374 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4376 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4377 fopen() in the content-scanning modules that did not already have it.
4383 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4384 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4386 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4388 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4390 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4391 to test Sieve filters that use "vacation".
4393 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4394 that changes the way the GnuTLS parameters are stored in the cache file.
4395 The new format can be generated externally. For backward compatibility,
4396 if the data in the cache doesn't make sense, Exim assumes it has read an
4397 old-format file, and it generates new data and writes a new file. This
4398 means that you can't go back to an older release without removing the
4401 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4402 work if there are any delivery delays because "one_time" forces the
4403 parent to be marked "delivered", so its unseen clone is never tried
4404 again. For this reason, Exim now forbids the simultaneous setting of
4407 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4408 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4409 entry if you want to know the details. The fix, however, neglected to
4410 consider the case when local delivery batching is involved. The test for
4411 "previously delivered" was not happening when checking to see if an
4412 address could be batched with a previous (undelivered) one; under
4413 certain circumstances this could lead to multiple deliveries to the same
4416 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4417 in its include files, and this causes problems building Exim.
4419 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4420 header_syntax) but Exim was just ignoring anything given after a slash.
4421 In particular, this caused confusion with an attempt to use "verify =
4422 reverse_host_lookup/defer_ok". An error is now given when options are
4423 supplied for verify items that do not have them. (Maybe reverse_host_
4424 lookup should have a defer_ok option, but that's a different point.)
4426 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4427 defined by RFC 821) to 2048, because there were problems with some AUTH
4428 commands, and RFC 1869 says the size should be increased for extended
4429 SMTP commands that take arguments.
4431 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4434 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4435 "unknown" error; now it says that the functionality isn't in the binary.
4437 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4438 an address' error message when a string expansion fails (syntax or
4439 whatever). Otherwise the password may appear in the log. Following change
4440 PH/42 below, there is no longer a chance of it appearing in a bounce
4443 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4445 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4446 of its addresses was ever tried. (Bugzilla bug #2.)
4448 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4449 the result incorrectly in the debug output. (It correctly added a newline
4450 to what was transported.)
4452 TF/01 Added $received_time.
4454 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4455 commented out examples of how to interface to a virus scanner and to
4456 SpamAssassin. Also added commented examples of av_scanner and
4457 spamd_address settings.
4459 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4460 and controls are allowed in which ACLs. There were a couple of minor
4461 errors. Some of the entries in the conditions table (which is a table of
4462 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4463 negation of where the condition IS allowed.
4465 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4467 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4468 header file does not have a version number, so I've had to invent a new
4469 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4470 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4471 radiusclient), but it was contributed by a Radius user.
4473 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4474 files or directories, for OpenSSL.
4476 PH/19 When an Exim process that is running as root has to create an Exim log
4477 file, it does so in a subprocess that runs as exim:exim so as to get the
4478 ownership right at creation (otherwise, other Exim processes might see
4479 the file with the wrong ownership). There was no test for failure of this
4480 fork() call, which would lead to the process getting stuck as it waited
4481 for a non-existent subprocess. Forks do occasionally fail when resources
4482 run out. I reviewed all the other calls to fork(); they all seem to check
4485 PH/20 When checking for unexpected SMTP input at connect time (before writing
4486 the banner), Exim was not dealing correctly with a non-positive return
4487 from the read() function. If the client had disconnected by this time,
4488 the result was a log entry for a synchronization error with an empty
4489 string after "input=" when read() returned zero. If read() returned -1
4490 (an event I could not check), uninitialized data bytes were printed.
4491 There were reports of junk text (parts of files, etc) appearing after
4494 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4496 PH/22 Added support for macro redefinition, and (re)definition in between
4497 driver and ACL definitions.
4499 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4500 forgetting to use the resulting value; it was using the unexpanded value.
4502 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4503 hadn't been configured. The fix is from Juergen Kreileder, who
4504 understands it better than I do:
4506 "Here's what I see happening with three configured cyrus_sasl
4507 authenticators configured (plain, login, cram-md5):
4509 On startup auth_cyrus_sasl_init() gets called for each of these.
4510 This means three calls to sasl_listmech() without a specified mech_list.
4511 => SASL tests which mechs of all available mechs actually work
4512 => three warnings about OTP not working
4513 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4515 With the patch, sasl_listmech() also gets called three times. But now
4516 SASL's mech_list option is set to the server_mech specified in the the
4517 authenticator. Or in other words, the answer from sasl_listmech()
4518 gets limited to just the mech you're testing for (which is different
4520 => the return list contains just 'plain' or 'login', 'cram-md5' or
4521 nothing depending on the value of ob->server_mech.
4523 I've just tested the patch: Authentication still works fine,
4524 unavailable mechs specified in the exim configuration are still
4525 caught, and the auth.log warnings about OTP are gone."
4527 PH/25 When debugging is enabled, the contents of the command line are added
4528 to the debugging output, even when log_selector=+arguments is not
4531 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4532 answer is "GNU", and only if the return is "GNU/something" is the answer
4535 PH/27 $acl_verify_message is now set immediately after the failure of a
4536 verification in an ACL, and so is available in subsequent modifiers. In
4537 particular, the message can be preserved by coding like this:
4539 warn !verify = sender
4540 set acl_m0 = $acl_verify_message
4542 Previously, $acl_verify_message was set only while expanding "message"
4543 and "log_message" when a very denied access.
4545 PH/28 Modified OS/os.c-Linux with
4547 -#ifndef OS_LOAD_AVERAGE
4548 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4550 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4551 nomenclature these days.)
4553 PH/29 Installed patch from the Sieve maintainer that adds the options
4554 sieve_useraddress and sieve_subaddress to the redirect router.
4556 PH/30 In these circumstances:
4557 . Two addresses routed to the same list of hosts;
4558 . First host does not offer TLS;
4559 . First host accepts first address;
4560 . First host gives temporary error to second address;
4561 . Second host offers TLS and a TLS session is established;
4562 . Second host accepts second address.
4563 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4564 and peerdn, if requested) that were in fact used only for the second
4567 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4568 attention to any local part prefix or suffix that was matched by the
4569 router that accepted the address. It now behaves in the same way as it
4570 does for delivery: the affixes are removed from the local part unless
4571 rcpt_include_affixes is set on the transport.
4573 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4574 timeout during the DATA phase of an incoming message.
4576 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4577 applied a patch sanctioned by the Sieve maintainer.
4579 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4580 the uid or gid is negative. A case of a negative gid caused this to be
4581 noticed. The fix allows for either to be negative.
4583 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4584 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4587 PH/36 The change PH/12 above was broken. Fixed it.
4589 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4590 the grounds that routing the same address twice would always produce the
4591 same answer. This might have been true once, but it is certainly no
4592 longer true now. Routing a child address may depend on the previous
4593 routing that produced that child. Some complicated redirection strategies
4594 went wrong when messages had multiple recipients, and made Exim's
4595 behaviour dependent on the order in which the addresses were given.
4597 I have moved the duplicate checking until after the routing is complete.
4598 Exim scans the addresses that are assigned to local and remote
4599 transports, and removes any duplicates. This means that more work will be
4600 done, as duplicates will always all be routed, but duplicates are
4601 presumably rare, so I don't expect this is of any significance.
4603 For deliveries to pipes, files, and autoreplies, the duplicate checking
4604 still happens during the routing process, since they are not going to be
4607 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4608 It corrects a timeout issue with spamd. This is Ian's comment: "The
4609 background is that sometimes spamd either never reads data from a
4610 connection it has accepted, or it never writes response data. The exiscan
4611 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4612 blindly assumes that writes won't block so it may never time out."
4614 PH/39 Allow G after quota size as well as K and M.
4616 PH/40 The value set for $authenticated_id in an authenticator may not contain
4617 binary zeroes or newlines because the value is written to log lines and
4618 to spool files. There was no check on this. Now the value is run through
4619 the string_printing() function so that such characters are converted to
4620 printable escape sequences.
4622 PH/41 $message_linecount is a new variable that contains the total number of
4623 lines in the message. Compare $body_linecount, which is the count for the
4626 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4627 bounce and delay warning messages, except in certain special cases, which
4630 (a) An SMTP error message from a remote host;
4631 (b) A message specified in a :fail: redirection;
4632 (c) A message specified in a "fail" command in a system filter;
4633 (d) A message specified in a FAIL return from the queryprogram router;
4634 (e) A message specified by the cannot_route_message router option.
4636 In these cases only, Exim does include the error details in bounce and
4637 warning messages. There are also a few cases where bland messages such
4638 as "unrouteable address" or "local delivery error" are given.
4640 PH/43 $value is now also set for the "else" part of a ${run expansion.
4642 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4643 being worked on, but at least Exim now implements the latest version to
4646 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4647 process to complete was treated as a delivery failure, a timeout while
4648 writing the message to the pipe was logged, but erroneously treated as a
4649 successful delivery. Such timeouts include transport filter timeouts. For
4650 consistency with the overall process timeout, these timeouts are now
4651 treated as errors, giving rise to delivery failures by default. However,
4652 there is now a new Boolean option for the pipe transport called
4653 timeout_defer, which, if set TRUE, converts the failures into defers for
4654 both kinds of timeout. A transport filter timeout is now identified in
4657 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4658 systems where "make" and "gmake" are different, calling "gmake" at top
4659 level broke things. I've arranged for the value of $(MAKE) to be passed
4660 from the Makefile to this script so that it can call the same version of
4664 A note about Exim versions 4.44 and 4.50
4665 ----------------------------------------
4667 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4668 changes of various kinds. As a consequence, a big documentation update was
4669 needed. This delayed the release for rather longer than seemed good, especially
4670 in the light of a couple of (minor) security issues. Therefore, the changes
4671 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4672 release. So 4.44 and 4.50 are in effect two different branches that both start
4675 I have left the 4.50 change log unchanged; it contains all the changes since
4676 4.43. The change log for 4.44 is below; many of its items are identical to
4677 those for 4.50. This seems to be the most sensible way to preserve the
4678 historical information.
4684 1. Minor wording change to the doc/README.SIEVE file.
4686 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4687 computation of the current number of files was incorrect.
4689 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4690 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4693 4. Give more explanation in the error message when the command for a transport
4694 filter fails to execute.
4696 5. There are several places where Exim runs a non-Exim command in a
4697 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4698 was being done only for the command run by the queryprogram router. It is
4699 now done for all such subprocesses. The other cases are: ${run, transport
4700 filters, and the commands run by the lmtp and pipe transports.
4702 6. Added CONFIGURE_GROUP build-time option.
4704 7. Some older OS have a limit of 256 on the maximum number of file
4705 descriptors. Exim was using setrlimit() to set 1000 as a large value
4706 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4707 systems. I've change it so that if it can't get 1000, it tries for 256.
4709 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4710 was an oversight, and furthermore, ever since the addition of extra
4711 controls (e.g. 4.43/32), the checks on when to allow different forms of
4712 "control" were broken. There should now be diagnostics for all cases when a
4713 control that does not make sense is encountered.
4715 9. Added the /retain_sender option to "control=submission".
4717 10. $recipients is now available in the predata ACL (oversight).
4719 11. Tidy the search cache before the fork to do a delivery from a message
4720 received from the command line. Otherwise the child will trigger a lookup
4721 failure and thereby defer the delivery if it tries to use (for example) a
4722 cached ldap connection that the parent has called unbind on.
4724 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4725 of $address_data from the recipient verification was clobbered by the
4726 sender verification.
4728 13. The value of address_data from a sender verification is now available in
4729 $sender_address_data in subsequent conditions in the ACL statement.
4731 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4733 15. Added a new option "connect=<time>" to callout options, to set a different
4736 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4737 was its contents. (It was OK if the option was not defined at all.)
4739 17. A "Completed" log line is now written for messages that are removed from
4740 the spool by the -Mrm option.
4742 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4743 information about exactly what failed.
4745 19. Added -dd to debug only the daemon process.
4747 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4748 handles timeouts, both on the server side and network timeouts. Renamed the
4749 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4751 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4752 It is now set to "smtps".
4754 22. $host_address is now set to the target address during the checking of
4755 ignore_target_hosts.
4757 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4758 being passed; this would have caused $sender_host_name to have been used if
4759 matching the list had actually called for a host name (not very likely,
4760 since this list is usually IP addresses). A host name is now passed as
4763 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4764 code that specifies a non-restarting handler (typically sigaction() in
4765 modern systems) in an attempt to fix a rare and obscure crash bug.
4767 25. Narrowed the window for a race in the daemon that could cause it to ignore
4768 SIGCHLD signals. This is not a major problem, because they are used only to
4769 wake it up if nothing else does.
4771 26. A malformed maildirsize file could cause Exim to calculate negative values
4772 for the mailbox size or file count. Odd effects could occur as a result.
4773 The maildirsize information is now recalculated if the size or filecount
4776 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4777 support for a long time. Removed HAVE_SYS_VFS_H.
4779 28. Installed the latest version of exipick from John Jetmore.
4781 29. In an address list, if the pattern was not a regular expression, an empty
4782 subject address (from a bounce message) matched only if the pattern was an
4783 empty string. Non-empty patterns were not even tested. This was the wrong
4784 because it is perfectly reasonable to use an empty address as part of a
4785 database query. An empty address is now tested by patterns that are
4786 lookups. However, all the other forms of pattern expect the subject to
4787 contain a local part and a domain, and therefore, for them, an empty
4788 address still always fails if the pattern is not itself empty.
4790 30. Exim went into a mad DNS loop when attempting to do a callout where the
4791 host was specified on an smtp transport, and looking it up yielded more
4792 than one IP address.
4794 31. Re-factored the code for checking spool and log partition space into a
4795 function that finds that data and another that does the check. The former
4796 is then used to implement four new variables: $spool_space, $log_space,
4797 $spool_inodes, and $log_inodes.
4799 32. The RFC2047 encoding function was originally intended for short strings
4800 such as real names; it was not keeping to the 75-character limit for
4801 encoded words that the RFC imposes. It now respects the limit, and
4802 generates multiple encoded words if necessary. To be on the safe side, I
4803 have increased the buffer size for the ${rfc2047: expansion operator from
4806 33. It is now permitted to omit both strings after an "if" condition; if the
4807 condition is true, the result is "true". As before, when the second string
4808 is omitted, a false condition yields an empty string. This makes it less
4809 cumbersome to write custom ACL and router conditions.
4811 34. Failure to deliver a bounce message always caused it to be frozen, even if
4812 there was an errors_to setting on the router. The errors_to setting is now
4815 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4816 canonical form (fully expanded) before being placed in
4817 $sender_host_address.
4819 36. The table in the code that translates DNS record types into text (T_A to
4820 "A" for instance) was missing entries for NS and CNAME. It is just possible
4821 that this could have caused confusion if both these types were looked up
4822 for the same domain, because the text type is used as part of Exim's
4823 per-process caching. But the chance of anyone hitting this buglet seems
4826 37. The dnsdb lookup has been extended in a number of ways.
4828 (1) There is a new type, "zns", which walks up the domain tree until it
4829 finds some nameserver records. It should be used with care.
4831 (2) There is a new type, "mxh", which is like "mx" except that it returns
4832 just the host names, not the priorities.
4834 (3) It is now possible to give a list of domains (or IP addresses) to be
4835 looked up. The behaviour when one of the lookups defers can be
4836 controlled by a keyword.
4838 (4) It is now possible to specify the separator character for use when
4839 multiple records are returned.
4841 38. The dnslists ACL condition has been extended: it is now possible to supply
4842 a list of IP addresses and/or domains to be looked up in a particular DNS
4845 39. Added log_selector=+queue_time_overall.
4847 40. When running the queue in the test harness, wait just a tad after forking a
4848 delivery process, to get repeatability of debugging output.
4850 41. Include certificate and key file names in error message when GnuTLS fails
4851 to set them up, because the GnuTLS error message doesn't include the name
4852 of the failing file when there is a problem reading it.
4854 42. Allow both -bf and -bF in the same test run.
4856 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4858 44. The "Exiscan patch" is now merged into the mainline Exim source.
4860 45. Sometimes the final signoff response after QUIT could fail to get
4861 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4862 < 0 before doing a fflush(). This bug looks as though it goes back to the
4863 introduction of TLS in release 3.20, but "sometimes" must have been rare
4864 because the tests only now provoked it.
4866 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4867 (this can affect the format of dates).
4869 47. exim_tidydb, when checking for the continued existence of a message for
4870 which it has found a message-specific retry record, was not finding
4871 messages that were in split spool directories. Consequently, it was
4872 deleting retry records that should have stayed in existence.
4874 48. Steve fixed some bugs in eximstats.
4876 49. The SPA authentication driver was not abandoning authentication and moving
4877 on to the next authenticator when an expansion was forced to fail,
4878 contradicting the general specification for all authenticators. Instead it
4879 was generating a temporary error. It now behaves as specified.
4881 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4882 (the order specifies the preference for clients). The order is now AES256,
4883 AES128, 3DES, ARCFOUR128.
4885 51. Small patch to Sieve code - explicitly set From: when generating an
4888 52. Exim crashed if a remote delivery caused a very long error message to be
4889 recorded - for instance if somebody sent an entire SpamAssassin report back
4890 as a large number of 550 error lines. This bug was coincidentally fixed by
4891 increasing the size of one of Exim's internal buffers (big_buffer) that
4892 happened as part of the Exiscan merge. However, to be on the safe side, I
4893 have made the code more robust (and fixed the comments that describe what
4896 53. Now that there can be additional text after "Completed" in log lines (if
4897 the queue_time_overall log selector is set), a one-byte patch to exigrep
4898 was needed to allow it to recognize "Completed" as not the last thing in
4901 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4902 patch that reportedly fixes this has been added. I am not expert enough to
4903 create a test for it. This is what the patch creator wrote:
4905 "I found a little strange behaviour of ldap code when working with
4906 Windows 2003 AD Domain, where users was placed in more than one
4907 Organization Units. When I tried to give exim partial DN, the exit code
4908 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4909 But simultaneously result of request was absolutely normal ldap result,
4910 so I produce this patch..."
4912 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4913 REFERENCE, so I have modified the code to exclude the patch when that macro
4916 55. Some experimental protocols are using DNS PTR records for new purposes. The
4917 keys for these records are domain names, not reversed IP addresses. The
4918 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4919 leaves it alone. Component reversal etc. now happens only for IP addresses.
4922 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4924 57. Double the size of the debug message buffer (to 2048) so that more of very
4925 long debug lines gets shown.
4927 58. The exicyclog utility now does better if the number of log files to keep
4928 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4930 59. Two changes related to the smtp_active_hostname option:
4932 (1) $smtp_active_hostname is now available as a variable.
4933 (2) The default for smtp_banner uses $smtp_active_hostname instead
4934 of $primary_hostname.
4936 60. The host_aton() function is supposed to be passed a string that is known
4937 to be a valid IP address. However, in the case of IPv6 addresses, it was
4938 not checking this. This is a hostage to fortune. Exim now panics and dies
4939 if the condition is not met. A case was found where this could be provoked
4940 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4941 components; fortuitously, this particular loophole had already been fixed
4942 by change 4.50/55 above.
4944 If there are any other similar loopholes, the new check in host_aton()
4945 itself should stop them being exploited. The report I received stated that
4946 data on the command line could provoke the exploit when Exim was running as
4947 exim, but did not say which command line option was involved. All I could
4948 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4949 running as the user.
4952 61. There was a buffer overflow vulnerability in the SPA authentication code
4953 (which came originally from the Samba project). I have added a test to the
4954 spa_base64_to_bits() function which I hope fixes it.
4957 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4958 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4960 63. The daemon start-up calls getloadavg() while still root for those OS that
4961 need the first call to be done as root, but it missed one case: when
4962 deliver_queue_load_max is set with deliver_drop_privilege. This is
4963 necessary for the benefit of the queue runner, because there is no re-exec
4964 when deliver_drop_privilege is set.
4966 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4967 This has been fixed.
4969 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4970 was in use, was not putting the data itself into the right store pool;
4971 consequently, it could be overwritten for a subsequent message in the same
4972 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4975 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4977 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4978 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4979 the function were treating the return as a boolean value, which happened to
4980 work because 0=false and not-0=true, but is not correct code.
4982 68. The host_aton() function was not handling scoped IPv6 addresses (those
4983 with, for example, "%eth0" on the end) correctly.
4985 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4986 negated items (that is, ~something) in unsigned ints. Some compilers
4987 apparently mutter when there is no cast.
4989 70. If an address verification called from an ACL failed, and did not produce a
4990 user-specific message (i.e. there was only a "system" message), nothing was
4991 put in $acl_verify_message. In this situation, it now puts the system
4994 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4995 session; change 4.31/43 added the unwanted input to the log line - except
4996 that it did not do this in the start of session case. It now does.
4998 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4999 This is wrong and can cause the other end to generate a synchronization
5000 error if it is another Exim or anything else that does the synchronization
5001 check. A QUIT command is no longer sent after a timeout.
5003 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5004 during host lookups.
5006 74. The defer_ok option of callout verification was not working if it was used
5007 when verifying addresses in header lines, that is, for this case:
5009 verify = header_sender/callout=defer_ok
5011 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5012 those file descriptors could be used for SMTP connections. If anything
5013 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5014 could be sent to the SMTP client, causing chaos. The daemon now opens
5015 stdin, stdout, and stderr to /dev/null when it puts itself into the
5018 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5019 log by default. The user can override this with suitable Perl magic.
5021 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5022 the log message when discard triggers, was not working for the DATA ACL or
5023 for the non-SMTP ACL.
5025 78. Error message wording change in sieve.c.
5027 79. If smtp_accept_max_per_host was set, the number of connections could be
5028 restricted to fewer than expected, because the daemon was trying to set up
5029 a new connection before checking whether the processes handling previous
5030 connections had finished. The check for completed processes is now done
5031 earlier. On busy systems, this bug wouldn't be noticed because something
5032 else would have woken the daemon, and it would have reaped the completed
5035 80. If a message was submitted locally by a user whose login name contained one
5036 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5037 It caused a spool format error. I have fixed the spool reading code. A
5038 related problem was that the "from" clause in the Received: line became
5039 illegal because of the space(s). It is now covered by ${quote_local_part.
5041 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5044 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5046 83. Patch from Sieve maintainer to fix unterminated string problem in
5047 "vacation" handling.
5049 84. Some minor changes to the Linux configuration files to help with other
5050 OS variants using glibc.
5052 85. One more patch for Sieve to update vacation handling to latest spec.
5055 ----------------------------------------------------
5056 See the note above about the 4.44 and 4.50 releases.
5057 ----------------------------------------------------
5063 1. Change 4.43/35 introduced a bug that caused file counts to be
5064 incorrectly computed when quota_filecount was set in an appendfile
5067 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5068 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5071 3. Give more explanation in the error message when the command for a transport
5072 filter fails to execute.
5074 4. There are several places where Exim runs a non-Exim command in a
5075 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5076 was being done only for the command run by the queryprogram router. It is
5077 now done for all such subprocesses. The other cases are: ${run, transport
5078 filters, and the commands run by the lmtp and pipe transports.
5080 5. Some older OS have a limit of 256 on the maximum number of file
5081 descriptors. Exim was using setrlimit() to set 1000 as a large value
5082 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5083 systems. I've change it so that if it can't get 1000, it tries for 256.
5085 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5086 was an oversight, and furthermore, ever since the addition of extra
5087 controls (e.g. 4.43/32), the checks on when to allow different forms of
5088 "control" were broken. There should now be diagnostics for all cases when a
5089 control that does not make sense is encountered.
5091 7. $recipients is now available in the predata ACL (oversight).
5093 8. Tidy the search cache before the fork to do a delivery from a message
5094 received from the command line. Otherwise the child will trigger a lookup
5095 failure and thereby defer the delivery if it tries to use (for example) a
5096 cached ldap connection that the parent has called unbind on.
5098 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5099 of $address_data from the recipient verification was clobbered by the
5100 sender verification.
5102 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5103 was its contents. (It was OK if the option was not defined at all.)
5105 11. A "Completed" log line is now written for messages that are removed from
5106 the spool by the -Mrm option.
5108 12. $host_address is now set to the target address during the checking of
5109 ignore_target_hosts.
5111 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5112 being passed; this would have caused $sender_host_name to have been used if
5113 matching the list had actually called for a host name (not very likely,
5114 since this list is usually IP addresses). A host name is now passed as
5117 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5118 code that specifies a non-restarting handler (typically sigaction() in
5119 modern systems) in an attempt to fix a rare and obscure crash bug.
5121 15. Narrowed the window for a race in the daemon that could cause it to ignore
5122 SIGCHLD signals. This is not a major problem, because they are used only to
5123 wake it up if nothing else does.
5125 16. A malformed maildirsize file could cause Exim to calculate negative values
5126 for the mailbox size or file count. Odd effects could occur as a result.
5127 The maildirsize information is now recalculated if the size or filecount
5130 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5131 support for a long time. Removed HAVE_SYS_VFS_H.
5133 18. Updated exipick to current release from John Jetmore.
5135 19. Allow an empty sender to be matched against a lookup in an address list.
5136 Previously the only cases considered were a regular expression, or an
5139 20. Exim went into a mad DNS lookup loop when doing a callout where the
5140 host was specified on the transport, if the DNS lookup yielded more than
5143 21. The RFC2047 encoding function was originally intended for short strings
5144 such as real names; it was not keeping to the 75-character limit for
5145 encoded words that the RFC imposes. It now respects the limit, and
5146 generates multiple encoded words if necessary. To be on the safe side, I
5147 have increased the buffer size for the ${rfc2047: expansion operator from
5150 22. Failure to deliver a bounce message always caused it to be frozen, even if
5151 there was an errors_to setting on the router. The errors_to setting is now
5154 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5155 canonical form (fully expanded) before being placed in
5156 $sender_host_address.
5158 24. Updated eximstats to version 1.33
5160 25. Include certificate and key file names in error message when GnuTLS fails
5161 to set them up, because the GnuTLS error message doesn't include the name
5162 of the failing file when there is a problem reading it.
5164 26. Expand error message when OpenSSL has problems setting up cert/key files.
5167 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5168 (this can affect the format of dates).
5170 28. exim_tidydb, when checking for the continued existence of a message for
5171 which it has found a message-specific retry record, was not finding
5172 messages that were in split spool directories. Consequently, it was
5173 deleting retry records that should have stayed in existence.
5175 29. eximstats updated to version 1.35
5176 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5177 1.35 - bugfix such that pie charts by volume are generated correctly
5179 30. The SPA authentication driver was not abandoning authentication and moving
5180 on to the next authenticator when an expansion was forced to fail,
5181 contradicting the general specification for all authenticators. Instead it
5182 was generating a temporary error. It now behaves as specified.
5184 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5185 (the order specifies the preference for clients). The order is now AES256,
5186 AES128, 3DES, ARCFOUR128.
5188 31. Small patch to Sieve code - explicitly set From: when generating an
5191 32. Exim crashed if a remote delivery caused a very long error message to be
5192 recorded - for instance if somebody sent an entire SpamAssassin report back
5193 as a large number of 550 error lines. This bug was coincidentally fixed by
5194 increasing the size of one of Exim's internal buffers (big_buffer) that
5195 happened as part of the Exiscan merge. However, to be on the safe side, I
5196 have made the code more robust (and fixed the comments that describe what
5199 33. Some experimental protocols are using DNS PTR records for new purposes. The
5200 keys for these records are domain names, not reversed IP addresses. The
5201 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5202 leaves it alone. Component reversal etc. now happens only for IP addresses.
5205 34. The host_aton() function is supposed to be passed a string that is known
5206 to be a valid IP address. However, in the case of IPv6 addresses, it was
5207 not checking this. This is a hostage to fortune. Exim now panics and dies
5208 if the condition is not met. A case was found where this could be provoked
5209 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5210 components; fortuitously, this particular loophole had already been fixed
5211 by change 4.50/55 or 4.44/33 above.
5213 If there are any other similar loopholes, the new check in host_aton()
5214 itself should stop them being exploited. The report I received stated that
5215 data on the command line could provoke the exploit when Exim was running as
5216 exim, but did not say which command line option was involved. All I could
5217 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5218 running as the user.
5221 35. There was a buffer overflow vulnerability in the SPA authentication code
5222 (which came originally from the Samba project). I have added a test to the
5223 spa_base64_to_bits() function which I hope fixes it.
5226 36. The daemon start-up calls getloadavg() while still root for those OS that
5227 need the first call to be done as root, but it missed one case: when
5228 deliver_queue_load_max is set with deliver_drop_privilege. This is
5229 necessary for the benefit of the queue runner, because there is no re-exec
5230 when deliver_drop_privilege is set.
5232 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5233 was in use, was not putting the data itself into the right store pool;
5234 consequently, it could be overwritten for a subsequent message in the same
5235 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5238 38. Sometimes the final signoff response after QUIT could fail to get
5239 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5240 < 0 before doing a fflush(). This bug looks as though it goes back to the
5241 introduction of TLS in release 3.20, but "sometimes" must have been rare
5242 because the tests only now provoked it.
5248 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5249 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5250 Now it returns an integer. A number of calls were still expecting a T/F
5251 return. Fortuitously, in all cases, the tests worked in OK situations,
5252 which is the norm. However, things would have gone wrong on any write
5253 failures on the smtp file descriptor. This function is used when sending
5254 messages over SMTP and also when doing verify callouts.
5256 2. When Exim is called to do synchronous delivery of a locally submitted
5257 message (the -odf or -odi options), it no longer closes stderr before doing
5260 3. Implemented the mua_wrapper option.
5262 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5264 5. Implemented the functions header_remove(), header_testname(),
5265 header_add_at_position(), and receive_remove_recipient(), and exported them
5268 6. If an ACL "warn" statement specified the addition of headers, Exim already
5269 inserted X-ACL-Warn: at the start if there was no header name. However, it
5270 was not making this test for the second and subsequent header lines if
5271 there were newlines in the string. This meant that an invalid header could
5272 be inserted if Exim was badly configured.
5274 7. Allow an ACL "warn" statement to add header lines at the start or after all
5275 the Received: headers, as well as at the end.
5277 8. Added the rcpt_4xx retry error code.
5279 9. Added postmaster_mailfrom=xxx to callout verification option.
5281 10. Added mailfrom=xxxx to the callout verification option, for verify=
5284 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5285 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5287 12. Inserted some casts to stop certain compilers warning when using pointer
5288 differences as field lengths or precisions in printf-type calls (mostly
5289 affecting debugging statements).
5291 13. Added optional readline() support for -be (dynamically loaded).
5293 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5294 same clock tick as a message's arrival, so that its received time was the
5295 same as the "first fail" time on the retry record, and that message
5296 remained on the queue past the ultimate address timeout, every queue runner
5297 would try a delivery (because it was past the ultimate address timeout) but
5298 after another failure, the ultimate address timeout, which should have then
5299 bounced the address, did not kick in. This was a "< instead of <=" error;
5300 in most cases the first failure would have been in the next clock tick
5301 after the received time, and all would be well.
5303 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5304 being recognized when the domain list was tested by the match_domain
5305 condition in an expansion string.
5307 16. Added the ${str2b64: operator.
5309 17. Exim was always calling setrlimit() to set a large limit for the number of
5310 processes, without checking whether the existing limit was already
5311 adequate. (It did check for the limit on file descriptors.) Furthermore,
5312 errors from getrlimit() and setrlimit() were being ignored. Now they are
5313 logged to the main and panic logs, but Exim does carry on, to try to do its
5314 job under whatever limits there are.
5316 18. Imported PCRE 5.0.
5318 19. Trivial typo in log message " temporarily refused connection" (the leading
5321 20. If the log selector return_path_on_delivery was set and an address was
5322 redirected to /dev/null, the delivery process crashed because it assumed
5323 that a return path would always be set for a "successful" delivery. In this
5324 case, the whole delivery is bypassed as an optimization, and therefore no
5327 21. Internal re-arrangement: the function for sending a challenge and reading
5328 a response while authentication was assuming a zero-terminated challenge
5329 string. It's now changed to take a pointer and a length, to allow for
5330 binary data in such strings.
5332 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5334 23. Exim was not respecting finduser_retries when seeking the login of the
5335 uid under which it was called; it was always trying 10 times. (The default
5336 setting of finduser_retries is zero.) Also, it was sleeping after the final
5337 failure, which is pointless.
5339 24. Implemented tls_on_connect_ports.
5341 25. Implemented acl_smtp_predata.
5343 26. If the domain in control=submission is set empty, Exim assumes that the
5344 authenticated id is a complete email address when it generates From: or
5345 Sender: header lines.
5347 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5348 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5349 chown and chgrp in /bin and hostname in /usr/bin.
5351 28. Exim was keeping the "process log" file open after each use, just as it
5352 does for the main log. This opens the possibility of it remaining open for
5353 long periods when the USR1 signal hits a daemon. Occasional processlog
5354 errors were reported, that could have been caused by this. Anyway, it seems
5355 much more sensible not to leave this file open at all, so that is what now
5358 29. The long-running daemon process does not normally write to the log once it
5359 has entered its main loop, and it closes the log before doing so. This is
5360 so that log files can straightforwardly be renamed and moved. However,
5361 there are a couple of unusual error situations where the daemon does write
5362 log entries, and I had neglected to close the log afterwards.
5364 30. The text of an SMTP error response that was received during a remote
5365 delivery was being truncated at 512 bytes. This is too short for some of
5366 the long messages that one sometimes sees. I've increased the limit to
5369 31. It is now possible to make retry rules that apply only when a message has a
5370 specific sender, in particular, an empty sender.
5372 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5373 it possible to be selective about when SMTP synchronization is enforced.
5375 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5377 32. Implemented hosts_connection_nolog.
5379 33. Added an ACL for QUIT.
5381 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5384 35. Added mailbox_size and mailbox_filecount to appendfile.
5386 36. Added control = no_multiline_responses to ACLs.
5388 37. There was a bug in the logic of the code that waits for the clock to tick
5389 in the case where the clock went backwards by a substantial amount such
5390 that the microsecond fraction of "now" was more than the microsecond
5391 fraction of "then" (but the whole seconds number was less).
5393 38. Added support for the libradius Radius client library this is found on
5394 FreeBSD (previously only the radiusclient library was supported).
5400 1. When certain lookups returned multiple values in the form name=value, the
5401 quoting of the values was not always being done properly. Specifically:
5402 (a) If the value started with a double quote, but contained no whitespace,
5404 (b) If the value contained whitespace other than a space character (i.e.
5405 tabs or newlines or carriage returns) it was not quoted.
5406 This fix has been applied to the mysql and pgsql lookups by writing a
5407 separate quoting function and calling it from the lookup code. The fix
5408 should probably also be applied to nisplus, ibase and oracle lookups, but
5409 since I cannot test any of those, I have not disturbed their existing code.
5411 2. A hit in the callout cache for a specific address caused a log line with no
5412 reason for rejecting RCPT. Now it says "Previous (cached) callout
5413 verification failure".
5415 3. There was an off-by-one bug in the queryprogram router. An over-long
5416 return line was truncated at 256 instead of 255 characters, thereby
5417 overflowing its buffer with the terminating zero. As well as fixing this, I
5418 have increased the buffer size to 1024 (and made a note to document this).
5420 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5421 when Exim is waiting for an SMTP response from a remote server, Exim
5422 restarts its select() call on the socket, thereby resetting its timeout.
5423 This is not a problem when such interrupts are rare. Somebody set up a cron
5424 job to run exiwhat every 2 minutes, which is less than the normal select()
5425 timeout (5 or 10 minutes). This meant that the select() timeout never
5426 kicked in because it was always reset. I have fixed this by comparing the
5427 time when an interrupt arrives with the time at the start of the first call
5428 to select(). If more time than the timeout has elapsed, the interrupt is
5429 treated as a timeout.
5431 5. Some internal re-factoring in preparation for the addition of Sieve
5432 extensions (by MH). In particular, the "personal" test is moved to a
5433 separate function, and given an option for scanning Cc: and Bcc: (which is
5434 not set for Exim filters).
5436 6. When Exim created an email address using the login of the caller as the
5437 local part (e.g. when creating a From: or Sender: header line), it was not
5438 quoting the local part when it contained special characters such as @.
5440 7. Installed new OpenBSD configuration files.
5442 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5443 try to make them clearer.
5445 9. Callout options, other than the timeout value, were being ignored when
5446 verifying sender addresses in header lines. For example, when using
5448 verify = header_sender/callout=no_cache
5450 the cache was (incorrectly) being used.
5452 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5453 only the Cygwin environment.
5455 11. When return_path_on_delivery was set as a log selector, if different remote
5456 addresses in the same message used different return paths and parallel
5457 remote delivery occurred, the wrong values would sometimes be logged.
5458 (Whenever a remote delivery process finished, the return path value from
5459 the most recently started remote delivery process was logged.)
5461 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5462 lines when AUTH and/or TLS are in use. This is the "received protocol"
5463 field. Exim used to use "asmtp" for authenticated SMTP, without any
5464 indication (in the protocol name) for TLS use. Now it follows the RFC and
5465 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5466 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5467 names appear in log lines as well as in Received: header lines.
5469 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5470 extensions, and comparison tests, and to fix some bugs.
5472 14. Changes to the "personal" filter test:
5474 (1) The test was buggy in that it was just doing the equivalent of
5475 "contains" tests on header lines. For example, if a user's address was
5476 anne@some.where, the "personal" test would incorrectly be true for
5478 To: susanne@some.where
5480 This test is now done by extracting each address from the header in turn,
5481 and checking the entire address. Other tests that are part of "personal"
5482 are now done using regular expressions (for example, to check local parts
5483 of addresses in From: header lines).
5485 (2) The list of non-personal local parts in From: addresses has been
5486 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5487 taken from the Sieve specification recommendations.
5489 (3) If the message contains any header line starting with "List-" it is
5490 treated as non-personal.
5492 (4) The test for "circular" in the Subject: header line has been removed
5493 because it now seems ill-conceived.
5495 15. Minor typos in src/EDITME comments corrected.
5497 16. Installed latest exipick from John Jetmore.
5499 17. If headers_add on a router specified a text string that was too long for
5500 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5501 of string_sprintf() is now avoided.
5503 18. $message_body_size was not set (it was always zero) when running the DATA
5504 ACL and the local_scan() function.
5506 19. For the "mail" command in an Exim filter, no default was being set for
5507 the once_repeat time, causing a random time value to be used if "once" was
5508 specified. (If the value happened to be <= 0, no repeat happened.) The
5509 default is now 0s, meaning "never repeat". The "vacation" command was OK
5510 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5511 (I found it when inspecting the code).
5513 20. There is now an overall timeout for performing a callout verification. It
5514 defaults to 4 times the callout timeout, which applies to individual SMTP
5515 commands during the callout. The overall timeout applies when there is more
5516 than one host that can be tried. The timeout is checked before trying the
5517 next host. This prevents very long delays if there are a large number of
5518 hosts and all are timing out (e.g. when the network connections are timing
5519 out). The value of the overall timeout can be changed by specifying an
5520 additional sub-option for "callout", called "maxwait". For example:
5522 verify = sender/callout=5s,maxwait=20s
5524 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5525 to the end before writing, but this should make it even safer).
5527 22. Exim was forgetting that it had advertised PIPELINING for the second and
5528 subsequent messages on an SMTP connection. It was also not resetting its
5529 memory on STARTTLS and an internal HELO.
5531 23. When Exim logs an SMTP synchronization error within a session, it now
5532 records whether PIPELINING has been advertised or not.
5534 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5535 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5536 rather than long int.
5538 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5540 26. Added the never_mail option to autoreply.
5546 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5547 crash if the getsockname() call failed; this can happen if a connection is
5548 closed very soon after it is established. The problem was simply in the
5549 order in which certain operations were done, causing Exim to try to write
5550 to the SMTP stream before it had set up the file descriptor. The bug has
5551 been fixed by making things happen in the correct order.
5557 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5558 before the connection was closed, thus losing the rejection response.
5560 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5561 some early Solaris releases, but causes trouble in current releases where
5562 socklen_t is defined.
5564 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5567 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5570 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5571 that releasing the top of it at the end releases what was used for sub-
5572 expansions (unless the block got too big). However, discard this block if
5573 the first thing is a variable or header, so that we can use its block when
5574 it is dynamic (useful for very large $message_headers, for example).
5576 6. Lookups now cache *every* query, not just the most recent. A new, separate
5577 store pool is used for this. It can be recovered when all lookup caches are
5578 flushed. Lookups now release memory at the end of their result strings.
5579 This has involved some general refactoring of the lookup sources.
5581 7. Some code has been added to the store_xxx() functions to reduce the amount
5582 of flapping under certain conditions.
5584 8. log_incoming_interface used to affect only the <= reception log lines. Now
5585 it causes the local interface and port to be added to several more SMTP log
5586 lines, for example "SMTP connection from", and rejection lines.
5588 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5590 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5592 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5593 could be overwritten at the end of the current message (or the start of a
5594 new message if it was set in a HELO ACL). The value is now preserved for
5595 the duration of the SMTP connection.
5597 12. If a transport had a headers_rewrite setting, and a matching header line
5598 contained an unqualified address, that address was qualified, even if it
5599 did not match any rewriting rules. The underlying bug was that the values
5600 of the flags that permit the existence of unqualified sender and recipient
5601 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5602 non-local messages, and by -bnq for local messages) were not being
5603 preserved with the message after it was received.
5605 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5606 "next input=" as part of the text comprising the host identity instead of
5607 the correct text. The code was using the same buffer for two different
5608 strings. However, depending on which order the printing function evaluated
5609 its arguments, the bug did not always show up. Under Linux, for example, my
5610 test suite worked just fine.
5612 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5613 doesn't work with some older versions of Perl. It has been changed to "my",
5614 which in any case is probably the better facility to use.
5616 15. A really picky compiler found some instances of statements for creating
5617 error messages that either had too many or two few arguments for the format
5620 16. The size of the buffer for calls to the DNS resolver has been increased
5621 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5622 for addresses that have a lot of PTR records. This alleviates a problem; it
5623 does not fully solve it.
5625 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5626 buffer now truncates the list and logs the incident, which is the same
5627 action as happens when Exim is looking up a host name and its aliases.
5628 Previously in this situation something unpredictable would happen;
5629 sometimes it was "internal error: store_reset failed".
5631 18. If a server dropped the connection unexpectedly when an Exim client was
5632 using GnuTLS and trying to read a response, the client delivery process
5633 crashed while trying to generate an error log message.
5635 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5636 string, for example:
5638 warn message = H1: something\nH2: something
5640 the text was added as a single header line from Exim's point of view
5641 though it ended up OK in the delivered message. However, searching for the
5642 second and subsequent header lines using $h_h2: did not work. This has been
5643 fixed. Similarly, if a system filter added multiple headers in this way,
5644 the routers could not see them.
5646 20. Expanded the error message when iplsearch is called with an invalid key to
5647 suggest using net-iplsearch in a host list.
5649 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5650 ACLs are no longer actually imposed (and a message to that effect is
5653 22. If a "gecos" field in a passwd entry contained escaped characters, in
5654 particular, if it contained a \" sequence, Exim got it wrong when building
5655 a From: or a Sender: header from that name. A second bug also caused
5656 incorrect handling when an unquoted " was present following a character
5657 that needed quoting.
5659 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5660 was not being matched caselessly.
5662 24. Arranged for all hyphens in the exim.8 source to be escaped with
5665 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5666 properly. Recipient callout cache records were still being keyed to include
5667 the sender, even when use_sender was set false. This led to far more
5668 callouts that were necessary. The sender is no longer included in the key
5669 when use_sender is false.
5671 26. Added "control = submission" modifier to ACLs.
5673 27. Added the ${base62d: operator to decode base 62 numbers.
5675 28. dnsdb lookups can now access SRV records.
5677 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5678 the configuration file.
5680 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5681 (-v) mode. This makes the output for a verbose queue run more intelligible.
5683 31. Added a use_postmaster feature to recipient callouts.
5685 32. Added the $body_zerocount variable, containing the number of binary zero
5686 bytes in the message body.
5688 33. The time of last modification of the "new" subdirectory is now used as the
5689 "mailbox time last read" when there is a quota error for a maildir
5692 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5694 35. Added +ignore_unknown as a special item in host lists.
5696 36. Code for decoding IPv6 addresses in host lists is now included, even if
5697 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5698 address was recognized as an IP address, but was then not correctly decoded
5699 into binary, causing unexpected and incorrect effects when compared with
5706 1. Very minor rewording of debugging text in manualroute to say "list of
5707 hosts" instead of "hostlist".
5709 2. If verify=header_syntax was set, and a header line with an unqualified
5710 address (no domain) and a large number of spaces between the end of the
5711 name and the colon was received, the reception process suffered a buffer
5712 overflow, and (when I tested it) crashed. This was caused by some obsolete
5713 code that should have been removed. The fix is to remove it!
5715 3. When running in the test harness, delay a bit after writing a bounce
5716 message to get a bit more predictability in the log output.
5718 4. Added a call to search_tidyup() just before forking a reception process. In
5719 theory, someone could use a lookup in the expansion of smtp_accept_max_
5720 per_host which, without the tidyup, could leave open a database connection.
5722 5. Added the variables $recipient_data and $sender_data which get set from a
5723 lookup success in an ACL "recipients" or "senders" condition, or a router
5724 "senders" option, similar to $domain_data and $local_part_data.
5726 6. Moved the writing of debug_print from before to after the "senders" test
5729 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5730 problems for message scanning, either using a data ACL, or using
5731 local_scan() because the Received: header was not generated till after they
5732 were called (in order to set the time as the time of reception completion).
5733 I have revised the way this works. The header is now generated after the
5734 body is received, but before the ACL or local_scan() are called. After they
5735 are run, the timestamp in the header is updated.
5741 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5742 before starting a queue runner without re-exec. This happened only when
5743 deliver_drop_privilege was set or when the Exim user was set to root. The
5744 effect of the bug was that timeouts during subsequent deliveries caused
5745 crashes instead of being properly handled. The handler is now left at its
5746 default (and expected) setting.
5748 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5749 message, again when deliver_drop_privilege is set or Exim is run as root.
5750 The bug described in (1) was not present in this case, but the tidying up
5751 of the other signals was missing. I have made the two cases consistent.
5753 3. The ignore_target_hosts setting on a manualroute router was being ignored
5754 for hosts that were looked up using the /MX notation.
5756 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5759 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5760 operated on the sender address. After changing the $sender_address to <>
5761 for the sender address verify, Exim was re-instated it as the original
5762 (before rewriting) address, but remembering that it had rewritten it, so it
5763 wasn't rewriting it again. This bug also had the effect of breaking the
5764 sender address verification caching when the sender address was rewritten.
5766 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5767 This has been changed so that if the ip literal address matches
5768 ignore_target_hosts, the router declines.
5770 7. Added expansion conditions match_domain, match_address, and match_local_
5771 part (NOT match_host).
5773 8. The placeholder for the Received: header didn't have a length field set.
5775 9. Added code to Exim itself and to exim_lock to test for a specific race
5776 condition that could lead to file corruption when using MBX delivery. The
5777 issue is with the lockfile that is created in /tmp. If this file is removed
5778 after a process has opened it but before that process has acquired a lock,
5779 there is the potential for a second process to recreate the file and also
5780 acquire a lock. This could lead to two Exim processes writing to the file
5781 at the same time. The added code performs the same test as UW imapd; it
5782 checks after acquiring the lock that its file descriptor still refers to
5783 the same named file.
5785 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5786 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5789 11. Added the smtp_active_hostname option. If used, this will typically be made
5790 to depend on the incoming interface address. Because $interface_address is
5791 not set up until the daemon has forked a reception process, error responses
5792 that can happen earlier (such as "too many connections") no longer contain
5795 12. If an expansion in a condition on a "warn" statement fails because a lookup
5796 defers, the "warn" statement is abandoned, and the next ACL statement is
5797 processed. Previously this caused the whole ACL to be aborted.
5799 13. Added the iplsearch lookup type.
5801 14. Added ident_timeout as a log selector.
5803 15. Added tls_certificate_verified as a log selector.
5805 16. Added a global option tls_require_ciphers (compare the smtp transport
5806 option of the same name). This controls incoming TLS connections.
5808 17. I finally figured out how to make tls_require_ciphers do a similar thing
5809 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5810 before starting the TLS session.
5812 18. Tabs are now shown as \t in -bP output.
5814 19. If the log selector return_path_on_delivery was set, Exim crashed when
5815 bouncing a message because it had too many Received: header lines.
5817 20. If two routers both had headers_remove settings, and the first one included
5818 a superfluous trailing colon, the final name in the first list and the
5819 first name in the second list were incorrectly joined into one item (with a
5820 colon in the middle).
5826 1. Added -C and -D options to the exinext utility, mainly to make it easier
5827 to include in the automated testing, but these could be helpful when
5828 multiple configurations are in use.
5830 2. The exinext utility was not formatting the output nicely when there was
5831 an alternate port involved in the retry record key, nor when there was a
5832 message id as well (for retries that were specific to a specific message
5833 and a specific host). It was also confused by IPv6 addresses, because of
5834 the additional colons they contain. I have fixed the IPv4 problem, and
5835 patched it up to do a reasonable job for IPv6.
5837 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5838 delivery, the log line now contains "pipelined" if PIPELINING was used.
5840 4. An SMTP transport process used to panic and die if the bind() call to set
5841 an explicit outgoing interface failed. This has been changed; it is now
5842 treated in the same way as a connect() failure.
5844 5. A reference to $sender_host_name in the part of a conditional expansion
5845 that was being skipped was still causing a DNS lookup. This no longer
5848 6. The def: expansion condition was not recognizing references to header lines
5849 that used bh_ and bheader_.
5851 7. Added the _cache feature to named lists.
5853 8. The code for checking quota_filecount in the appendfile transport was
5854 allowing one more file than it should have been.
5856 9. For compatibility with Sendmail, the command line option
5864 and sets the incoming protocol and host name (for trusted callers). The
5865 host name and its colon can be omitted when only the protocol is to be set.
5866 Note the Exim already has two private options, -pd and -ps, that refer to
5867 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5868 "s", but I don't think that's a major issue.
5870 10. A number of refactoring changes to the code, none of which should affect
5873 (a) The number of logging options was getting close to filling up the
5874 32-bit word that was used as a bit map. I have split them into two classes:
5875 those that are passed in the argument to log_write(), and those that are
5876 only ever tested independently outside of that function. These are now in
5877 separate 32-bit words, so there is plenty of room for expansion again.
5878 There is no change in the user interface or the logging behaviour.
5880 (b) When building, for example, log lines, the code previously used a
5881 macro that called string_cat() twice, in order to add two strings. This is
5882 not really sufficiently general. Furthermore, there was one instance where
5883 it was actually wrong because one of the argument was used twice, and in
5884 one call a function was used. (As it happened, calling the function twice
5885 did not affect the overall behaviour.) The macro has been replaced by a
5886 function that can join an arbitrary number of extra strings onto a growing
5889 (c) The code for expansion conditions now uses a table and a binary chop
5890 instead of a serial search (which was left over from when there were very
5891 few conditions). Also, it now recognizes conditions like "pam" even when
5892 the relevant support is not compiled in: a suitably worded error message is
5893 given if an attempt is made to use such a condition.
5895 11. Added ${time_interval:xxxxx}.
5897 12. A bug was causing one of the ddress fields not to be passed back correctly
5898 from remote delivery subprocesses. The field in question was not being
5899 subsequently used, so this caused to problems in practice.
5901 13. Added new log selectors queue_time and deliver_time.
5903 14. Might have fixed a bug in maildirsizefile handling that threw up
5904 "unexpected character" debug warnings, and recalculated the data
5905 unnecessarily. In any case, I expanded the warning message to give more
5908 15. Added the message "Restricted characters in address" to the statements in
5909 the default ACL that block characters like @ and % in local parts.
5911 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5912 Three changes have been made:
5914 (a) There was a serious bug; a negative response to MAIL caused the whole
5915 recipient domain to be cached as invalid, thereby blocking all messages
5916 to all local parts at the same domain, from all senders. This bug has
5917 been fixed. The domain is no longer cached after a negative response to
5918 MAIL if the sender used is not empty.
5920 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5923 (c) A new callout option, "use_sender" has been added for people who want
5924 the modified behaviour.
5930 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5933 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5934 indeed breaks things for older releases.
5936 3. Added additional logging to the case where there is a problem reading data
5937 from a filter that is running in a subprocess using a pipe, in order to
5938 try to track down a specific problem.
5940 4. Testing facility fudge: when running in the test harness and attempting
5941 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5942 getting "No route to host". Convert this to a timeout.
5944 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5947 6. Some OS don't have socklen_t but use size_t instead. This affects the
5948 fifth argument of getsockopt() amongst other things. This is now
5949 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5950 can be set for individual OS. I have set it for SunOS5, OSF1, and
5951 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5952 some earlier ones do not.
5954 7. Change 4.30/15 was not doing the test caselessly.
5956 8. The standard form for an IPv6 address literal was being rejected by address
5957 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5958 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5959 this, as well as the form without the "IPv6" on the front (but only when
5960 address literals are enabled, of course).
5962 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5964 10. Exim crashed if a message with an empty sender address specified by -f
5965 encountered a router with an errors_to setting. This could be provoked only
5966 by a command such as
5970 where an empty string was supplied; "<>" did not hit this bug.
5972 11. Installed PCRE release 4.5.
5974 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5975 remained set. It is now erased.
5977 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5978 times from message ids (which are base 36 rather than the normal 62).
5980 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5981 were being counted as actual protocol errors, and logged if the log
5982 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5983 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5984 and DATA following a set of rejected RCPTs do not count as protocol errors.
5985 In other words, Exim assumes they were pipelined, though this may not
5986 actually be the case. Of course, in all cases the client gets an
5987 appropriate error code.
5989 15. If a lookup fails in an ACL condition, a message about the failure may
5990 be available; it is used if testing the ACL cannot continue, because most
5991 such messages specify what the cause of the deferral is. However, some
5992 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5993 that caused an old message to be retained and used if a later statement
5994 caused a defer, replacing the real cause of the deferral.
5996 16. If an IP address had so many PTR records that the DNS lookup buffer
5997 was not large enough to hold them, Exim could crash while trying to process
5998 the truncated data. It now detects and logs this case.
6000 17. Further to 4.21/58, another change has been made: if (and only if) the
6001 first line of a message (the first header line) ends with CRLF, a bare LF
6002 in a subsequent header line has a space inserted after it, so as not to
6003 terminate the header.
6005 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6006 unnecessarily, used atoi() instead of strtol(), and didn't check the
6007 termination when getting file sizes from file names by regex.
6009 19. Completely re-implemented the support for maildirsize files, in the light
6010 of a number of problems with the previous contributed implementation
6011 (4.30/29). In particular:
6013 . If the quota is zero, the maildirsize file is maintained, but no quota is
6016 . If the maildir directory does not exist, it is created before any attempt
6017 to write a maildirsize file.
6019 . The quota value in the file is just a cache; if the quota is changed in
6020 the transport, the new value overrides.
6022 . A regular expression is available for excluding directories from the
6025 20. The autoreply transport checks the characters in options that define the
6026 message's headers; it allows continued headers, but it was checking with
6027 isspace() after an embedded newline instead of explicitly looking for a
6030 21. If all the "regular" hosts to which an address was routed had passed their
6031 expiry times, and had not reached their retry times, the address was
6032 bounced, even if fallback hosts were defined. Now Exim should go on to try
6035 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6036 equivalent code in the SMTP transport. Some hosts send humungous responses
6037 to HELO/EHLO, more than 1024 it seems.
6039 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6040 gives compiler warnings in some environments. I've now done it "properly",
6043 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6044 (because of problems with the built-in one) was declared to return uschar *
6045 instead of char *, causing compiler failure.
6047 25. Fixed a file descriptor leak when processing alias/forward files.
6049 26. Fixed a minor format string issue in dbfn.c.
6051 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6053 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6054 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6057 29. When a sender address is verified, it is cached, to save repeating the test
6058 when there is more than one recipient in a message. However, when the
6059 verification involves a callout, it is possible for different callout
6060 options to be set for different recipients. It is too complicated to keep
6061 track of this in the cache, so now Exim always runs a verification when a
6062 callout is required, relying on the callout cache for the optimization.
6063 The overhead is duplication of the address routing, but this should not be
6066 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6067 to be verified with callout=postmaster, and the main callout worked but the
6068 postmaster check failed, the verification correctly failed. However, if a
6069 subsequent RCPT command asked for sender verification *without* the
6070 postmaster check, incorrect caching caused this verification also to fail,
6073 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6074 it was not caching the DNS options (qualify_single, search_parents) that
6075 were used when the lookup failed. A subsequent lookup with different
6076 options therefore always gave the same answer, though there were cases
6077 where it should not have. (Example: a "domains = !$mx_any" option on a
6078 dnslookup router: the "domains" option is always processed without any
6079 widening, but the router might have qualify_single set.) Now Exim uses the
6080 cached value only when the same options are set.
6082 32. Added John Jetmore's "exipick" utility to the distribution.
6084 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6085 than a timeout (e.g. a certificate is required, and is not provided), an
6086 Exim server now closes the connection immediately. Previously it waited for
6087 the client to close - but if the client is SSL, it seems that they each
6088 wait for each other, leading to a delay before one of them times out.
6090 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6091 maintained 0.8.x compatibility because I don't think many are using it, and
6092 it is clearly obsolete.
6094 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6097 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6098 client certificate was expired. A simple patch fixes this, though I don't
6099 understand the full logic of why the verify callback is called multiple
6102 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6103 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6104 which causes problems with some clients (such as the Certicom SSL Plus
6105 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6106 disables the coutermeasure allowing Eudora to connect."
6108 38. Exim was not checking that a write() to a log file succeeded. This could
6109 lead to Bad Things if a log got too big, in particular if it hit a file
6110 size limit. Exim now panics and dies if it cannot write to a log file, just
6111 as it does if it cannot open a log file.
6113 39. Modified OS/Makefile-Linux so that it now contains
6115 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6117 The two -D definitions ensure that Exim is compiled with large file
6118 support, which makes it possible to handle log files that are bigger than
6121 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6122 instance) a domain was checked against a named list that involved a lookup,
6123 causing $domain_data to be set, then another domain was checked against the
6124 same list, then the first domain was re-checked, the value of $domain_data
6125 after the final check could be wrong. In particular, if the second check
6126 failed, it could be set empty. This bug probably also applied to
6129 41. The strip_trailing_dot option was not being applied to the address given
6130 with the -f command-line option.
6132 42. The code for reading a message's header from the spool was incrementing
6133 $received_count, but never initializing it. This meant that the value was
6134 incorrect (doubled) while delivering a message in the same process in which
6135 it was received. In the most common configuration of Exim, this never
6136 happens - a fresh exec is done - but it can happen when
6137 deliver_drop_privilege is set.
6139 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6140 it now includes up to 150 characters of the unexpected data in the log
6143 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6144 and building data strings. The size of both of these buffers was 10 000
6145 bytes - far larger than anybody would *ever* want, thought I. Needless to
6146 say, somebody hit the limit. I have increased the maximum line length to
6147 20 000 and the maximum data length of concatenated lines to 100 000. I have
6148 also fixed two bugs, because there was no checking on these buffers. Tsk,
6149 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6150 buffer is too small.
6152 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6153 lsearch lookups. Now it does.
6155 46. When parsing a route_list item in a manualroute router, a fixed-length
6156 buffer was used for the list of hosts. I made this 1024 bytes long,
6157 thinking that nobody would ever have a list of hosts that long. Wrong.
6158 Somebody had a whole pile of complicated expansion conditions, and the
6159 string was silently truncated, leading to an expansion error. It turns out
6160 that it is easier to change to an unlimited length (owing to other changes
6161 that have happened since this code was originally written) than to build
6162 structure for giving a limitation error. The length of the item that
6163 expands into the list of hosts is now unlimited.
6165 47. The lsearch lookup could not handle data where the length of text line was
6166 more than 4095 characters. Such lines were truncated, leading to shortened
6167 data being returned. It should now handle lines of any length.
6169 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6170 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6173 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6174 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6175 no longer happens in comment lines. A list of replacements is now placed
6176 at the head of all of the source files, except those whose only change is
6177 to replace PERL_COMMAND in the very first #! line.
6179 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6180 messages on the queue, with a bottom-up merge sort, using code contributed
6181 by Michael Haardt. This should make operations like -bp somewhat faster on
6182 large queues. It won't affect queue runners, except when queue_run_in_order
6185 51. Installed eximstats 1.31 in the distribution.
6187 52. Added support for SRV lookups to the dnslookup router.
6189 53. If an ACL referred to $message_body or $message_body_end, the value was not
6190 reset for any messages that followed in the same SMTP session.
6192 54. The store-handling optimization for building very long strings was not
6193 differentiating between the different store pools. I don't think this
6194 actually made any difference in practice, but I've tidied it.
6196 55. While running the routers to verify a sender address, $sender_address
6197 was still set to the sender address. This is wrong, because when routing to
6198 send a bounce to the sender, it would be empty. Therefore, I have changed
6199 it so that, while verifying a sender address, $sender_address is set to <>.
6200 (There is no change to what happens when verifying a recipient address.)
6202 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6203 target A or AAAA records (if not already returned) without resetting the
6204 qualify_single or search_parents options of the DNS resolver. These are
6205 inappropriate in this case because the targets of MX and SRV records must
6206 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6207 target that, when qualified, matched something in the local domain. These
6208 two options are now turned off when doing these lookups.
6210 57. It seems that at least some releases of Reiserfs (which does not have the
6211 concept of a fixed number of inodes) returns zero and not -1 for the
6212 number of available inodes. This interacted badly with check_spool_inodes,
6213 which assumed that -1 was the "no such thing" setting. What I have done is
6214 to check that the total number of inodes is greater than zero before doing
6215 the test of how many are available.
6217 58. When a "warn" ACL statement has a log_message modifier, the message is
6218 remembered, and not repeated. This is to avoid a lot of repetition when a
6219 message has many recipients that cause the same warning to be written.
6220 However, Exim was preserving the list of already written lines for an
6221 entire SMTP session, which doesn't seem right. The memory is now reset if a
6222 new message is started.
6224 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6225 debugging output unless log_rewrite was also set.
6227 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6228 of (int)(handle) when we know that handle contains (void *)(-1).
6230 61. The Exim daemon panic-logs an error return when it closes the incoming
6231 connection. However "connection reset by peer" seems to be common, and
6232 isn't really an error worthy of noting specially, so that particular error
6235 62. When Exim is trying to find all the local interfaces, it used to panic and
6236 die if the ioctl to get the interface flags failed. However, it seems that
6237 on at least one OS (Solaris 9) it is possible to have an interface that is
6238 included in the list of interfaces, but for which you get a failure error
6239 for this call. This happens when the interface is not "plumbed" into a
6240 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6241 failure of the "get flags" call assumes that the interface is down.
6243 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6244 makes life easier for people who are doing arithmetic on fields extracted
6245 from dates, where you often get leading zeros that should not be
6246 interpreted as octal.
6248 64. Added qualify_domain to the redirect router, to override the global
6251 65. If a pathologically long header line contained very many addresses (the
6252 report of this problem mentioned 10 000) and each of them was rewritten,
6253 Exim could use up a very large amount of memory. (It kept on making new
6254 copies of the header line as it rewrote, and never released the old ones.)
6255 At the expense of a bit more processing, the header rewriting function has
6256 been changed so that it no longer eats memory in this way.
6258 66. The generation of the Received: header has been moved from the time that a
6259 message starts to be received, to the time that it finishes. The timestamp
6260 in the Received: header should now be very close to that of the <= log
6261 line. There are two side-effects of this change:
6263 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6264 logged header lines no longer include the local Received: line, because
6265 it has not yet been created. The same applies to a copy of the message
6266 that is returned to a non-SMTP sender when a message is rejected.
6268 (b) When a filter file is tested using -bf, no additional Received: header
6269 is added to the test message. After some thought, I decided that this
6272 This change does not affect the value of $received_for. It is still set
6273 after address rewriting, but before local_scan() is called.
6275 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6277 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6278 gave an unhelpful panic error message, and a defer error. I have managed to
6279 change this behaviour so that it now rejects any supplied certificate,
6280 which seems right, as the list of acceptable certificates is empty.
6282 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6283 gave an unhelpful defer error. I have not managed to make this reject any
6284 supplied certificates, but the error message it gives is "no certificate
6285 supplied", which is not helpful.
6287 70. exigrep's output now also includes lines that are not associated with any
6288 message, but which match the given pattern. Implemented by a patch from
6289 Martin Sluka, which also tidied up the Perl a bit.
6291 71. Recipient callout verification, like sender verification, was using <> in
6292 the MAIL FROM command. This isn't really the right thing, since the actual
6293 sender may affect whether the remote host accepts the recipient or not. I
6294 have changed it to use the actual sender in the callout; this means that
6295 the cache record is now keyed on a recipient/sender pair, not just the
6296 recipient address. There doesn't seem to be a real danger of callout loops,
6297 since a callout by the remote host to check the sender would use <>.
6298 [SEE ABOVE: changed after hitting problems.]
6300 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6301 temporary errors. However, in the case of such a code being given after
6302 the end of a data transmission (i.e. after ".") Exim was failing to write
6303 a retry record for the message. (Yes, there was some broken host that was
6304 actually sending 8xx at this point.)
6306 73. An unknown lookup type in a host list could cause Exim to panic-die when
6307 the list was checked. (An example that provoked this was putting <; in the
6308 middle of a list instead of at the start.) If this happened during a DATA
6309 ACL check, a -D file could be left lying around. This kind of configuration
6310 error no longer causes Exim to die; instead it causes a defer error. The
6311 incident is still logged to the main and panic logs.
6313 74. Buglet left over from Exim 3 conversion. The message "too many messages
6314 in one connection" was written to the rejectlog but not the mainlog, except
6315 when address rewriting (yes!) was being logged.
6317 75. Added write_rejectlog option.
6319 76. When a system filter was run not as root (that is, when system_filter_user
6320 was set), the values of the $n variables were not being returned to the
6321 main process; thus, they were not subsequently available in the $sn
6324 77. Added +return_path_on_delivery log selector.
6326 78. A connection timeout was being treated differently from recipients deferred
6327 when testing hosts_max_try with a message that was older than the host's
6328 retry timeout. (The host should not be counted, thus allowing all hosts to
6329 be tried at least once before bouncing.) This may have been the cause of an
6330 occasionally reported bug whereby a message would remain on the queue
6331 longer than the retry timeout, but would be bounced if a delivery was
6332 forced. I say "may" because I never totally pinned down the problem;
6333 setting up timeout/retry tests is difficult. See also the next item.
6335 79. The ultimate address timeout was not being applied to errors that involved
6336 a combination of host plus message (for example, a timeout on a MAIL
6337 command). When an address resolved to a number of possible hosts, and they
6338 were not all tried for each delivery (e.g. because of hosts_max_try), a
6339 message could remain on the queue longer than the retry timeout.
6341 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6344 81. Fixed an obscure SMTP outgoing bug which required at least the following
6345 conditions: (a) there was another message waiting for the same server;
6346 (b) the server returned 5xx to all RCPT commands in the first message so
6347 that the message was not completed; (c) the server dropped the connection
6348 or gave a negative response to the RSET that Exim sends to abort the
6349 transaction. The observed case was a dropped connection after DATA that had
6350 been sent in pipelining mode. That is, the server had advertised PIPELINING
6351 but was not implementing it correctly. The effect of the bug was incorrect
6352 behaviour, such as trying another host, and this could lead to a crash.
6358 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6359 and daemon.c were passed as pointers to ints; they should have been
6360 pointers to socklen_t variables (which are typically unsigned ints).
6362 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6365 3. Fixed a really odd bug that affected only the testing scheme; patching a
6366 certain fixed string in the binary changed the value of another string that
6367 happened to be identical to the end of the original first string.
6369 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6370 name", it returns that address as the IP address. On some operating
6371 systems (e.g. Solaris), it also passes back the IP address string as the
6372 "host name". However, on others (e.g. Linux), it passes back an empty
6373 string. Exim wasn't checking for this, and was changing the host name to an
6374 empty string, assuming it had been canonicalized.
6376 5. Although rare, it is permitted to have more than one PTR record for a given
6377 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6378 all the names associated with an address, because they do in Solaris.
6379 However, it seems that they do not in Linux for data that comes from the
6380 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6381 I found this out when I moved to a new Linux workstation and tried to run
6382 the Exim test suite.
6384 To get round this problem I have changed the code so that it now does its
6385 own call to the DNS to look up PTR records when searching for a host name.
6386 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6387 addresses that are only in /etc/hosts are still found.
6389 This behaviour is, however, controlled by an option called host_lookup_
6390 order, which defaults to "bydns:byaddr". If people want to use the other
6391 order, or indeed, just use one or the other means of lookup, they can
6392 specify it in this variable.
6394 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6395 some operating systems, this comes back from gethostbyaddr() as an empty
6396 string, and this is what Exim used to test for. However, it seems that in
6397 other systems, "." is yielded. Exim now tests for this case too.
6399 7. The values of check_spool_space and check_log_space are now held internally
6400 as a number of kilobytes instead of an absolute number of bytes. If a
6401 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6402 kilobyte. This means that much larger values can be stored.
6404 8. Exim monitor: an attempt to get the action menu when not actually pointing
6405 at a message produces an empty menu entitled "No message selected". This
6406 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6407 no entries in it ("Shell widget menu has zero width and/or height"). So I
6408 have added a single, blank menu entry in this case.
6410 9. Added ${quote_local_part.
6412 10. MIME decoding is now applied to the contents of Subject: header lines when
6415 11. Now that a reference to $sender_host_address automatically causes a reverse
6416 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6417 host lookup before query-style lookups in lists that might use this
6418 variable. This has therefore been abolished, and the "net-" prefix is no
6419 longer necessary for query-style lookups.
6421 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6422 have been a typo for LFLAGS, so it has been changed.
6424 13. The install script calls Exim with "-C /dev/null" in order to find the
6425 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6426 to be output. However, since Exim outputs its version number before the
6427 error, it didn't break the script. It just looked ugly. I fixed this by
6428 always allowing "-C /dev/null" if the caller is root.
6430 14. Ignore overlarge ACL variable number when reading spool file - insurance
6431 against a later release with more variables having written the file.
6433 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6434 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6435 this, as well as the form without the "IPv6" on the front.
6437 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6438 OS/Makefile-Darwin file.
6440 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6441 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6443 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6444 variable. However, some OS use ints for inodes. I've added cast to long int
6445 to get rid of the compiler warning.
6447 19. I had forgotten to lock out "/../" in configuration file names when
6448 ALT_CONFIG_PREFIX was set.
6450 20. Routers used for verification do not need to specify transports. However,
6451 if such a router generated a host list, and callout was configured, Exim
6452 crashed, because it could not find a port number from the (non-existent)
6453 transport. It now assumes port 25 in this circumstance.
6455 21. Added the -t option to exigrep.
6457 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6458 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6459 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6460 them is set, LOOKUP_LSEARCH is forced.
6462 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6464 24. Added sender and host information to the "rejected by local_scan()" log
6465 line; previously there was no indication of these.
6467 25. Added .include_if_exists.
6469 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6470 Linux. It turns out that not all file systems support this. Apparently some
6471 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6472 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6473 not supported on the file descriptor, is now ignored when Exim is trying to
6474 sync a directory. This applies only to Linux.
6476 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6478 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6479 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6482 29. Added support for maildirsize files from supplied patch (modified a bit).
6484 30. The use of :fail: followed by an empty string could lead Exim to respond to
6485 sender verification failures with (e.g.):
6487 550 Verification failed for <xxx>
6488 550 Sender verify failed
6490 where the first response line was missing the '-' that indicates it is not
6491 the final line of the response.
6493 31. The loop for finding the name of the user that called Exim had a hardwired
6494 limit of 10; it now uses the value of finduser_retries, which is used for
6495 all other user lookups.
6497 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6500 33. Exim was neglecting to zero errno before one call of strtol() when
6501 expanding a string and expecting an integer value. On some systems this
6502 resulted in spurious "integer overflow" errors. Also, it was casting the
6503 result into an int without checking.
6505 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6506 did not work. The code looks as if it has *never* worked, though it appears
6507 to have been documented since at least release 1.62. I have made it work.
6509 35. The "timeout_DNS" error in retry rules, also documented since at least
6510 1.62, also never worked. As it isn't clear exactly what this means, and
6511 clearly it isn't a major issue, I have abolished the feature by treating it
6512 as "timeout", and writing a warning to the main and panic logs.
6514 36. The display of retry rules for -brt wasn't always showing the error code
6517 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6518 timeout_connect_A, timeout_connect_MX.
6520 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6521 to the empty sender.
6523 39. The daemon was not analysing the content of -oX till after it had closed
6524 stderr and disconnected from the controlling terminal. This meant that any
6525 syntax errors were only noted on the panic log, and the return code from
6526 the command was 0. By re-arranging the code a little, I've made the
6527 decoding happen first, so such errors now appear on stderr, and the return
6528 code is 1. However, the actual setting up of the sockets still happens in
6529 the disconnected process, so errors there are still only recorded on the
6532 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6533 connections (as happens on some IP stacks) was logged at start up time as
6534 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6535 it from "IPv6 and IPv4", which means that two separate sockets are being
6538 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6539 says whether AF_INET or AF_INET6 was passed as an argument.
6541 42. Exiwhat output was messed up when time zones were included in log
6544 43. Exiwhat now gives more information about the daemon's listening ports,
6545 and whether -tls-on-connect was used.
6547 44. The "port" option of the smtp transport is now expanded.
6549 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6550 silently ignored. Now an error message is written to the main and panic
6553 46. There's a new ACL modifier called "logwrite" which writes to a log file
6554 as soon as it is encountered.
6556 47. Added $local_user_uid and $local_user_gid at routing time.
6558 48. Exim crashed when trying to verify a sender address that was being
6561 49. Exim was recognizing only a space character after ".include". It now also
6562 recognizes a tab character.
6564 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6565 extracting the relevant information from the specification. The man page no
6566 longer contains scrambled data for the -d option, and I've added a section
6567 at the front about calling Exim under different names.
6569 51. Added "extra_headers" argument to the "mail" command in filter files.
6571 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6574 53. Installed eximstats 1.29.
6576 54. Added transport_filter_timeout as a generic transport option.
6578 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6579 Cc: header lines. This was required by RFC 822, but it not required by RFC
6582 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6583 incoming messages that did not have them. Now it does so only if the
6584 message originates locally, that is, if there is no associated remote host
6585 address. When Resent- header lines are present, this applies to the Resent-
6586 lines rather than the non-Resent- lines.
6588 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6589 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6591 58. Messages for configuration errors now include the name of the main
6592 configuration file - useful now that there may be more than one file in a
6593 list (.included file names were always shown).
6595 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6596 for those rare installations that do not start the daemon as root or run it
6597 setuid root. I've cut out the call to initgroups() if the daemon is not
6600 60. The Exim user and group can now be bound into the binary as text strings
6601 that are looked up at the start of Exim's processing.
6603 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6605 62. Added $mailstore_basename variable.
6607 63. Installed patch to sieve.c from Michael Haardt.
6609 64. When Exim failed to open the panic log after failing to open the main log,
6610 the original message it was trying to log was written to stderr and debug
6611 output, but if they were not available (the usual case in production), it
6612 was lost. Now it is written to syslog before the two lines that record the
6613 failures to open the logs.
6615 65. Users' Exim filters run in subprocesses under the user's uid. It is
6616 possible for a "deliver" command or an alias in a "personal" command to
6617 provoke an address rewrite. If logging of address rewriting is configured,
6618 this fails because the process is not running as root or exim. There may be
6619 a better way of dealing with this, but for the moment (because 4.30 needs
6620 to be released), I have disabled address rewrite logging when running a
6621 filter in a non-root, non-exim process.
6627 1. The buildconfig auxiliary program wasn't quoting the value set for
6628 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6629 not defined. This bug was masked in 4.22 by the effect that was fixed in
6632 2. Some messages that were rejected after a message id was allocated were
6633 shown as "incomplete" by exigrep. It no longer does this for messages that
6634 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6636 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6637 have allow_domain_literals set, the ID did not get logged in the <= line.
6638 Domain literals are now always recognized in Message-ID: header lines.
6640 4. The first argument for a ${extract expansion item is the key name or field
6641 number. Leading and trailing spaces in this item were not being ignored,
6642 causing some misleading effects.
6644 5. When deliver_drop_privilege was set, single queue runner processes started
6645 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6646 same command in the process it spins off) were not dropping privilege.
6648 6. When the daemon running as "exim" started a queue runner, it always
6649 re-executed Exim in the spun-off process. This is a waste of effort when
6650 deliver_drop_privilege is set. The new process now just calls the
6651 queue-runner function directly.
6657 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6660 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6661 ignored. Though the use of -oP was forcing the writing of a pid file, it
6662 was always written to the default place.
6664 3. If the message "no IP address found for host xxxx" is generated during
6665 incoming verification, it is now followed by identification of the incoming
6666 connection (so you can more easily find what provoked it).
6668 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6670 5. Added some features to "harden" Exim a bit more against certain attacks:
6672 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6673 be put in Local/Makefile. This is like the never_users runtime option,
6674 but it cannot be overridden. The default setting is "root".
6676 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6677 prefix string with which any file named in a -C command line option
6680 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6681 is retained for -C and -D only if the caller of Exim is root. Without
6682 it, the exim user may also use -C and -D and retain privilege.
6684 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6685 command line option is disabled.
6687 6. Macro names set by the -D option must start with an upper case letter, just
6688 like macro names defined in the configuration file.
6690 7. Added "dereference=" facility to LDAP.
6692 8. Two instances of the typo "uknown" in the source files are fixed.
6694 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6695 the Configure-Makefile script screwed up while processing it.
6697 10. Incorporated PCRE 4.4.
6699 11. The SMTP synchronization check was not operating right at the start of an
6700 SMTP session. For example, it could not catch a HELO sent before the client
6701 waited for the greeting. There is now a check for outstanding input at the
6702 point when the greeting is written. Because of the duplex, asynchronous
6703 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6704 way, but not yet received, when the check is performed.
6706 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6707 on TCP/IP sockets, because this apparently causes some broken clients to
6710 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6711 unchanged) from the Cygwin maintainer.
6713 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6714 was supported instead of testing for mbx. Effectively a typo.
6716 15. The spa authenticator server code was not checking that the input it
6717 received was valid base64.
6719 16. The debug output line for the "set" modifier in ACLs was not showing the
6720 name of the variable that was being set.
6722 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6724 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6725 lookup on its own. Something else was needed to trigger the lookup. For
6726 example, a match in host_lookup or the need for a host name in a host list.
6727 Now, if $sender_host_name is referenced and the host name has not yet been
6728 looked up, a lookup is performed. If the lookup fails, the variable remains
6729 empty, and $host_lookup_failed is set to "1".
6731 19. Added "eqi" as a case-independent comparison operator.
6733 20. The saslauthd authentication condition could segfault if neither service
6734 nor realm was specified.
6736 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6737 error message that was logged was misleading, and incoming SMTP
6738 connections were dropped. The message is now more accurate, and temporary
6739 errors are given to SMTP connections.
6741 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6742 (except RSET) with a 421 error, until QUIT is received. However, it was
6743 failing to send a response to QUIT.
6745 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6746 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6747 ACL is now run after the helo_try_verify_hosts code.
6749 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6750 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6751 case-independent, so other case variants are also recognized). Apparently
6752 some systems use these upper case variants.
6754 25. If more than two messages were waiting for the same host, and a transport
6755 filter was specified for the transport, Exim sent two messages over the
6756 same TCP/IP connection, and then failed with "socket operation on non-
6757 socket" when it tried to send the third.
6759 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6761 27. The extern definition of crypt16() in expand.c was not being excluded when
6762 the OS had its own crypt16() function.
6764 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6765 as a preferred synonym for return_size_limit, both as an option and as an
6768 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6770 30. Changed the default configuration ACL to relax the local part checking rule
6771 for addresses that are not in any local domains. For these addresses,
6772 slashes and pipe symbols are allowed within local parts, but the sequence
6773 /../ is explicitly forbidden.
6775 31. SPA server authentication was not clearing the challenge buffer before
6778 32. log_message in a "warn" ACL statement was writing to the reject log as
6779 well as to the main log, which contradicts the documentation and doesn't
6780 seem right (because no rejection is happening). So I have stopped it.
6782 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6783 However, I am unable to do any testing of this.
6785 34. Fixed an infelicity in the appendfile transport. When checking directories
6786 for a mailbox, to see if any needed to be created, it was accidentally
6787 using path names with one or more superfluous leading slashes; tracing
6788 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6790 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6791 contents are added to the log line that is written for every discarded
6792 recipient. (Previously a log_message setting was ignored.)
6794 36. The ${quote: operator now quotes the string if it is empty.
6796 37. The install script runs exim in order to find its version number. If for
6797 some reason other than non-existence or emptiness, which it checks, it
6798 could not run './exim', it was installing it with an empty version number,
6799 i.e. as "exim-". This error state is now caught, and the installation is
6802 38. An argument was missing from the function that creates an error message
6803 when Exim fails to connect to the socket for saslauthd authentication.
6804 This could cause Exim to crash, or give a corrupted message.
6806 39. Added isip, isip4, and isip6 to ${if conditions.
6808 40. The ACL variables $acl_xx are now saved with the message, and can be
6809 accessed later in routers, transports, and filters.
6811 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6812 strings in the file are not string-expanded.
6814 42. If a MAIL command specified a SIZE value that was too large to fit into an
6815 int variable, the check against message_size_limit failed. Such values are
6816 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6817 day this will have to be increased, but I don't think I want to be around
6818 when emails are that large.
6825 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6826 iconv() is not standard in FreeBSD.
6828 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6829 IPv6 enabled. The observed symptom was a segmentation fault on return from
6830 the function os_common_find_running_interfaces() in src/os.c.
6832 3. In the check_special_case() function in daemon.c I had used "errno" as an
6833 argument name, which causes warnings on some systems. This was basically a
6834 typo, since it was named "eno" in the comments!
6836 4. The code that waits for the clock to tick (at a resolution of some fraction
6837 of a second) so as to ensure message-id uniqueness was always waiting for
6838 at least one whole tick, when it could have waited for less. [This is
6839 almost certainly not relevant at current processor speeds, where it is
6840 unlikely to ever wait at all. But we try to future-proof.]
6842 5. The function that sleeps for a time interval that includes fractions of a
6843 second contained a race. It did not block SIGALRM between setting the
6844 timer, and suspending (a couple of lines later). If the interval was short
6845 and the sigsuspend() was delayed until after it had expired, the suspension
6846 never ended. On busy systems this could lead to processes getting stuck for
6849 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6850 process, before it forks any delivery processes. The open lookup caching
6851 mechanism meant that the open file or database connection was passed into
6852 the delivery process. The problem was that delivery processes always tidy
6853 up cached lookup data. This could cause a problem for the next delivery
6854 process started by the queue runner, because the external queue runner
6855 process does not know about the closure. So the next delivery process
6856 still has data in the lookup cache. In the case of a file lookup, there was
6857 no problem because closing a file descriptor in a subprocess doesn't affect
6858 the parent. However, if the lookup was caching a connection to a database,
6859 the connection was closed, and the second delivery process was likely to
6860 see errors such as "PGSQL: query failed: server closed the connection
6861 unexpectedly". The problem has been fixed by closing all cached lookups
6862 in a queue runner before running a delivery process.
6864 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6865 seem to have the "const" qualifier which it has on other OS. I've
6868 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6869 *of the same type* (client or server) with the same public name that an
6870 error should be diagnosed.
6872 9. When Exim looked up a host name for an IP address, but failed to find the
6873 original IP address when looking up the host name (a safety check), it
6874 output the message "<ip address> does not match any IP for NULL", which was
6875 confusing, to say the least. The bug was that the host name should have
6876 appeared instead of "NULL".
6878 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6879 user that is built into the binary, and the -C or -D options is used, root
6880 privilege is dropped before the configuration file is read. In addition,
6881 logging is switched to stderr instead of the normal log files. If the
6882 configuration then re-defines the Exim user, the unprivileged environment
6883 is probably not what is expected, so Exim logs a panic warning message (but
6886 However, if deliver_drop_privilege is set, the unprivileged state may well
6887 be exactly what is intended, so the warning has been cut out in that case,
6888 and Exim is allowed to try to write to its normal log files.
6894 1. smtp_return_error_details was not giving details for temporary sender
6895 or receiver verification errors.
6897 2. Diagnose a configuration error if two authenticators have the same public
6900 3. Exim used not to create the message log file for a message until the first
6901 delivery attempt. This could be confusing when incoming messages were held
6902 for policy or load reasons. The message log file is now created at the time
6903 the message is received, and an initial "Received" line is written to it.
6905 4. The automatically generated man page for command line options had a minor
6906 bug that caused no ill effects; however, a more serious problem was that
6907 the procedure for building the man page automatically didn't always
6908 operate. Consequently, release 4.20 contains an out-of-date version. This
6909 shouldn't happen again.
6911 5. When building Exim with embedded Perl support, the script that builds the
6912 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6913 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6915 6. The freeze_tell option was not being used for messages that were frozen on
6916 arrival, either by an ACL or by local_scan().
6918 7. Added the smtp_incomplete_transaction log selector.
6920 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6921 was accepting AUTH without a new EHLO.
6923 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6924 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6927 10. Logging of TCP/IP connections (when configured) now happens in the main
6928 daemon process instead of the child process, so that the TCP/IP connection
6929 count is more accurate (but it can never be perfect).
6931 11. The use of "drop" in a nested ACL was not being handled correctly in the
6932 outer ACL. Now, if condition failure induced by the nested "drop" causes
6933 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6934 or "require"), the connection is dropped.
6936 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6937 that yield "discard" can now be used with an "accept" or a "discard" verb,
6938 but an error is generated for any others (because I can't see a useful way
6939 to define what should happen).
6941 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6942 are now processed in the same way as lines in the Exim configuration file.
6943 In particular, continuation lines are supported.
6945 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6947 15. Added -ti meaning -t -i.
6949 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6950 domains, and warn by logging if others are found.
6952 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6953 returned from GIFCONF when Exim is trying to find the list of interfaces on
6954 a host. The code in os.c has been modified to copy each ifreq to an aligned
6955 structure in all cases.
6957 Also, in some cases, the returned ifreq's were being copied to a 'struct
6958 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6959 means the last couple of bytes of an IPv6 address could be chopped if the
6960 ifreq contained only a normal sockaddr (14 bytes storage).
6962 18. Named domain lists were not supported in the hosts_treat_as_local option.
6963 An entry such as +xxxx was not recognized, and was treated as a literal
6966 19. Ensure that header lines added by a DATA ACL are included in the reject log
6967 if the ACL subsequently rejects the message.
6969 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6970 MD5 (which is deprecated).
6972 21. When testing a filter file using -bf, Exim was writing a message when it
6973 took the sender from a "From " line in the message, but it was not doing so
6974 when it took $return_path from a Return-Path: header line. It now does.
6976 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6977 with a valid header line field name (a series of printing characters
6978 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6980 23. Changed "disc" in the source to "disk" to conform to the documentation and
6981 the book and for uniformity.
6983 24. Ignore Sendmail's -Ooption=value command line item.
6985 25. When execve() failed while trying to run a command in a pipe transport,
6986 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6987 could be confused with a return value of 69 from the command itself. This
6988 has been changed to 127, the value the shell returns if it is asked to run
6989 a non-existent command. The wording for the related log line suggests a
6990 non-existent command as the problem.
6992 26. If received_header_text expands to an empty string, do not add a Received:
6993 header line to the message. (Well, it adds a token one on the spool, but
6994 marks it "old" so that it doesn't get used or transmitted.)
6996 27. Installed eximstats 1.28 (addition of -nt option).
6998 28. There was no check for failure on the call to getsockname() in the daemon
6999 code. This can fail if there is a shortage of resources on the system, with
7000 ENOMEM, for example. A temporary error is now given on failure.
7002 29. Contrary to the C standard, it seems that in some environments, the
7003 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7004 program. Exim now does this explicitly; it affects the formatting of
7005 timestamps using strftime().
7007 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7008 complaints. I've now initialized all the variables, to avoid this.
7010 32. Header lines added by a system filter were not being "seen" during
7011 transport-time rewrites.
7013 33. The info_callback() function passed to OpenSSL is set up with type void
7014 (*)(SSL *, int, int), as described somewhere. However, when calling the
7015 function (actually a macro) that sets it up, the type void(*)() is
7016 expected. I've put in a cast to prevent warnings from picky compilers.
7018 34. If a DNS black list lookup found a CNAME record, but there were no A
7019 records associated with the domain it pointed at, Exim crashed.
7021 35. If a DNS black list lookup returned more than one A record, Exim ignored
7022 all but the first. It now scans all returned addresses if a particular IP
7023 value is being sought. In this situation, the contents of the
7024 $dnslist_value variable are a list of all the addresses, separated by a
7027 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7028 to accept a host name and all its aliases if the forward lookup for any of
7029 them yielded the IP address of the incoming connection. Now it accepts only
7030 those names whose forward lookup yields the correct IP address. Any other
7031 names are discarded. This closes a loophole whereby a rogue DNS
7032 administrator could create reverse DNS records to break through a
7033 wildcarded host restriction in an ACL.
7035 37. If a user filter or a system filter that ran in a subprocess used any of
7036 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7037 the wrong values were passed to the pipe command ($thisaddress had the
7038 value of $0, $0 had the value of $1, etc). This bug was introduced by
7039 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7041 38. Improved the line breaking for long SMTP error messages from ACLs.
7042 Previously, if there was no break point between 40 and 75 characters, Exim
7043 left the rest of the message alone. Two changes have been made: (a) I've
7044 reduced the minimum length to 35 characters; (b) if it can't find a break
7045 point between 35 and 75 characters, it looks ahead and uses the first one
7046 that it finds. This may give the occasional overlong line, but at least the
7047 remaining text gets split now.
7049 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7050 file descriptors might be low, and that setting 1000 would always raise it.
7051 It turns out that in some environments, the limit is already over 1000 and
7052 that lowering it causes trouble. So now Exim takes care not to decrease it.
7054 40. When delivering a message, the value of $return_path is set to $sender_
7055 address at the start of routing (routers may change the value). By an
7056 oversight, this default was not being set up when an address was tested by
7057 -bt or -bv, which affected the outcome if any router or filter referred to
7060 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7061 log only when "message" or "log_message" are set. However, if one of the
7062 conditions was an address verification, or a call to a nested ACL, the
7063 messages generated by the underlying test were being passed through. This
7064 no longer happens. The underlying message is available in $acl_verify_
7065 message for both "message" and "log_message" expansions, so it can be
7066 passed through if needed.
7068 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7069 new expansion $bh_ to give the encoded byte string without charset
7070 translation. Translation happens only if iconv() is available; HAVE_ICONV
7071 indicates this at build time. HEADERS_CHARSET gives the charset to
7072 translate to; headers_charset can change it in the configuration, and
7073 "headers charset" can change it in an individual filter file.
7075 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7076 that creates RFC 2047 encoded "words" labels them as that charset instead
7077 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7078 expansion operator; (ii) when Exim creates a From: line for a local
7079 message; (iii) when a header line is rewritten to include a "phrase" part.
7081 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7082 buggy, causing it to skip the first lines of messages whose message ID
7083 ended in 'D'. This would not have bitten before Exim release 4.14, because
7084 message IDs were unlikely to end in 'D' before then. The effect was to have
7085 incorrect size information for certain domains.
7087 45. #include "config.h" was missing at the start of the crypt16.c module. This
7088 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7091 46. If there was a timeout during a "random" callout check, Exim treated it as
7092 a failure of the random address, and carried on sending RSET and the real
7093 address. If the delay was just some slowness somewhere, the response to the
7094 original RCPT would be taken as a response to RSET and so on, causing
7095 mayhem of various kinds.
7097 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7098 when I implemented it. It didn't allow for the fact that some option values
7099 may legitimately be negative (e.g. size_addition), and it didn't even do
7100 the right test for positive values.
7102 48. Domain names in DNS records are case-independent. Exim always looks them up
7103 in lower case. Some resolvers return domain names in exactly the case they
7104 appear in the zone file, that is, they may contain uppercase letters. Not
7105 all resolvers do this - some return always lower case. Exim was treating a
7106 change of case by a resolver as a change of domain, similar to a widening
7107 of a domain abbreviation. This triggered its re-routing code and so it was
7108 trying to route what was effectively the same domain again. This normally
7109 caused routing to fail (because the router wouldn't handle the domain
7110 twice). Now Exim checks for this case specially, and just changes the
7111 casing of the domain that it ultimately uses when it transmits the message
7114 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7117 50. If a filter generated a file delivery with a non-absolute name (possible if
7118 no home directory exists for the router), the forbid_file option was not
7121 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7122 the existing equality matching.
7124 52. Exim was using ints instead of ino_t variables in some places where it was
7125 dealing with inode numbers.
7127 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7128 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7129 called TMPDIR, and if it finds it is different, it changes its value.
7131 54. The smtp_printf() function is now made available to local_scan() so
7132 additional output lines can be written before returning. There is also an
7133 smtp_fflush() function to enable the detection of a dropped connection.
7134 The variables smtp_input and smtp_batched_input are exported to
7137 55. Changed the default runtime configuration: the message "Unknown user"
7138 has been removed from the ACL, and instead placed on the localuser router,
7139 using the cannot_route_message feature. This means that any verification
7140 failures that generate their own messages won't get overridden. Similarly,
7141 the "Unrouteable address" message that was in the ACL for unverifiable
7142 relay addresses has also been removed.
7144 56. Added hosts_avoid_esmtp to the smtp transport.
7146 57. The exicyclog script was not checking for the esoteric option
7147 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7148 will work only if exicyclog is run under the appropriate euid.
7150 58. Following a discussion on the list, the rules by which Exim recognises line
7151 endings on incoming messages have been changed. The -dropcr and drop_cr
7152 options are now no-ops, retained only for backwards compatibility. The
7153 following line terminators are recognized: LF CRLF CR. However, special
7154 processing applies to CR:
7156 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7157 nor a local message in the state where . is a terminator.
7159 (ii) If a bare CR is encountered in a header line, an extra space is added
7160 after the line terminator so as not to end the header. The reasoning
7161 behind this is that bare CRs in header lines are most likely either
7162 to be mistakes, or people trying to play silly games.
7164 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7165 was being incorrectly given as 18 bytes larger than it should have been.
7166 This is a VOB (very old bug).
7168 60. This may never have affected anything current, but just in case it has:
7169 When the local host is found other than at the start of a list of hosts,
7170 the local host, those with the same MX, and any that follow, are discarded.
7171 When the list in question was part of a longer list of hosts, the following
7172 hosts (not currently being processed) were also being discarded. This no
7173 longer happens. I'm not sure if this situation could ever has previously
7176 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7179 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7180 header. This is something that is recommended in a new Internet Draft, and
7181 is something that is documented as being done by Sendmail. There are two
7182 possible values. For messages generated by the autoreply transport, Exim
7185 Auto-Submitted: auto-replied
7187 whereas for all other generated messages (e.g. bounces) it adds
7189 Auto-Submitted: auto-generated
7191 63. The "personal" condition in filters now includes a test for the
7192 Auto-Submitted: header. If it contains the string "auto-" the message it
7193 not considered personal.
7195 64. Added rcpt_include_affixes as a generic transport option.
7197 65. Added queue_only_override (default true).
7199 66. Added the syslog_duplication option.
7201 67. If what should have been the first header line of a message consisted of
7202 a space followed by a colon, Exim was mis-interpreting it as a header line.
7203 It isn't of course - it is syntactically invalid and should therefore be
7204 treated as the start of the message body. The misbehaviour could have
7205 caused a number of strange effects, including loss of data in subsequent
7206 header lines, and spool format errors.
7208 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7209 client host had authenticated. This control can now be exercised by an ACL
7210 for more flexibility.
7212 69. By default, callouts do not happen when testing with -bh. There is now a
7213 variant, -bhc, which does actually run the callout code, including
7214 consulting and updating the callout cache.
7216 70. Added support for saslauthd authentication, courtesy of Alexander
7219 71. If statvfs() failed on the spool or log directories while checking their
7220 size for availability, Exim confusingly gave the error "space shortage".
7221 Furthermore, in debugging mode it crashed with a floating point exception.
7222 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7223 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7224 serious problem, Exim now writes to the main and panic logs when this
7225 happens, with details of the failure. It then refuses to accept the
7226 incoming message, giving the message "spool directory problem" or "log
7227 directory problem" with a 421 code for SMTP messages.
7229 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7230 0, 1, and 2 exist, because some OS complain for execs without them (see
7231 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7232 descriptors. However, the code omitted to check that the open succeeded,
7233 causing mysterious errors if for some reason the permissions on /dev/null
7234 got screwed. Now Exim writes a message to the main and panic logs, and
7235 bombs out if it can't open /dev/null.
7237 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7238 interact so that it is all more flexible. It is supposed to remain
7239 backwards compatible. Also added extra_local_interfaces.
7241 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7242 to bomb out with an assertion failure - to the client this appears as a
7243 connection drop. This problem occurs in the part of the code that was taken
7244 from the Samba project. Fortunately, the assertion is in a very simple
7245 function, so I have fixed this by reproducing the function inline in the
7246 one place where it is called, and arranging for authentication to fail
7247 instead of killing the process with assert().
7249 75. The SPA client code was not working when the server requested OEM rather
7250 than Unicode encoding.
7252 76. Added code to make require_files with a specific uid setting more usable in
7253 the case where statting the file as root fails - usually a non-root-mounted
7254 NFS file system. When this happens and the failure is EACCES, Exim now
7255 forks a subprocess and does the per-uid checking as the relevant uid.
7257 77. Added process_log_path.
7259 78. If log_file_path was not explicitly set, a setting of check_log_space or
7260 check_log_inodes was ignored.
7262 79. If a space check for the spool or log partitions fails, the incident is now
7263 logged. Of course, in the latter case the data may get lost...
7265 80. Added the %p formatting code to string_format() so that it can be used to
7266 print addresses in debug_print(). Adjusted all the address printing in the
7267 debugging in store.c to use %p rather than %d.
7269 81. There was a concern that a line of code in smtp_in.c could overflow a
7270 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7271 initially expressed, the concern was not well-founded, because trailing
7272 spaces are removed early. However, if the trailing spaces were followed by
7273 a NULL, they did not get removed, so the overflow was possible. Two fixes
7276 (a) I re-wrote the offending code in a cleaner fashion.
7277 (b) If an incoming SMTP command contains a NULL character, it is rejected
7280 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7281 runs initgroups(), so that if the Exim user is in any additional groups,
7282 they will be used during message reception.
7288 The change log for 4.20 and earlier releases has been archived.