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 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
18 JH/03 Debug output for ACL now gives the config file name and line number for
21 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
23 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
25 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
26 buffer overrun for (non-chunking) other transports.
28 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
29 TLS1.3, means that a server rejecting a client certificate is not visible
30 to the client until the first read of encrypted data (typically the
31 response to EHLO). Add detection for that case and treat it as a failed
32 TLS connection attempt, so that the normal retry-in-clear can work (if
35 JB/01 Big 2375: fix expansions of 822 addresses having comments in local-part
36 and/or domain. Found and fixed by Jason Betts.
42 JH/01 Remove code calling the customisable local_scan function, unless a new
43 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
45 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
46 non-signal-safe functions being used.
48 JH/03 Bug 2269: When presented with a received message having a stupidly large
49 number of DKIM-Signature headers, disable DKIM verification to avoid
50 a resource-consumption attack. The limit is set at twenty.
52 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
53 report of oldest_pass in ${authres } in consequence, and separate out
54 some descriptions of reasons for verification fail.
56 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
57 files in the spool were present and unlocked. A queue-runner could spot
58 them, resulting in a duplicate delivery. Fix that by doing the unlock
59 after the unlink. Investigation by Tim Stewart. Take the opportunity to
60 add more error-checking on spoolfile handling while that code is being
63 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
64 No known attacks, no CVE, this is defensive hardening.
66 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
67 a queue-runner could start a delivery while other operations were ongoing.
68 Cutthrough delivery was a common victim, resulting in duplicate delivery.
69 Found and investigated by Tim Stewart. Fix by using the open message data
70 file handle rather than opening another, and not locally closing it (which
71 releases a lock) for that case, while creating the temporary .eml format
72 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
74 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
75 $sender_verify_failure/$recipient_verify_failure to "random".
77 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
80 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
81 Previously this would segfault.
83 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
86 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
87 like zero, since the resolver should be doing this for us, But we need one
88 as a CNAME but no MX presence gets the CNAME returned; we need to check
89 that doesn't point to an MX to declare it "no MX returned" rather than
90 "error, loop". A new main option is added so the older capability of
91 following some limited number of chain links is maintained.
93 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
95 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
96 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
97 now-deprecated earlier definitions used only specified the range up to TLS
98 1.2 (in the older-version library docs).
100 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
102 JH/15 Rework TLS client-side context management. Stop using a global, and
103 explicitly pass a context around. This enables future use of TLS for
104 connections to service-daemons (eg. malware scanning) while a client smtp
105 connection is using TLS; with cutthrough connections this is quite likely.
107 JH/16 Fix ARC verification to do AS checks in reverse order.
109 JH/17 Support a "tls" option on the ${readsocket } expansion item.
111 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
112 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
113 Previously the "utf8" would be re-prepended for every additional message.
115 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
116 Previously thery were accepted, resulting in issues when attempting to
117 forward messages to a non-supporting MTA.
119 PP/02 Let -n work with printing macros too, not just options.
121 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
122 one parent address was copied, and bogus data was used at delivery-logging
123 time. Either a crash (after delivery) or bogus log data could result.
124 Discovery and analysis by Tim Stewart.
126 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
127 Previously if the string ended mid-character, we did not insert the
128 promised '?' replacement.
130 PP/04 Documentation: current string operators work on bytes, not codepoints.
132 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
133 should pack well giving a smaller memory footprint so better caching and
134 therefore performance. Group the declarations where this can't be done so
135 that the byte-sized flag variables are not interspersed among pointer
136 variables, giving a better chance of good packing by the compiler.
138 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
139 non-null, to avoid issues with sites running BATV. Previously reports were
140 sent with an empty envelope sender so looked like bounces.
142 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
143 The ignore_error flag wasn't being returned from the filter subprocess so
144 was not set for later routers. Investigation and fix by Matthias Kurz.
146 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
147 and a msg:complete for the whole, when a message is manually removed using
148 -Mrm. Developement by Matthias Kurz, hacked on by JH.
150 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
151 a "Gnu special" function, asprintf() in the DB utility binary builds; I
152 hope that is portable enough.
154 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
155 requiring a known-CA anchor certificate; make it now rely entirely on the
156 TLSA as an anchor. Checking the name on the leaf cert against the name
157 on the A-record for the host is still done for TA (but not for EE mode).
159 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
160 included in delivery lines for non-proxied connections, when compiled with
161 SUPPORT_SOCKS and running with proxy logging enabled.
163 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
164 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
165 move the existing event to fire before the normal logging of message
166 failure so that custom logging is bracketed by normal logging.
168 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
169 msg:fail:internal event. Developement by Matthias Kurz.
171 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
172 far too small for todays use of crypto signatures stored there. Go all
173 the way to the max DNS message size of 64kB, even though this might be
174 overmuch for IOT constrained device use.
176 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
177 copy a string over itself. The library routine is documented as not
178 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
180 JH/32 For main options check_spool_space and check_inode_space, where the
181 platform supports 64b integers, support more than the previous 2^31 kB
182 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
183 the previous G, M, k.
185 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
186 $authenticated_fail_id variable on authentication failure. Previously
189 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
190 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
191 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
192 GNUTLS_SEC_PARAM_MEDIUM.
194 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
195 side. Previously we would continue as if no SNI had been received.
197 JH/36 Harden the handling of string-lists. When a list consisted of a sole
198 "<" character, which should be a list-separator specification, we walked
199 off past the nul-terimation.
201 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
202 causes) even when the retry time is not yet met. Previously they were
203 not, meaning that when (say) an account was over-quota and temp-rejecting,
204 and multiple senders' messages were queued, only one sender would get
205 notified on each configured delay_warning cycle.
207 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
209 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
210 cipher-suites, an error can be left on the stack even for a succeeding
211 accept; this results in impossible error messages when a later operation
214 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
215 return error codes indicating retry. Under TLS1.3 this becomes required.
217 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
218 it only wrote the new authenticators, resulting in a lack of tracking of
219 peer changes of ESMTP extensions until the next cache flush.
221 JH/41 Fix the loop reading a message header line to check for integer overflow,
222 and more-often against header_maxsize. Previously a crafted message could
223 induce a crash of the recive process; now the message is cleanly rejected.
225 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
226 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
232 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
233 When redis_servers is set to a list of > 1 element, and the Redis servers
234 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
235 case of MOVED into a DEFER case instead, thus moving the query onto the
236 next server in the list. For a cluster of N elements, all N servers must
237 be defined in redis_servers.
239 GF/02 Catch and remove uninitialized value warning in exiqsumm
240 Check for existence of @ARGV before looking at $ARGV[0]
242 JH/01 Replace the store_release() internal interface with store_newblock(),
243 which internalises the check required to safely use the old one, plus
244 the allocate and data copy operations duplicated in both (!) of the
245 extant use locations.
247 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
248 modifier. This matches the restriction on the commandline.
250 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
251 Previously only the last row was returned.
253 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
254 we assumed that tags in the header were well-formed, and parsed the
255 element content after inspecting only the first char of the tag.
256 Assumptions at that stage could crash the receive process on malformed
259 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
260 While running the DKIM ACL we operate on the Permanent memory pool so that
261 variables created with "set" persist to the DATA ACL. Also (at any time)
262 DNS lookups that fail create cache records using the Permanent pool. But
263 expansions release any allocations made on the current pool - so a dnsdb
264 lookup expansion done in the DKIM ACL releases the memory used for the
265 DNS negative-cache, and bad things result. Solution is to switch to the
266 Main pool for expansions.
267 While we're in that code, add checks on the DNS cache during store_reset,
268 active in the testsuite.
269 Problem spotted, and debugging aided, by Wolfgang Breyha.
271 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
272 When none of the hosts presented to a transport match an already-open
273 connection, close it and proceed with the list. Previously we would
274 queue the message. Spotted by Lena with Yahoo, probably involving
277 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
278 Previously a spurious "250 OK id=" response was appended to the proper
281 JH/08 The "support for" informational output now, which built with Content
282 Scanning support, has a line for the malware scanner interfaces compiled
283 in. Interface can be individually included or not at build time.
285 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
286 by the template makefile "src/EDITME". The "STREAM" support for an older
287 ClamAV interface method is removed.
289 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
290 rows affected is given instead).
292 JH/11 The runtime Berkeley DB library version is now additionally output by
293 "exim -d -bV". Previously only the compile-time version was shown.
295 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
296 SMTP connection. Previously, when one had more receipients than the
297 first, an abortive onward connection was made. Move to full support for
298 multiple onward connections in sequence, handling cutthrough connection
299 for all multi-message initiating connections.
301 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
302 routers. Previously, a multi-recipient message would fail to match the
303 onward-connection opened for the first recipient, and cause its closure.
305 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
306 a timeout on read on a GnuTLS initiating connection, resulting in the
307 initiating connection being dropped. This mattered most when the callout
308 was marked defer_ok. Fix to keep the two timeout-detection methods
311 JH/15 Relax results from ACL control request to enable cutthrough, in
312 unsupported situations, from error to silently (except under debug)
313 ignoring. This covers use with PRDR, frozen messages, queue-only and
316 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
318 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
319 metadata, resulting in a crash in free().
321 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
322 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
323 Broken also in d185889f4, with init system revamp.
325 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
326 Previously we abruptly closed the connection after reading a malware-
327 found indication; now we go on to read the "scan ok" response line,
330 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
331 ACL. Previously, a crash would result.
333 JH/19 Speed up macro lookups during configuration file read, by skipping non-
334 macro text after a replacement (previously it was only once per line) and
335 by skipping builtin macros when searching for an uppercase lead character.
337 JH/20 DANE support moved from Experimental to mainline. The Makefile control
338 for the build is renamed.
340 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
341 was allocated for every new TLS startup, meaning one per message. Fix
342 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
344 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
345 reported the original. Fix to report (as far as possible) the ACL
346 result replacing the original.
348 JH/23 Fix memory leak during multi-message connections using STARTTLS under
349 OpenSSL. Certificate information is loaded for every new TLS startup,
350 and the resources needed to be freed.
352 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
354 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
357 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
358 DATA response info to the (existing) per-recipient response info for
359 the "C=" log element. It can have useful tracking info from the
360 destination system. Patch from Simon Arlott.
362 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
363 length value. Previously this would segfault.
365 HS/02 Support Avast multiline protoocol, this allows passing flags to
366 newer versions of the scanner.
368 JH/28 Ensure that variables possibly set during message acceptance are marked
369 dead before release of memory in the daemon loop. This stops complaints
370 about them when the debug_store option is enabled. Discovered specifically
371 for sender_rate_period, but applies to a whole set of variables.
372 Do the same for the queue-runner and queue-list loops, for variables set
373 from spool message files. Do the same for the SMTP per-message loop, for
374 certain variables indirectly set in ACL operations.
376 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
377 as a multi-recipient message from a mailinglist manager). The coding had
378 an arbitrary cutoff number of characters while checking for more input;
379 enforced by writing a NUL into the buffer. This corrupted long / fast
380 input. The problem was exposed more widely when more pipelineing of SMTP
381 responses was introduced, and one Exim system was feeding another.
382 The symptom is log complaints of SMTP syntax error (NUL chars) on the
383 receiving system, and refused recipients seen by the sending system
384 (propating to people being dropped from mailing lists).
385 Discovered and pinpointed by David Carter.
387 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
388 replaced by the ${authresults } expansion.
390 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
392 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
393 allows proper process termination in container environments.
395 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
396 Previously the "final dot" had a newline after it; ensure it is CR,LF.
398 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
399 and "err_perm", deprecated since 4.83 when the RFC-defined words
400 "temperror" and "permerror" were introduced.
402 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
403 transport-filters or DKIM-signing. The restriction was lost in the
404 consolidation of verify-callout and delivery SMTP handling.
405 Extend the restriction to also cover ARC-signing.
407 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
408 in defer=pass mode supply a 450 to the initiator. Previously the message
411 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
412 tls_require_ciphers is used as before.
414 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
415 "pass_unscanned". Only tmpfails from the scanner are written to
416 the paniclog, as they may require admin intervention (permission
417 denied, license issues). Other scanner errors (like decompression
418 bombs) do not cause a paniclog entry.
420 JH/36 Fix reinitialisation of DKIM logging variable between messages.
421 Previously it was possible to log spurious information in receive log
424 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
425 triggered odd behaviour from Outlook Express clients.
427 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
430 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
431 since the IETF WG has not yet settled on that versus the original
432 "bare" representation.
434 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
435 Previously the millisecond value corrupted the output.
436 Fix also for syslog_pid=no and log_selector +pid, for which the pid
437 corrupted the output.
443 JH/01 Rework error string handling in TLS interface so that the caller in
444 more cases is responsible for logging. This permits library-sourced
445 string to be attached to addresses during delivery, and collapses
446 pairs of long lines into single ones.
448 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
449 during configuration. Wildcards are allowed and expanded.
451 JH/02 Rework error string handling in DKIM to pass more info back to callers.
452 This permits better logging.
454 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
455 do not close it down and have the child transport start it up again on
456 the passed-on TCP connection. Instead, proxy the child (and any
457 subsequent ones) for TLS via a unix-domain socket channel. Logging is
458 affected: the continued delivery log lines do not have any DNSSEC, TLS
459 Certificate or OCSP information. TLS cipher information is still logged.
461 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
462 identical IP addresses on different listening ports. Will also affect
465 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
466 add noisy ifdef guards to special-case this sillyness.
467 Patch from Bernd Kuhls.
469 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
470 than 255 are no longer allowed.
472 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
473 Disable the session-cache too, which might reduce our load. Since we
474 currrectly use a new context for every connection, both as server and
475 client, there is no benefit for these.
476 GnuTLS appears to not support tickets server-side by default (we don't
477 call gnutls_session_ticket_enable_server()) but client side is enabled
478 by default on recent versions (3.1.3 +) unless the PFS priority string
481 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
482 <https://reproducible-builds.org/specs/source-date-epoch/>.
484 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
485 the check for any unsuccessful recipients did not notice the limit, and
486 erroneously found still-pending ones.
488 JH/08 Pipeline CHUNKING command and data together, on kernels that support
489 MSG_MORE. Only in-clear (not on TLS connections).
491 JH/09 Avoid using a temporary file during transport using dkim. Unless a
492 transport-filter is involved we can buffer the headers in memory for
493 creating the signature, and read the spool data file once for the
494 signature and again for transmission.
496 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
497 4.77 as the kernel support then wasn't solid, having issues in 64bit
498 mode. Now, it's been long enough. Add support for FreeBSD also.
500 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
501 case where the routing stage had gathered several addresses to send to
502 a host before calling the transport for the first, we previously failed
503 to close down TLS in the old transport process before passing the TCP
504 connection to the new process. The new one sent a STARTTLS command
505 which naturally failed, giving a failed delivery and bloating the retry
506 database. Investigation and fix prototype from Wolfgang Breyha.
508 JH/12 Fix check on SMTP command input synchronisation. Previously there were
509 false-negatives in the check that the sender had not preempted a response
510 or prompt from Exim (running as a server), due to that code's lack of
511 awareness of the SMTP input buffering.
513 PP/04 Add commandline_checks_require_admin option.
514 Exim drops privileges sanely, various checks such as -be aren't a
515 security problem, as long as you trust local users with access to their
516 own account. When invoked by services which pass untrusted data to
517 Exim, this might be an issue. Set this option in main configuration
518 AND make fixes to the calling application, such as using `--` to stop
521 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
522 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
523 responses to those, into a single TLS record each way (this usually means
524 a single packet). As a side issue, smtp_enforce_sync now works on TLS
527 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
528 affects you only if you're dancing at the edge of the param size limits.
529 If you are, and this message makes sense to you, then: raise the
530 configured limit or use OpenSSL 1.1. Nothing we can do for older
533 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
534 cmdline element to get the documented default one. Previously it was
537 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
538 are unsuable for non-message-specific reasons - eg. connection timeout,
539 banner-time rejection.
541 JH/16 Fix logging of delivery remote port, when specified by router, under
544 PP/06 Repair manualroute's ability to take options in any order, even if one
545 is the name of a transport.
548 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
550 JH/17 Change the list-building routines interface to use the expanding-string
551 triplet model, for better allocation and copying behaviour.
553 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
554 Previously it was constructed the first time a possibly-matching string
555 was met in the configuration file input during startup; now it is done
558 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
559 compatible one, to avoid the (poorly documented) possibility of a config
560 file in the working directory redirecting the DB files, possibly correpting
561 some existing file. CVE-2017-10140 assigned for BDB.
563 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
564 cache-hot. Previously, although the result was properly cached, the
565 initial verify call returned a defer.
567 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
568 the main verify for receipient in uncached-mode.
570 JH/22 Retire historical build files to an "unsupported" subdir. These are
571 defined as "ones for which we have no current evidence of testing".
573 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
574 if present. Previously it was ignored.
576 JH/24 Start using specified-initialisers in C structure init coding. This is
577 a C99 feature (it's 2017, so now considered safe).
579 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
580 if was a fixed-sized field and bitmask ops via macros; it is now more
583 PP/07 GitHub PR 56: Apply MariaDB build fix.
584 Patch provided by Jaroslav Škarvada.
586 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
587 during Coverity cleanups [4.87 JH/47]
588 Diagnosis and fix provided by Michael Fischer v. Mollard.
590 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
591 the right size to place the terminating semicolon on its own folded
592 line, the header hash was calculated to an incorrect value thanks to
593 the (relaxed) space the fold became.
595 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
596 and confused the parent.
598 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
599 which could crash as a result. This could lead to undeliverable messages.
601 JH/28 Logging: "next input sent too soon" now shows where input was truncated
604 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
605 matters on fast-turnover and PID-randomising systems, which were getting
606 out-of-order delivery.
608 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
609 a possibly-overlapping copy. The symptom was that "Remote host closed
610 connection in response to HELO" was logged instead of the actual 4xx
613 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
614 Previously only that bufferd was discarded, resulting in SYMTP command
617 JH/32 DKIM: when a message has multiple signatures matching an identity given
618 in dkim_verify_signers, run the dkim acl once for each. Previously only
619 one run was done. Bug 2189.
621 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
622 "panic the current process" to "deliberately defer". The panic log is
623 still written with the problem list name; the mail and reject logs now
624 get a temp-reject line for the message that was being handled, saying
625 something like "domains check lookup or other defer". The SMTP 451
626 message is still "Temporary local problem".
628 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
629 A crafted sequence of BDAT commands could result in in-use memory beeing
630 freed. CVE-2017-16943.
632 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
633 from SMTP input. Previously it was always done; now only done for DATA
634 and not BDAT commands. CVE-2017-16944.
636 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
637 to the message (such as an overlong header line). Previously this was
638 not done and we did not exit BDAT mode. Followon from the previous item
639 though a different problem.
645 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
646 than -2003 did; needs libidn2 in addition to libidn.
648 JH/02 The path option on a pipe transport is now expanded before use.
650 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
651 Patch provided by "Björn", documentation fix added too.
653 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
654 missing a wire-to-host endian conversion.
656 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
657 close after a BDAT command line could be taken as a following command,
658 giving a synch failure. Fix by only checking for synch immediately
659 before acknowledging the chunk.
661 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
662 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
663 macro. Patches provided by Josh Soref.
665 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
666 Previously we did not; the RFC seems ambiguous and VRFY is not listed
667 by IANA as a service extension. However, John Klensin suggests that we
670 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
671 the dkim code may be unix-mode line endings rather than smtp wire-format
672 CRLF, so prepend a CR to any bare LF.
674 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
675 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
677 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
678 the first were themselves being wrongly included in the feed into dkim
679 processing; with most chunk sizes in use this resulted in an incorrect
680 body hash calculated value.
682 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
683 DKIM signature block, for verification. Although advised against by
684 standards it is specifically not ruled illegal.
686 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
688 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
689 missing a body hash (the bh= tag).
691 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
692 It seems that HAProxy sends the Proxy Protocol information in clear and
693 only then does a TLS startup, so do the same.
695 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
696 TCP connections (such as for Spamd) unless the daemon successfully set
697 Fast Open mode on its listening sockets. This fixes breakage seen on
698 too-old kernels or those not configured for Fast Open, at the cost of
699 requiring both directions being enabled for TFO, and TFO never being used
700 by non-daemon-related Exim processes.
702 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
703 endings, at least on the first header line. Try to canonify any that get
704 past that check, despite the cost.
706 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
707 now limited to an arbitrary five deep, while parsing addresses with the
708 strip_excess_angle_brackets option enabled.
710 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
711 instead leave the unprompted TLS handshake in socket buffer for the
712 TLS library to consume.
714 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
716 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
718 JH/16 Drop variables when they go out of scope. Memory management drops a whole
719 region in one operation, for speed, and this leaves assigned pointers
720 dangling. Add checks run only under the testsuite which checks all
721 variables at a store-reset and panics on a dangling pointer; add code
722 explicitly nulling out all the variables discovered. Fixes one known
723 bug: a transport crash, where a dangling pointer for $sending_ip_address
724 originally assigned in a verify callout, is re-used.
726 PP/06 Drop '.' from @INC in various Perl scripts.
728 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
730 PP/08 Reduce a number of compilation warnings under clang; building with
731 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
732 should be warning-free.
734 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
736 HS/01 Fix portability problems introduced by PP/08 for platforms where
737 realloc(NULL) is not equivalent to malloc() [SunOS et al].
739 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
740 chunk. This allows us to accept broken chunked messages. We need a more
741 general solution here.
743 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
744 already-broken messages in the queue.
746 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
748 JH/19 Fix reference counting bug in routing-generated-address tracking.
754 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
755 supports it and a size is available (ie. the sending peer gave us one).
757 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
758 years of being deprecated). The replacements are the ACLs
759 acl_smtp_mime and acl_not_smtp_mime.
761 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
762 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
763 or even an in-clear connection were permitted. Now, if the host lookup
764 was dnssec and dane was requested then the host is only used if the
765 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
766 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
767 if one fails this test.
768 This means that a poorly-configured remote DNS will make it incommunicado;
769 but it protects against a DNS-interception attack on it.
771 JH/04 Bug 1810: make continued-use of an open smtp transport connection
772 non-noisy when a race steals the message being considered.
774 JH/05 If main configuration option tls_certificate is unset, generate a
775 self-signed certificate for inbound TLS connections.
777 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
778 in rewrites and routers.
780 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
781 and logged a warning sing 4.83; now they are a configuration file error.
783 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
784 (lacking @domain). Apply the same qualification processing as RCPT.
786 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
788 JH/10 Support ${sha256:} applied to a string (as well as the previous
791 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
792 a cutthrough deliver is pending, as we always want to make a connection.
793 This also avoids re-routing the message when later placing the cutthrough
794 connection after a verify cache hit.
795 Do not update it with the verify result either.
797 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
798 when routing results in more than one destination address.
800 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
801 signing (which inhibits the cutthrough capability). Previously only
802 the presence of an option was tested; now an expansion evaluating as
803 empty is permissible (obviously it should depend only on data available
804 when the cutthrough connection is made).
806 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
807 the relevant preceding SMTP command did not note the pipelining mode.
809 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
810 Previously they were not counted.
812 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
813 as one having no matching records. Previously we deferred the message
814 that needed the lookup.
816 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
817 distinguished as "(=".
819 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
820 for missing MX records. Previously it only worked for missing A records.
822 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
824 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
825 after the data-go-ahead and data-ack. Patch from Jason Betts.
827 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
828 even for a "none" policy. Patch from Tony Meyer.
830 JH/22 Fix continued use of a connection for further deliveries. If a port was
831 specified by a router, it must also match for the delivery to be
834 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
835 When one of the recipients of a message was unsuitable for the connection
836 (has no matching addresses), we lost track of needing to mark it
837 deferred. As a result mail would be lost.
839 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
841 JH/25 Decoding ACL controls is now done using a binary search; the source code
842 takes up less space and should be simpler to maintain. Merge the ACL
843 condition decode tables also, with similar effect.
845 JH/26 Fix problem with one_time used on a redirect router which returned the
846 parent address unchanged. A retry would see the parent address marked as
847 delivered, so not attempt the (identical) child. As a result mail would
850 JH/27 Fix a possible security hole, wherein a process operating with the Exim
851 UID can gain a root shell. Credit to http://www.halfdog.net/ for
852 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
855 JH/28 Enable {spool,log} filesystem space and inode checks as default.
856 Main config options check_{log,spool}_{inodes,space} are now
857 100 inodes, 10MB unless set otherwise in the configuration.
859 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
860 Previously it only applied to the main-section connection policy
863 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
865 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
866 by me. Added RFC7919 DH primes as an alternative.
868 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
869 are not in the system include path.
871 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
872 GnuTLS, when a session startup failed (eg because the client disconnected)
873 Exim did stdio operations after fclose. This was exposed by a recent
874 change which nulled out the file handle after the fclose.
876 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
877 signed directly by the cert-signing cert, rather than an intermediate
878 OCSP-signing cert. This is the model used by LetsEncrypt.
880 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
882 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
883 an incoming connection.
885 HS/02 Bug 1802: Do not half-close the connection after sending a request
888 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
889 fallback to "prime256v1".
891 JH/34 SECURITY: Use proper copy of DATA command in error message.
892 Could leak key material. Remotely exploitable. CVE-2016-9963.
898 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
899 and 3.4.4 - once the server is enabled to respond to an OCSP request
900 it does even when not requested, resulting in a stapling non-aware
901 client dropping the TLS connection.
903 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
904 support variable-length bit vectors. No functional change.
906 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
907 The I= interface field on outgoing lines is now after the H= remote
908 host field, same as incoming lines. There is a separate
909 outgoing_interface log selector which allows you to disable the
912 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
913 If not running log_selector +smtp_connection the mainlog would be held
914 open indefinitely after a "too many connections" event, including to a
915 deleted file after a log rotate. Leave the per net connection logging
916 leaving it open for efficiency as that will be quickly detected by the
917 check on the next write.
919 HS/01 Bug 1671: Fix post transport crash.
920 Processing the wait-<transport> messages could crash the delivery
921 process if the message IDs didn't exist for some reason. When
922 using 'split_spool_directory=yes' the construction of the spool
923 file name failed already, exposing the same netto behaviour.
925 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
926 mime_regex ACL conditions.
928 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
929 to DSN fail messages (bounces): remote IP, remote greeting, remote response
930 to HELO, local diagnostic string.
932 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
933 log line to debug. Even when configured with a tls authenticator many
934 client connections are expected to not authenticate in this way, so
935 an authenticate fail is not an error.
937 HS/02 Add the Exim version string to the process info. This way exiwhat
938 gives some more detail about the running daemon.
940 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
941 matter for fast-change records such as DNSBLs.
943 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
944 retry record, even if constant. There may be multiple transports with
945 different interface settings and the retry behaviour needs to be kept
948 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
950 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
952 JH/10 Bug 840: fix log_defer_output option of pipe transport
954 JH/11 Bug 830: use same host for all RCPTS of a message, even under
955 hosts_randomize. This matters a lot when combined with mua_wrapper.
957 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
958 ${quote_pgsql:<string>} operator.
960 JH/13 Bug 1708: avoid misaligned access in cached lookup.
962 JH/14 Change header file name for freeradius-client. Relevant if compiling
963 with Radius support; from the Gentoo tree and checked under Fedora.
965 JH/15 Bug 1712: Introduce $prdr_requested flag variable
967 JH/16 Bug 1714: Permit an empty string as expansion result for transport
968 option transport_filter, meaning no filtering.
970 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
972 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
973 defaults to "*" (all hosts). The variable is now available when not built
974 with TLS, default unset, mainly to enable keeping the testsuite sane.
975 If a server certificate is not supplied (via tls_certificate) an error is
976 logged, and clients will find TLS connections fail on startup. Presumably
977 they will retry in-clear.
978 Packagers of Exim are strongly encouraged to create a server certificate
979 at installation time.
981 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
982 with the $config_file variable.
984 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
985 in transport context, after the attempt, and per-recipient. The latter type
986 is per host attempted. The event data is the error message, and the errno
987 information encodes the lookup type (A vs. MX) used for the (first) host,
988 and the trailing two digits of the smtp 4xx response.
990 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
991 to write to mainlog (or rejectlog, paniclog) in the window between file
992 creation and permissions/ownership being changed. Particularly affects
993 installations where exicyclog is run as root, rather than exim user;
994 result is that the running daemon panics and dies.
996 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
998 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
999 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1000 "pri" and "weight". Note that the previous implicit priority given by the
1001 list order is no longer honoured.
1003 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1004 for DKIM processing.
1006 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1007 by defining SUPPORT_SOCKS.
1009 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1010 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1011 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1012 variables are renamed to proxy_{local,external}_{address,port}.
1014 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1015 for a build by defining SUPPORT_I18N
1017 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1018 of the query string, and make ${quote_redis:} do that quoting.
1020 JH/29 Move Events support from Experimental to mainline, enabled by default
1021 and removable for a build by defining DISABLE_EVENT.
1023 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1025 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1026 cached by the daemon.
1028 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1029 by defining LOOKUP_REDIS. The libhiredis library is required.
1031 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1032 keys are given for lookup.
1034 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1035 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1036 only supported when built with TLS support. The PolarSSL SHA routines
1037 are still used when the TLS library is too old for convenient support.
1039 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1040 openssl_options), for security. OpenSSL forces this from version 1.1.0
1041 server-side so match that on older versions.
1043 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1044 allocation for $value could be released as the expansion processing
1045 concluded, but leaving the global pointer active for it.
1047 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1048 and to use the domains and local_parts ACL conditions.
1050 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1051 incorrectly not doubled on cutthrough transmission, hence seen as a
1052 body-termination at the receiving system - resulting in truncated mails.
1053 Commonly the sender saw a TCP-level error, and retransmitted the message
1054 via the normal store-and-forward channel. This could result in duplicates
1055 received - but deduplicating mailstores were liable to retain only the
1056 initial truncated version.
1058 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1060 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1062 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1063 we're in there, support oversigning also; bug 1309.
1065 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1067 HS/04 Add support for keep_environment and add_environment options.
1069 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1070 either intentional arithmetic overflow during PRNG, or testing config-
1073 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1074 delivery resulted in actual delivery. Cancel cutthrough before DATA
1077 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1078 rejects a recipient: pass the reject to the originator.
1080 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1081 Many were false-positives and ignorable, but it's worth fixing the
1084 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1085 for the new environment-manipulation done at startup. Move the routines
1086 from being local to tls.c to being global via the os.c file.
1088 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1089 an extract embedded as result-arg for a map, the first arg for extract
1090 is unavailable so we cannot tell if this is a numbered or keyed
1091 extraction. Accept either.
1097 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1100 JH/02 The smtp transport option "multi_domain" is now expanded.
1102 JH/03 The smtp transport now requests PRDR by default, if the server offers
1105 JH/04 Certificate name checking on server certificates, when exim is a client,
1106 is now done by default. The transport option tls_verify_cert_hostnames
1107 can be used to disable this per-host. The build option
1108 EXPERIMENTAL_CERTNAMES is withdrawn.
1110 JH/05 The value of the tls_verify_certificates smtp transport and main options
1111 default to the word "system" to access the system default CA bundle.
1112 For GnuTLS, only version 3.0.20 or later.
1114 JH/06 Verification of the server certificate for a TLS connection is now tried
1115 (but not required) by default. The verification status is now logged by
1116 default, for both outbound TLS and client-certificate supplying inbound
1119 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1122 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1123 Status Notification (bounce) messages are now MIME format per RFC 3464.
1124 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1125 under the control of the dsn_advertise_hosts option, and routers may
1126 have a dsn_lasthop option.
1128 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1129 default, modifiable by a malware= option. The list separator for
1130 the options can now be changed in the usual way. Bug 68.
1132 JH/10 The smtp_receive_timeout main option is now expanded before use.
1134 JH/11 The incoming_interface log option now also enables logging of the
1135 local interface on delivery outgoing connections.
1137 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1138 if the interface and destination host and port all match.
1140 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1143 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1144 Patch from Andrew Lewis.
1146 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1147 now supports optional time-restrictions, weighting, and priority
1148 modifiers per server. Patch originally by <rommer@active.by>.
1150 JH/16 The spamd_address main option now supports a mixed list of local
1151 and remote servers. Remote servers can be IPv6 addresses, and
1152 specify a port-range.
1154 JH/17 Bug 68: The spamd_address main option now supports an optional
1155 timeout value per server.
1157 JH/18 Bug 1581: Router and transport options headers_add/remove can
1158 now have the list separator specified.
1160 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1163 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1166 JH/21 Support for the A6 type of dns record is withdrawn.
1168 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1169 rather than the verbs used.
1171 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1172 from 255 to 1024 chars.
1174 JH/24 Verification callouts now attempt to use TLS by default.
1176 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1177 are generic router options now. The defaults didn't change.
1179 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1180 Original patch from Alexander Shikoff, worked over by JH.
1182 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1183 files if ZCAT_COMMAND is not executable.
1185 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1187 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1189 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1190 Normally benign, it bites when the pair was led to by a CNAME;
1191 modern usage is to not canonicalize the domain to a CNAME target
1192 (and we were inconsistent anyway for A-only vs AAAA+A).
1194 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1196 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1197 when evaluating $sender_host_dnssec.
1199 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1200 $sender_helo_dnssec variable.
1202 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1204 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1206 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1208 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1209 documented as working, but never had. Support all but $spam_report.
1211 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1212 added for tls authenticator.
1214 HS/03 Add perl_taintmode main config option
1220 TL/01 When running the test suite, the README says that variables such as
1221 no_msglog_check are global and can be placed anywhere in a specific
1222 test's script, however it was observed that placement needed to be near
1223 the beginning for it to behave that way. Changed the runtest perl
1224 script to read through the entire script once to detect and set these
1225 variables, reset to the beginning of the script, and then run through
1226 the script parsing/test process like normal.
1228 TL/02 The BSD's have an arc4random API. One of the functions to induce
1229 adding randomness was arc4random_stir(), but it has been removed in
1230 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1231 function when detected.
1233 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1234 cause callback expansion.
1236 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1237 syntax errors in an expansion can be treated as a string instead of
1238 logging or causing an error, due to the internal use of bool_lax
1239 instead of bool when processing it.
1241 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1242 server certificates when making smtp deliveries.
1244 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1246 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1248 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1250 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1251 Merged patch from Sebastian Wiedenroth.
1253 JH/05 Fix results-pipe from transport process. Several recipients, combined
1254 with certificate use, exposed issues where response data items split
1255 over buffer boundaries were not parsed properly. This eventually
1256 resulted in duplicates being sent. This issue only became common enough
1257 to notice due to the introduction of connection certificate information,
1258 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1260 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1261 size buffer was used, resulting in syntax errors when an expansion
1264 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1265 version 3.3.6 or later.
1267 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1268 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1269 both become "event_action", the variables become $event_name, $event_data
1270 and $event_defer_errno. There is a new variable $verify_mode, usable in
1271 routers, transports and related events. The tls:cert event is now also
1272 raised for inbound connections, if the main configuration event_action
1275 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1276 early OCSP support, but no stapling (appears to be less than 1.0.0).
1278 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1279 server certificate names available under the smtp transport option
1280 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1283 JH/10 Time-related extraction expansions from certificates now use the main
1284 option "timezone" setting for output formatting, and are consistent
1285 between OpenSSL and GnuTLS compilations. Bug 1541.
1287 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1288 encoded parameter in the incoming message. Bug 1558.
1290 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1291 include certificate info, eximon was claiming there were spoolfile
1294 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1296 JH/14 Log delivery-related information more consistently, using the sequence
1297 "H=<name> [<ip>]" wherever possible.
1299 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1300 are problematic for Debian distribution, omit them from the release
1303 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1305 JH/16 Fix string representation of time values on 64bit time_t architectures.
1308 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1309 output list separator was used.
1314 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1315 checkers that were complaining about end of non-void function with no
1318 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1319 This was a regression introduced in 4.83 by another bugfix.
1321 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1323 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1324 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1330 TF/01 Correctly close the server side of TLS when forking for delivery.
1332 When a message was received over SMTP with TLS, Exim failed to clear up
1333 the incoming connection properly after forking off the child process to
1334 deliver the message. In some situations the subsequent outgoing
1335 delivery connection happened to have the same fd number as the incoming
1336 connection previously had. Exim would try to use TLS and fail, logging
1337 a "Bad file descriptor" error.
1339 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1340 utilities have not been installed.
1342 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1343 temporary space as the ACL may create new global variables.
1345 TL/01 LDAP support uses per connection or global context settings, depending
1346 upon the detected version of the libraries at build time.
1348 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1349 to extract and use the src ip:port in logging and expansions as if it
1350 were a direct connection from the outside internet. PPv2 support was
1351 updated based on HAProxy spec change in May 2014.
1353 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1355 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1356 Properly escape header and check for NULL return.
1358 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1361 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1363 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1364 characters in header names, implemented as a verify condition.
1365 Contributed by Michael Fischer v. Mollard.
1367 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1368 results permerror and temperror. Previous values are deprecated but
1369 still accepted. In a future release, err_perm and err_temp will be
1370 completely removed, which will be a backward incompatibility if the
1371 ACL tests for either of these two old results. Patch contributed by
1372 user bes-internal on the mailing list.
1374 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1376 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1377 selectors, in both main and reject logs.
1379 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1382 JH/07 Add malware type "sock" for talking to simple daemon.
1384 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1386 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1387 routers/transports under cutthrough routing.
1389 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1390 numbers. Touch up "bool" conditional to keep the same definition.
1392 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1394 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1396 JH/12 Expand items in router/transport headers_add or headers_remove lists
1397 individually rather than the list as a whole. Bug 1452.
1399 Required for reasonable handling of multiple headers_ options when
1400 they may be empty; requires that headers_remove items with embedded
1401 colons must have them doubled (or the list-separator changed).
1403 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1404 view the policy declared in the DMARC record. Currently, $dmarc_status
1405 is a combined value of both the record presence and the result of the
1408 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1410 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1411 dnslookup router and the smtp transport (applying to the forward
1414 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1415 of ldap servers used for a specific lookup. Patch provided by Heiko
1418 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1419 New variable $lookup_dnssec_authenticated for observability.
1421 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1422 Patch submitted by Lars Timman.
1424 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1426 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1427 Requires trusted mode and valid format message id, aborts otherwise.
1428 Patch contributed by Heiko Schlichting.
1430 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1431 certextract with support for various fields. Bug 1358.
1433 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1434 is requested by default, modifiable by smtp transport option
1437 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1438 operate on certificate variables to give certificate fingerprints
1439 Also new ${sha256:cert_variable}.
1441 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1443 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1444 Christian Aistleitner.
1446 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1448 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1449 file. Patch from Wolfgang Breyha.
1451 JH/25 Expand the coverage of the delivery $host and $host_address to
1452 client authenticators run in verify callout. Bug 1476.
1454 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1455 align with daemon_smtp_ports. Bug 72.
1457 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1458 support and error reporting did not work properly.
1460 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1461 and is readable. Patch from Andrew Colin Kissa.
1463 TL/14 Enhance documentation of ${run expansion and how it parses the
1464 commandline after expansion, particularly in the case when an
1465 unquoted variable expansion results in an empty value.
1467 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1469 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1470 by renaming away from it. Addresses GH issue 15, reported by
1473 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1475 TL/15 SECURITY: prevent double expansion in math comparison functions
1476 (can expand unsanitized data). Not remotely exploitable.
1483 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1485 PP/02 Make -n do something, by making it not do something.
1486 When combined with -bP, the name of an option is not output.
1488 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1491 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1492 $sender_host_name and config options to manage this, and basic check
1495 PP/05 DSCP support for outbound connections and control modifier for inbound.
1497 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1498 (Only plugin which currently uses this is kerberos4, which nobody should
1499 be using, but we should make it available and other future plugins might
1500 conceivably use it, even though it would break NAT; stuff *should* be
1501 using channel bindings instead).
1503 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1504 name; added for Sendmail compatibility; requires admin caller.
1505 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1506 just ignore it); requires trusted caller.
1507 Also parse but ignore: -Ac -Am -X<logfile>
1510 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1512 TL/02 Add +smtp_confirmation as a default logging option.
1514 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1515 Patch by Magnus Holmgren from 2007-02-20.
1517 TL/04 Bugzilla 1281 - Spec typo.
1518 Bugzilla 1283 - Spec typo.
1519 Bugzilla 1290 - Spec grammar fixes.
1521 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1523 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1525 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1526 mailing list by by Dmitry Isaikin.
1528 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1530 JH/02 Support "G" suffix to numbers in ${if comparisons.
1532 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1534 NM/01 Bugzilla 1197 - Spec typo
1535 Bugzilla 1196 - Spec examples corrections
1537 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1539 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1540 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1543 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1544 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1546 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1547 "acl {{name}{arg}...}", and optional args on acl condition
1550 JH/05 Permit multiple router/transport headers_add/remove lines.
1552 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1554 JH/07 Avoid using a waiting database for a single-message-only transport.
1555 Performance patch from Paul Fisher. Bugzilla 1262.
1557 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1560 JH/09 Add $headers_added variable, with content from use of ACL modifier
1561 add_header (but not yet added to the message). Bugzilla 199.
1563 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1564 Pulled from Bugzilla 817 by Wolfgang Breyha.
1566 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1568 (nb: this is the same fix as in Exim 4.80.1)
1570 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1573 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1576 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1578 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1579 advertise SMTP AUTH mechanism to us, instead of a generic
1580 protocol violation error. Also, make Exim more robust to bad
1581 data from the Dovecot auth socket.
1583 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1585 When a queue runner is handling a message, Exim first routes the
1586 recipient addresses, during which it prunes them based on the retry
1587 hints database. After that it attempts to deliver the message to
1588 any remaining recipients. It then updates the hints database using
1591 So if a recipient address works intermittently, it can get repeatedly
1592 deferred at routing time. The retry hints record remains fresh so the
1593 address never reaches the final cutoff time.
1595 This is a fairly common occurrence when a user is bumping up against
1596 their storage quota. Exim had some logic in its local delivery code
1597 to deal with this. However it did not apply to per-recipient defers
1598 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1600 This change adds a proper retry rule check during routing so that the
1601 final cutoff time is checked against the message's age. We only do
1602 this check if there is an address retry record and there is not a
1603 domain retry record; this implies that previous attempts to handle
1604 the address had the retry_use_local_parts option turned on. We use
1605 this as an approximation for the destination being like a local
1606 delivery, as in LMTP.
1608 I suspect this new check makes the old local delivery cutoff check
1609 redundant, but I have not verified this so I left the code in place.
1611 TF/02 Correct gecos expansion when From: is a prefix of the username.
1613 Test 0254 submits a message to Exim with the header
1617 When I ran the test suite under the user fanf2, Exim expanded
1618 the header to contain my full name, whereas it should have added
1619 a Resent-Sender: header. It erroneously treats any prefix of the
1620 username as equal to the username.
1622 This change corrects that bug.
1624 GF/01 DCC debug and logging tidyup
1625 Error conditions log to paniclog rather than rejectlog.
1626 Debug lines prefixed by "DCC: " to remove any ambiguity.
1628 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1630 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1631 Bug spotted by Jeremy Harris; was flawed since initial commit.
1632 Would have resulted in OCSP responses post-SNI triggering an Exim
1633 NULL dereference and crash.
1635 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1637 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1638 Bug detection, analysis and fix by Samuel Thibault.
1639 Bugzilla 1331, Debian bug #698092.
1641 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1643 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1644 Server implementation by Todd Lyons, client by JH.
1645 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1646 config variable "prdr_enable" controls whether the server
1647 advertises the facility. If the client requests PRDR a new
1648 acl_data_smtp_prdr ACL is called once for each recipient, after
1649 the body content is received and before the acl_smtp_data ACL.
1650 The client is controlled by both of: a hosts_try_prdr option
1651 on the smtp transport, and the server advertisement.
1652 Default client logging of deliveries and rejections involving
1653 PRDR are flagged with the string "PRDR".
1655 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1656 fclose(). Diagnosis by Todd Lyons.
1658 PP/17 Update configure.default to handle IPv6 localhost better.
1659 Patch by Alain Williams (plus minor tweaks).
1662 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1663 This is now consistent with GnuTLS, and is now documented: the
1664 previous undocumented portable approach to treating the option as
1665 unset was to force an expansion failure. That still works, and
1666 an empty string is now equivalent.
1668 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1669 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1670 not performing validation itself.
1672 PP/20 Added force_command boolean option to pipe transport.
1673 Patch from Nick Koston, of cPanel Inc.
1675 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1678 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1680 PP/21 Fix eximon continuous updating with timestamped log-files.
1681 Broken in a format-string cleanup in 4.80, missed when I repaired the
1682 other false fix of the same issue.
1683 Report and fix from Heiko Schlichting.
1686 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1687 Report from Prashanth Katuri.
1689 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1690 It's SecureTransport, so affects any MacOS clients which use the
1691 system-integrated TLS libraries, including email clients.
1693 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1694 using a MIME ACL for non-SMTP local injection.
1695 Report and assistance in diagnosis by Warren Baker.
1697 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1699 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1701 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1702 last id that failed so it may be referenced in subsequent ACL's.
1704 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1707 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1708 ldap library initialization, allowing self-signed CA's to be
1709 used. Also properly sets require_cert option later in code by
1710 using NULL (global ldap config) instead of ldap handle (per
1711 session). Bug diagnosis and testing by alxgomz.
1713 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1714 the src/util/ subdirectory.
1716 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1717 renamed to Transport Post Delivery Action by Jeremy Harris, as
1720 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1721 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1722 redis_servers = needs to be configured which will be used by the redis
1723 lookup. Patch from Warren Baker, of The Packet Hub.
1725 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1727 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1728 hostname or reverse DNS when processing a host list. Used suggestions
1729 from multiple comments on this bug.
1731 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1733 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1734 Missed a few lines, added it to make the runtest require no keyboard
1737 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1738 contains upper case chars. Make router use caseful_local_part.
1740 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1741 support when GnuTLS has been built with p11-kit.
1747 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1749 This, or similar/improved, will also be change PP/11 of 4.82.
1755 PP/01 Handle short writes when writing local log-files.
1756 In practice, only affects FreeBSD (8 onwards).
1757 Bugzilla 1053, with thanks to Dmitry Isaikin.
1759 NM/01 Bugzilla 949 - Documentation tweak
1761 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1764 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1766 PP/02 Implemented gsasl authenticator.
1768 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1770 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1771 `pkg-config foo` for cflags/libs.
1773 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1774 with rest of GSASL and with heimdal_gssapi.
1776 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1777 `pkg-config foo` for cflags/libs for the TLS implementation.
1779 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1780 properties get this fed in as external SSF. A number of robustness
1781 and debugging improvements to the cyrus_sasl authenticator.
1783 PP/08 cyrus_sasl server now expands the server_realm option.
1785 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1786 Patch by Jeremy Harris.
1788 PP/10 Added dbmjz lookup type.
1790 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1792 PP/12 MAIL args handles TAB as well as SP, for better interop with
1793 non-compliant senders.
1794 Analysis and variant patch by Todd Lyons.
1796 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1797 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1798 Patch from Dirk Mueller <dmueller@suse.com>
1800 PP/13 tls_peerdn now print-escaped for spool files.
1801 Observed some $tls_peerdn in wild which contained \n, which resulted
1802 in spool file corruption.
1804 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1805 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1806 or write after TLS renegotiation, which otherwise led to messages
1809 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1810 as a tracking header (ie: a signed header comes before the signature).
1811 Patch from Wolfgang Breyha.
1813 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1814 comma-sep list; embedded commas doubled.
1816 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1818 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1820 Report and patch from Dmitry Banschikov.
1822 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1823 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1824 failures. We appear to now support TLS1.1+ with Exim.
1826 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1827 lets Exim select keys and certificates based upon TLS SNI from client.
1828 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1829 before an outbound SMTP session. New log_selector, +tls_sni.
1831 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1832 NULL dereference. Report and patch from Alun Jones.
1834 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1835 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1836 Not seeing resolver debug output on NetBSD, but suspect this is a
1837 resolver implementation change.
1839 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1840 Left warnings. Added "eximon gdb" invocation mode.
1842 PP/21 Defaulting "accept_8bitmime" to true, not false.
1844 PP/22 Added -bw for inetd wait mode support.
1846 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1847 locate the relevant includes and libraries. Made this the default.
1849 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1850 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1852 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1853 This may cause build issues on older platforms.
1855 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1856 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1857 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1858 Added SNI support via GnuTLS too.
1859 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1861 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1863 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1864 Applied second patch from Janne, implementing suggestion to default
1865 multiple-strings-in-record handling to match SPF spec.
1867 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1869 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1870 read-only, out of scope).
1871 Patch from Wolfgang Breyha, report from Stuart Northfield.
1873 PP/29 Fix three issues highlighted by clang analyser static analysis.
1874 Only crash-plausible issue would require the Cambridge-specific
1875 iplookup router and a misconfiguration.
1876 Report from Marcin Mirosław.
1878 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1880 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1881 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1882 As part of this, removing so much warning spew let me fix some minor
1883 real issues in debug logging.
1885 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1886 assignment on my part. Fixed.
1888 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1889 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1890 Janne Snabb (who went above and beyond: thank you).
1892 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1893 string otherwise requires a connection and a bunch more work and it's
1894 relatively easy to get wrong. Should also expose TLS library linkage
1897 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1898 64-bit ${eval} (JH/03).
1900 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1901 GNU libc to support some of the 64-bit stuff, should not lead to
1902 conflicts. Defined before os.h is pulled in, so if a given platform
1903 needs to override this, it can.
1905 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1906 protection layer was required, which is not implemented.
1907 Bugzilla 1254, patch from Wolfgang Breyha.
1909 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1910 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1911 tls_dhparam take prime identifiers. Also unbreak combination of
1912 OpenSSL+DH_params+TLSSNI.
1914 PP/39 Disable SSLv2 by default in OpenSSL support.
1920 PP/01 Solaris build fix for Oracle's LDAP libraries.
1921 Bugzilla 1109, patch from Stephen Usher.
1923 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1925 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1928 TF/02 Fix a couple more cases where we did not log the error message
1929 when unlink() failed. See also change 4.74-TF/03.
1931 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1932 lock up or crash if it happened to be inside a call to libc when it
1933 got a SIGUSR1 from exiwhat.
1935 The SIGUSR1 handler appends the current process status to the process
1936 log which is later printed by exiwhat. It used to use the general
1937 purpose logging code to do this, but several functions it calls are
1938 not safe for signals.
1940 The new output code in the SIGUSR1 handler is specific to the process
1941 log, and simple enough that it's easy to inspect for signal safety.
1942 Removing some special cases also simplifies the general logging code.
1943 Removing the spurious timestamps from the process log simplifies
1946 TF/04 Improved ratelimit ACL condition.
1948 The /noupdate option has been deprecated in favour of /readonly which
1949 has clearer semantics. The /leaky, /strict, and /readonly update modes
1950 are mutually exclusive. The update mode is no longer included in the
1951 database key; it just determines when the database is updated. (This
1952 means that when you upgrade Exim will forget old rate measurements.)
1954 Exim now checks that the per_* options are used with an update mode that
1955 makes sense for the current ACL. For example, when Exim is processing a
1956 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1957 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1958 must specify per_mail/readonly. If you omit the update mode it defaults to
1959 /leaky where that makes sense (as before) or /readonly where required.
1961 The /noupdate option is now undocumented but still supported for
1962 backwards compatibility. It is equivalent to /readonly except that in
1963 ACLs where /readonly is required you may specify /leaky/noupdate or
1964 /strict/noupdate which are treated the same as /readonly.
1966 A useful new feature is the /count= option. This is a generalization
1967 of the per_byte option, so that you can measure the throughput of other
1968 aggregate values. For example, the per_byte option is now equivalent
1969 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1971 The per_rcpt option has been generalized using the /count= mechanism
1972 (though it's more complicated than the per_byte equivalence). When it is
1973 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1974 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1975 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1976 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1977 Note that using per_rcpt with a non-readonly update mode in more than
1978 one ACL will cause the recipients to be double-counted. (The per_mail
1979 and per_byte options don't have this problem.)
1981 The handling of very low rates has changed slightly. If the computed rate
1982 is less than the event's count (usually one) then this event is the first
1983 after a long gap. In this case the rate is set to the same as this event's
1984 count, so that the first message of a spam run is counted properly.
1986 The major new feature is a mechanism for counting the rate of unique
1987 events. The new per_addr option counts the number of different
1988 recipients that someone has sent messages to in the last time period. It
1989 behaves like per_rcpt if all the recipient addresses are different, but
1990 duplicate recipient addresses do not increase the measured rate. Like
1991 the /count= option this is a general mechanism, so the per_addr option
1992 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1993 example, measure the rate that a client uses different sender addresses
1994 with the options per_mail/unique=$sender_address. There are further
1995 details in the main documentation.
1997 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1999 TF/06 Removed a few PCRE remnants.
2001 TF/07 Automatically extract Exim's version number from tags in the git
2002 repository when doing development or release builds.
2004 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2005 Bugzilla 879. Patch from Paul Fisher.
2007 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2008 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2011 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2013 PP/05 Variable $av_failed, true if the AV scanner deferred.
2014 Bugzilla 1078. Patch from John Horne.
2016 PP/06 Stop make process more reliably on build failure.
2017 Bugzilla 1087. Patch from Heiko Schlittermann.
2019 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2020 Bugzilla 1089. Patch from Heiko Schlittermann.
2022 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2023 Bugzilla 1131. Patch from Holger Weiß.
2025 PP/09 Handle IPv6 addresses with SPF.
2026 Bugzilla 860. Patch from Wolfgang Breyha.
2028 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2030 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2033 PP/11 match_* no longer expand right-hand-side by default.
2034 New compile-time build option, EXPAND_LISTMATCH_RHS.
2035 New expansion conditions, "inlist", "inlisti".
2037 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2039 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2041 PP/14 fix log_write() format string regression from TF/03.
2042 Bugzilla 1152. Patch from Dmitry Isaikin.
2048 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2050 PP/02 Harmonised TLS library version reporting; only show if debugging.
2051 Layout now matches that introduced for other libraries in 4.74 PP/03.
2053 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2055 PP/04 New "dns_use_edns0" global option.
2057 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2060 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2061 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2063 TK/01 Updated PolarSSL code to 0.14.2.
2064 Bugzilla 1097. Patch from Andreas Metzler.
2066 PP/07 Catch divide-by-zero in ${eval:...}.
2067 Fixes bugzilla 1102.
2069 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2072 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2073 format-string attack -- SECURITY: remote arbitrary code execution.
2075 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2076 time unintentionally subject to list matching rules, letting the header
2077 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2078 arbitrary string expansion). This allowed for information disclosure.
2080 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2081 INT_MIN/-1 -- value coerced to INT_MAX.
2087 NM/01 Workaround for PCRE version dependency in version reporting
2090 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2091 This fixes portability to compilers other than gcc, notably
2092 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2094 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2095 makefiles for portability to HP-UX and POSIX correctness.
2097 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2098 Also via indented variable definition in the Makefile.
2099 (Debugging by Oliver Heesakkers).
2101 PP/02 Restore caching of spamd results with expanded spamd_address.
2102 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2104 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2105 Improves build reliability. Fix from: Frank Elsner
2107 NM/02 Fix wide character breakage in the rfc2047 coding
2108 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2110 NM/03 Allow underscore in dnslist lookups
2111 Fixes bug 1026. Patch from Graeme Fowler
2113 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2114 Code patches from Adam Ciarcinski of NetBSD.
2116 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2119 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2120 is logged, to avoid truncation. Patch from John Horne.
2122 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2123 Patch from Jakob Hirsch.
2125 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2126 SQL string expansion failure details.
2127 Patch from Andrey Oktyabrski.
2129 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2130 Patch from Simon Arlott.
2132 PP/09 New lookups functionality failed to compile on old gcc which rejects
2133 extern declarations in function scope.
2134 Patch from Oliver Fleischmann
2136 PP/10 Use sig_atomic_t for flags set from signal handlers.
2137 Check getgroups() return and improve debugging.
2138 Fixed developed for diagnosis in bug 927 (which turned out to be
2141 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2142 Patch from Mark Zealey.
2144 PP/12 Bugzilla 1056: Improved spamd server selection.
2145 Patch from Mark Zealey.
2147 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2148 Based on patch from Heiko Schlittermann.
2150 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2151 Patch from Uwe Doering, sign-off by Michael Haardt.
2153 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2154 variable declaration deep within a block. Bug and patch from
2157 PP/15 lookups-Makefile IRIX compatibility coercion.
2159 PP/16 Make DISABLE_DKIM build knob functional.
2161 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2162 Patch by Simon Arlott
2164 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2165 variable argument macros. Our copy now differs from upstream.
2171 TF/01 Failure to get a lock on a hints database can have serious
2172 consequences so log it to the panic log.
2174 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2175 controlled using the smtp_confirmation log selector.
2177 TF/03 Include the error message when we fail to unlink a spool file.
2179 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2180 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2181 for maintaining out-of-tree patches for some time.
2183 PP/01 Bugzilla 139: Documentation and portability issues.
2184 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2185 Handle per-OS dynamic-module compilation flags.
2187 PP/02 Let /dev/null have normal permissions.
2188 The 4.73 fixes were a little too stringent and complained about the
2189 permissions on /dev/null. Exempt it from some checks.
2190 Reported by Andreas M. Kirchwitz.
2192 PP/03 Report version information for many libraries, including
2193 Exim version information for dynamically loaded libraries. Created
2194 version.h, now support a version extension string for distributors
2195 who patch heavily. Dynamic module ABI change.
2197 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2198 privilege escalation vulnerability whereby the Exim run-time user
2199 can cause root to append content of the attacker's choosing to
2202 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2205 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2206 If dropping privileges for untrusted macros, we disabled normal logging
2207 on the basis that it would fail; for the Exim run-time user, this is not
2208 the case, and it resulted in successful deliveries going unlogged.
2209 Fixed. Reported by Andreas Metzler.
2215 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2216 only prepend for the Resent-* case. Fixes regression introduced in
2217 Exim 4.70 by NM/22 for Bugzilla 607.
2219 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2220 increasing numbers of administrators be bitten by this.
2222 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2224 PP/03 Bugzilla 994: added openssl_options main configuration option.
2226 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2228 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2230 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2232 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2233 without a peer certificate, leading to a segfault because of an
2234 assumption that peers always have certificates. Be a little more
2235 paranoid. Problem reported by Martin Tscholak.
2237 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2238 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2239 NB: ClamAV planning to remove STREAM in "middle of 2010".
2240 CL also introduces -bmalware, various -d+acl logging additions and
2241 more caution in buffer sizes.
2243 PP/09 Implemented reverse_ip expansion operator.
2245 PP/10 Bugzilla 937: provide a "debug" ACL control.
2247 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2249 PP/12 Bugzilla 973: Implement --version.
2251 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2253 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2255 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2257 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2258 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2259 ignore trailing whitespace.
2261 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2263 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2266 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2267 Notification from Dr Andrew Aitchison.
2269 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2270 ExtendedDetectionInfo response format.
2271 Notification from John Horne.
2273 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2276 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2277 XSL and documented dependency on system catalogs, with examples of how
2280 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2283 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2284 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2285 configuration file which is writeable by the Exim user or group.
2287 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2288 of configuration files to cover files specified with the -C option if
2289 they are going to be used with root privileges, not just the default
2292 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2293 option (effectively making it always true).
2295 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2296 files to be used while preserving root privileges.
2298 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2299 that rogue child processes cannot use them.
2301 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2302 run-time user, instead of root.
2304 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2305 Exim run-time user without dropping privileges.
2307 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2308 result string, instead of calling string_vformat() twice with the same
2311 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2312 for other users. Others should always drop root privileges if they use
2313 -C on the command line, even for a whitelisted configure file.
2315 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2317 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2323 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2324 $data_path, and $header_path variables; fixed documentation bugs and
2327 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2328 exipick to access non-standard spools, including the "frozen" queue
2331 NM/01 Bugzilla 965: Support mysql stored procedures.
2332 Patch from Alain Williams
2334 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2336 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2337 Patch from Andreas Metzler
2339 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2340 Patch from Kirill Miazine
2342 NM/05 Bugzilla 671: Added umask to procmail example.
2344 JJ/03 installed exipick 20100323.0, fixing doc bug
2346 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2347 directory. Notification and patch from Dan Rosenberg.
2349 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2351 TK/02 Improve log output when DKIM signing operation fails.
2353 MH/01 Treat the transport option dkim_domain as a colon separated
2354 list, not as a single string, and sign the message with each element,
2355 omitting multiple occurences of the same signer.
2357 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2358 Bugzilla 985, 986. Patch by Simon Arlott
2360 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2361 Patch by Simon Arlott
2363 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2364 MBX locking. Notification from Dan Rosenberg.
2370 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2372 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2374 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2376 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2378 NM/04 Bugzilla 915: Flag broken perl installation during build.
2384 TK/01 Added patch by Johannes Berg that expands the main option
2385 "spamd_address" if it starts with a dollar sign.
2387 TK/02 Write list of recipients to X-Envelope-Sender header when building
2388 the mbox-format spool file for content scanning (suggested by Jakob
2391 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2392 (http://www.dcc-servers.net/) support via dccifd. Activated by
2393 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2395 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2396 by Mark Daniel Reidel <mr@df.eu>.
2398 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2399 When building exim an external PCRE library is now needed -
2400 PCRE is a system library on the majority of modern systems.
2401 See entry on PCRE_LIBS in EDITME file.
2403 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2404 conversation. Added nologin parameter to request.
2405 Patch contributed by Kirill Miazine.
2407 TF/01 Do not log submission mode rewrites if they do not change the address.
2409 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2411 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2412 log files in place. Contributed by Roberto Lima.
2414 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2416 TF/03 Bugzilla 615: When checking the local_parts router precondition
2417 after a local_part_suffix or local_part_prefix option, Exim now
2418 does not use the address's named list lookup cache, since this
2419 contains cached lookups for the whole local part.
2421 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2422 Robert Millan. Documentation is in experimental-spec.txt.
2424 TF/04 Bugzilla 668: Fix parallel build (make -j).
2426 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2428 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2429 Patch provided by Jan Srzednicki.
2431 TF/05 Leading white space used to be stripped from $spam_report which
2432 wrecked the formatting. Now it is preserved.
2434 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2435 that they are available at delivery time.
2437 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2439 TF/08 TLS error reporting now respects the incoming_interface and
2440 incoming_port log selectors.
2442 TF/09 Produce a more useful error message if an SMTP transport's hosts
2443 setting expands to an empty string.
2445 NM/06 Bugzilla 744: EXPN did not work under TLS.
2446 Patch provided by Phil Pennock.
2448 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2449 Patch provided by Richard Godbee.
2451 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2452 acl_smtp_notquit, added index entry.
2454 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2455 Patch provided by Eugene Bujak.
2457 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2458 accept(). Patch provided by Maxim Dounin.
2460 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2461 Patch provided by Phil Pennock.
2463 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2465 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2466 Patch provided by Brad "anomie" Jorsch.
2468 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2469 Patch provided by Dean Brooks.
2471 TK/05 Add native DKIM support (does not depend on external libraries).
2473 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2474 Patch provided by Graeme Fowler.
2476 NM/16 Bugzilla 851: Documentation example syntax fix.
2478 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2480 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2483 NM/19 Bugzilla 745: TLS version reporting.
2484 Patch provided by Phil Pennock.
2486 NM/20 Bugzilla 167: bool: condition support.
2487 Patch provided by Phil Pennock.
2489 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2490 clients. Patch provided by Phil Pennock.
2492 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2493 Patch provided by Brad "anomie" Jorsch.
2495 NM/23 Bugzilla 687: Fix misparses in eximstats.
2496 Patch provided by Heiko Schlittermann.
2498 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2499 Patch provided by Heiko Schlittermann.
2501 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2502 plus update to original patch.
2504 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2506 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2507 Patch provided by David Brownlee.
2509 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2511 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2513 NM/30 Bugzilla 888: TLS documentation bugfixes.
2515 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2517 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2518 Unlike the original bugzilla I have changed all shell scripts in src tree.
2520 NM/33 Bugzilla 898: Transport filter timeout fix.
2521 Patch by Todd Rinaldo.
2523 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2524 Patch by Serge Demonchaux.
2526 NM/35 Bugzilla 39: Base64 decode bug fixes.
2527 Patch by Jakob Hirsch.
2529 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2531 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2533 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2535 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2541 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2542 ALT-N's libdkim that I have put here:
2543 http://duncanthrax.net/exim-experimental/
2545 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2546 were called 'true' and it seems that C99 defines that as a reserved
2547 keyword to be used with 'bool' variable types. That means you could
2548 not include C99-style headers which use bools without triggering
2549 build errors in sieve.c.
2551 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2552 as mailq or other aliases. Changed the --help handling significantly
2553 to do whats expected. exim_usage() emits usage/help information.
2555 SC/01 Added the -bylocaldomain option to eximstats.
2557 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2559 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2561 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2567 PH/01 Another patch from the Sieve maintainer.
2569 PH/02 When an IPv6 address is converted to a string for single-key lookup
2570 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2571 dots are used instead of colons so that keys in lsearch files need not
2572 contain colons. This was done some time before quoting was made available
2573 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2574 (notated using the quote facility) so as to distinguish them from IPv4
2575 keys. This meant that lookups for IP addresses in host lists did not work
2576 for iplsearch lookups.
2578 This has been fixed by arranging for IPv6 addresses to be expressed with
2579 colons if the lookup type is iplsearch. This is not incompatible, because
2580 previously such lookups could never work.
2582 The situation is now rather anomalous, since one *can* have colons in
2583 ordinary lsearch keys. However, making the change in all cases is
2584 incompatible and would probably break a number of configurations.
2586 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2589 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2590 conversion specification without a maximum field width, thereby enabling
2591 a rogue spamd server to cause a buffer overflow. While nobody in their
2592 right mind would setup Exim to query an untrusted spamd server, an
2593 attacker that gains access to a server running spamd could potentially
2594 exploit this vulnerability to run arbitrary code as the Exim user.
2596 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2597 $primary_hostname instead of what libspf2 thinks the hosts name is.
2599 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2600 a directory entry by the name of the lookup key. Previously, if a
2601 symlink pointed to a non-existing file or a file in a directory that
2602 Exim lacked permissions to read, a lookup for a key matching that
2603 symlink would fail. Now it is enough that a matching directory entry
2604 exists, symlink or not. (Bugzilla 503.)
2606 PH/03 The body_linecount and body_zerocount variables are now exported in the
2609 PH/04 Added the $dnslist_matched variable.
2611 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2612 This means they are set thereafter only if the connection becomes
2615 PH/06 Added the client_condition to authenticators so that some can be skipped
2616 by clients under certain conditions.
2618 PH/07 The error message for a badly-placed control=no_multiline_responses left
2619 "_responses" off the end of the name.
2621 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2623 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2624 (without spaces) instead of just copying the configuration text.
2626 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2628 PH/11 Added $max_received_linelength.
2630 PH/12 Added +ignore_defer and +include_defer to host lists.
2632 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2633 way in which PCRE > 7.0 is built.
2635 PH/14 Implemented queue_only_load_latch.
2637 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2638 MAIL command. The effect was to mangle the value on 64-bit systems.
2640 PH/16 Another patch from the Sieve maintainer.
2642 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2644 PH/18 If a system quota error occurred while trying to create the file for
2645 a maildir delivery, the message "Mailbox is full" was not appended to the
2646 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2647 only to a quota excession during the actual writing of the file.
2649 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2650 characters?) which causes problems in log lines. The DN values are now
2651 passed through string_printing() before being added to log lines.
2653 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2654 and InterBase are left for another time.)
2656 PH/21 Added message_body_newlines option.
2658 PH/22 Guard against possible overflow in moan_check_errorcopy().
2660 PH/23 POSIX allows open() to be a macro; guard against that.
2662 PH/24 If the recipient of an error message contained an @ in the local part
2663 (suitably quoted, of course), incorrect values were put in $domain and
2664 $local_part during the evaluation of errors_copy.
2670 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2671 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2674 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2675 issue a MAIL command.
2677 PH/02 In an ACL statement such as
2679 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2681 if a client was not listed at all, or was listed with a value other than
2682 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2683 the condition was not true (as it should be), so access was not denied.
2684 The bug was that the ! inversion was incorrectly passed on to the second
2685 item. This has been fixed.
2687 PH/03 Added additional dnslists conditions == and =& which are different from
2688 = and & when the dns lookup returns more than one IP address.
2690 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2691 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2693 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2694 FSYNC, which compiles an option called disable_fsync that allows for
2695 bypassing fsync(). The documentation is heavily laced with warnings.
2697 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2699 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2700 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2701 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2702 including adding "make clean"; (3) Added -fPIC when compiling the test
2703 dynamically loaded module, to get rid of a warning.
2705 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2706 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2707 = 0s. The bug is otherwise harmless.
2709 PH/07 There was a bug in the dovecot authenticator such that the value of
2710 $auth1 could be overwritten, and so not correctly preserved, after a
2711 successful authentication. This usually meant that the value preserved by
2712 the server_setid option was incorrect.
2714 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2716 PH/09 Installed PCRE release 7.0.
2718 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2719 run for batched SMTP input. It is now run at the start of every message
2720 in the batch. While fixing this I discovered that the process information
2721 (output by running exiwhat) was not always getting set for -bs and -bS
2722 input. This is fixed, and it now also says "batched" for BSMTP.
2724 PH/11 Added control=no_pipelining.
2726 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2727 patch, slightly modified), and move the expansion of helo_data till after
2728 the connection is made in the smtp transport (so it can use these
2731 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2733 PH/14 Added log_selector = +pid.
2735 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2737 PH/16 Add ${if forany and ${if forall.
2739 PH/17 Added dsn_from option to vary the From: line in DSNs.
2741 PH/18 Flush SMTP output before performing a callout, unless control =
2742 no_callout_flush is set.
2744 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2745 was true (the default) a successful delivery failed to delete the retry
2746 item, thus causing premature timeout of the address. The bug is now
2749 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2751 PH/21 Long custom messages for fakedefer and fakereject are now split up
2752 into multiline responses in the same way that messages for "deny" and
2753 other ACL rejections are.
2755 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2756 with slight modification.
2758 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2759 draft, changing the syntax and factoring some duplicate code".
2761 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2762 for deliveries of the second and subsequent messages over the same SMTP
2765 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2766 ${reduce, with only minor "tidies".
2768 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2770 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2771 expansion side effects.
2773 PH/27 When a message times out after an over-quota error from an Exim-imposed
2774 quota, the bounce message says "mailbox is full". This message was not
2775 being given when it was a system quota that was exceeded. It now should
2778 MH/03 Made $recipients available in local_scan(). local_scan() already has
2779 better access to the recipient list through recipients_list[], but
2780 $recipients can be useful in postmaster-provided expansion strings.
2782 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2783 in the case of a MAIL command with additional options following the
2784 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2785 were accidentally chopped off.
2787 PH/29 SMTP synchronization checks are implemented when a command is read -
2788 there is a check that no more input is waiting when there shouldn't be
2789 any. However, for some commands, a delay in an ACL can mean that it is
2790 some time before the response is written. In this time, more input might
2791 arrive, invalidly. So now there are extra checks after an ACL has run for
2792 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2793 pipelining has not been advertised.
2795 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2797 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2798 returned in the SMTP response when smtp_return_error_details was false.
2799 This has been fixed.
2801 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2802 instead of the C I/O that was originally supplied, because problems were
2803 reported on Solaris.
2805 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2806 Exim which did not show up earlier: it was assuming that a call to
2807 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2808 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2809 was a macro that became an assignment, so it seemed to work. This has
2810 changed to a proper function call with a void return, hence the compile
2811 error. Exim's code has been fixed.
2813 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2816 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2818 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2820 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2821 because it turns out to be too big or there is a timeout) there is no
2822 "Completed" line in the log. When some messages of this type were
2823 selected by exigrep, they were listed as "not completed". Others were
2824 picked up by some special patterns. I have improved the selection
2825 criteria to be more general.
2827 PH/38 The host_find_failed option in the manualroute router can now be set
2828 to "ignore", to completely ignore a host whose IP address cannot be
2829 found. If all hosts are ignored, the behaviour is controlled by the new
2830 host_all_ignored option.
2832 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2833 homing or because of multiple MX records with /mx) generated more than
2834 one IP address, and the following item turned out to be the local host,
2835 all the secondary addresses of the first item were incorrectly removed
2836 from the list, along with the local host and any following hosts (which
2837 is what is supposed to happen).
2839 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2840 whoever called Exim into the -H file. In the case of the daemon it was
2841 behaving confusingly. When first started, it used values for whoever
2842 started the daemon, but after a SIGHUP it used the Exim user (because it
2843 calls itself on a restart). I have changed the code so that it now always
2846 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2847 message are rejected with the same error (e.g. no authentication or bad
2848 sender address), and a DATA command is nevertheless sent (as can happen
2849 with PIPELINING or a stupid MUA), the error message that was given to the
2850 RCPT commands is included in the rejection of the DATA command. This is
2851 intended to be helpful for MUAs that show only the final error to their
2854 PH/42 Another patch from the Sieve maintainer.
2856 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2857 Eximstats - Fixed some broken HTML links and added missing column headers
2859 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2860 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2862 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2864 PH/43 Yet another patch from the Sieve maintainer.
2866 PH/44 I found a way to check for a TCP/IP connection going away before sending
2867 the response to the final '.' that terminates a message, but only in the
2868 case where the client has not sent further data following the '.'
2869 (unfortunately, this is allowed). However, in many cases there won't be
2870 any further data because there won't be any more messages to send. A call
2871 to select() can be used: if it shows that the input is "ready", there is
2872 either input waiting, or the socket has been closed. An attempt to read
2873 the next input character can distinguish the two cases. Previously, Exim
2874 would have sent an OK response which the client would never have see.
2875 This could lead to message repetition. This fix should cure that, at
2876 least in a lot of common cases.
2878 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2879 advertised in response to EHLO.
2885 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2886 fixed by 4.65/MH/01 (is this a record?) are fixed:
2888 (i) An empty string was always treated as zero by the numeric comparison
2889 operators. This behaviour has been restored.
2891 (ii) It is documented that the numeric comparison operators always treat
2892 their arguments as decimal numbers. This was broken in that numbers
2893 starting with 0 were being interpreted as octal.
2895 While fixing these problems I realized that there was another issue that
2896 hadn't been noticed. Values of message_size_limit (both the global option
2897 and the transport option) were treated as octal if they started with 0.
2898 The documentation was vague. These values are now always treated as
2899 decimal, and I will make that clear in the documentation.
2905 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2906 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2909 MH/01 Don't check that the operands of numeric comparison operators are
2910 integers when their expansion is in "skipping" mode (fixes bug
2911 introduced by 4.64-PH/07).
2913 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2914 child addresses, Exim now panics and dies. Previously, because the count
2915 is held in a short int, deliveries were likely to be lost. As such a
2916 large number of recipients for a single message is ridiculous
2917 (performance will be very, very poor), I have chosen to impose a limit
2918 rather than extend the field.
2924 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2925 leftover -K file (the existence of which was triggered by #402).
2926 While we were at it, introduced process PID as part of the -K
2927 filename. This should rule out race conditions when creating
2930 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2931 processing considerably. Previous code took too long for large mails,
2932 triggering a timeout which in turn triggers #401.
2934 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2935 in the DK code in transports.c. sendfile() is not really portable,
2936 hence the _LINUX specificness.
2938 TF/01 In the add_headers option to the mail command in an Exim filter,
2939 there was a bug that Exim would claim a syntax error in any
2940 header after the first one which had an odd number of characters
2943 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2944 callout verification, Exim cached a "reject" for the entire domain. This
2945 is correct for most verifications, but it is not correct for a recipient
2946 verification with use_sender or use_postmaster set, because in that case
2947 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2948 case of MAIL FROM:<> rejection from other early rejections (e.g.
2949 rejection of HELO). When verifying a recipient using a non-null MAIL
2950 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2951 Whatever the result of the callout, the value of the domain cache is
2952 left unchanged (for any other kind of callout, getting as far as trying
2953 RCPT means that the domain itself is ok).
2955 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2958 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2959 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2960 session, the daemon ignores ECONNECT errors and logs others; it now
2961 ignores EPIPE as well.
2963 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2964 (quoted-printable decoding).
2966 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2967 later the small subsequent patch to fix an introduced bug.
2969 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2971 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2973 PH/08 An error is now given if message_size_limit is specified negative.
2975 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2976 to be given (somewhat) arbitrary names.
2978 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2981 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2982 miscellaneous code fixes
2984 PH/10 Added the log_reject_target ACL modifier to specify where to log
2987 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2988 hostname. This is wrong, because it relates to the incoming message (and
2989 probably the interface on which it is arriving) and not to the outgoing
2990 callout (which could be using a different interface). This has been
2991 changed to use the value of the helo_data option from the smtp transport
2992 instead - this is what is used when a message is actually being sent. If
2993 there is no remote transport (possible with a router that sets up host
2994 addresses), $smtp_active_hostname is used.
2996 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2997 tweaks were necessary in order to get it to work (see also 21 below):
2998 (a) The code assumed that strncpy() returns a negative number on buffer
2999 overflow, which isn't the case. Replaced with Exim's string_format()
3001 (b) There were several signed/unsigned issues. I just did the minimum
3002 hacking in of casts. There is scope for a larger refactoring.
3003 (c) The code used strcasecmp() which is not a standard C function.
3004 Replaced with Exim's strcmpic() function.
3005 (d) The code set only $1; it now sets $auth1 as well.
3006 (e) A simple test gave the error "authentication client didn't specify
3007 service in request". It would seem that Dovecot has changed its
3008 interface. Fortunately there's a specification; I followed it and
3009 changed what the client sends and it appears to be working now.
3011 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3014 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3015 address A is aliased to B and C, where B exists and C does not. Without
3016 -v the output is "A verified" because verification stops after a
3017 successful redirection if more than one address is generated. However,
3018 with -v the child addresses are also verified. Exim was outputting "A
3019 failed to verify" and then showing the successful verification for C,
3020 with its parentage. It now outputs "B failed to verify", showing B's
3021 parentage before showing the successful verification of C.
3023 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3024 look up a TXT record in a specific list after matching in a combined
3027 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3028 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3029 they consult the DNS. I had assumed they would set it the way they
3030 wanted; and indeed my experiments on Linux seem to show that in some
3031 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3032 To be on the safe side, however, I have now made the interface to
3033 host_find_byname() similar to host_find_bydns(), with an argument
3034 containing the DNS resolver options. The host_find_byname() function now
3035 sets these options at its start, just as host_find_bydns() does. The smtp
3036 transport options dns_qualify_single and dns_search_parents are passed to
3037 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3038 of host_find_byname() use the default settings of RES_DEFNAMES
3039 (qualify_single) but not RES_DNSRCH (search_parents).
3041 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3042 spool_read_header() do less string testing, by means of a preliminary
3043 switch on the second character of optional "-foo" lines. (This is
3044 overdue, caused by the large number of possibilities that now exist.
3045 Originally there were few.) While I was there, I also converted the
3046 str(n)cmp tests so they don't re-test the leading "-" and the first
3047 character, in the hope this might squeeze out yet more improvement.
3049 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3050 flag allowing group syntax was set by the header_syntax check but not
3051 turned off, possible causing trouble later; (2) The flag was not being
3052 set at all for the header_verify test, causing "group"-style headers to
3053 be rejected. I have now set it in this case, and also caused header_
3054 verify to ignore an empty address taken from a group. While doing this, I
3055 came across some other cases where the code for allowing group syntax
3056 while scanning a header line wasn't quite right (mostly, not resetting
3057 the flag correctly in the right place). These bugs could have caused
3058 trouble for malformed header lines. I hope it is now all correct.
3060 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3061 with the "reply" argument non-NULL. The code, however (which originally
3062 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3063 but it didn't always do it. This confused somebody who was copying the
3064 code for some other use. I have removed all the tests.
3066 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3067 feature that was used to support insecure browsers during the U.S. crypto
3068 embargo. It requires special client support, and Exim is probably the
3069 only MTA that supported it -- and would never use it because real RSA is
3070 always available. This code has been removed, because it had the bad
3071 effect of slowing Exim down by computing (never used) parameters for the
3072 RSA_EXPORT functionality.
3074 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3075 authenticator to fail if there's a tab character in the incoming data
3076 (there should never be unless someone is messing about, as it's supposed
3077 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3078 if the connection is using TLS or if the remote IP is the same as the
3079 local IP, and the "valid-client-cert option" if a client certificate has
3082 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3083 authenticators. This can be used for authorization after authentication
3084 succeeds. (In the case of plaintext, it servers for both authentication
3087 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3088 if any retry times were supplied.
3090 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3091 connection, where there is no client IP address to check. In this
3092 situation, the verify now always succeeds.
3094 PH/25 Applied John Jetmore's -Mset patch.
3096 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3098 PH/27 In a string expansion for a processed (not raw) header when multiple
3099 headers of the same name were present, leading whitespace was being
3100 removed from all of them, but trailing whitespace was being removed only
3101 from the last one. Now trailing whitespace is removed from each header
3102 before concatenation. Completely empty headers in a concatenation (as
3103 before) are ignored.
3105 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3106 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3108 PH/29 [Removed. This was a change that I later backed out, and forgot to
3109 correct the ChangeLog entry (that I had efficiently created) before
3110 committing the later change.]
3112 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3113 address errors (4xx response to RCPT) over the same connection as other
3114 messages routed to the same hosts. Such deliveries are always "forced",
3115 so retry times are not inspected. This resulted in far too many retries
3116 for the affected addresses. The effect occurred only when there were more
3117 hosts than the hosts_max_try setting in the smtp transport when it had
3118 the 4xx errors. Those hosts that it had tried were not added to the list
3119 of hosts for which the message was waiting, so if all were tried, there
3120 was no problem. Two fixes have been applied:
3122 (i) If there are any address or message errors in an SMTP delivery, none
3123 of the hosts (tried or untried) are now added to the list of hosts
3124 for which the message is waiting, so the message should not be a
3125 candidate for sending over the same connection that was used for a
3126 successful delivery of some other message. This seems entirely
3127 reasonable: after all the message is NOT "waiting for some host".
3128 This is so "obvious" that I'm not sure why it wasn't done
3129 previously. Hope I haven't missed anything, but it can't do any
3130 harm, as the worst effect is to miss an optimization.
3132 (ii) If, despite (i), such a delivery is accidentally attempted, the
3133 routing retry time is respected, so at least it doesn't keep
3134 hammering the server.
3136 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3137 in ${readsocket because some servers need this prod.
3139 PH/32 Added some extra debug output when updating a wait-xxx database.
3141 PH/33 The hint "could be header name not terminated by colon", which has been
3142 given for certain expansion errors for a long time, was not being given
3143 for the ${if def:h_colon_omitted{... case.
3145 PH/34 The spec says: "With one important exception, whenever a domain list is
3146 being scanned, $domain contains the subject domain." There was at least
3147 one case where this was not true.
3149 PH/35 The error "getsockname() failed: connection reset by peer" was being
3150 written to the panic log as well as the main log, but it isn't really
3151 panic-worthy as it just means the connection died rather early on. I have
3152 removed the panic log writing for the ECONNRESET error when getsockname()
3155 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3156 runs only) independently of the message's sender address. This meant
3157 that, if the 4xx error was in fact related to the sender, a different
3158 message to the same recipient with a different sender could confuse
3159 things. In particular, this can happen when sending to a greylisting
3160 server, but other circumstances could also provoke similar problems.
3161 I have changed the default so that the retry time for these errors is now
3162 based a combination of the sender and recipient addresses. This change
3163 can be overridden by setting address_retry_include_sender=false in the
3166 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3167 remote server are returned as part of bounce messages. This was not
3168 happening for LMTP over a pipe (the lmtp transport), but now it is the
3169 same for both kinds of LMTP.
3171 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3172 in header lines that were in fact CNAMEs. This is no longer the case.
3174 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3175 and queue runs started by the daemon processed all messages. This has
3176 been fixed so that -R and -S can now usefully be given with -q<time>.
3178 PH/40 Import PCRE release 6.7 (fixes some bugs).
3180 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3182 PH/42 Give an error if -q is specified more than once.
3184 PH/43 Renamed the variables $interface_address and $interface_port as
3185 $received_ip_address and $received_port, to make it clear that these
3186 values apply to message reception, and not to the outgoing interface when
3187 a message is delivered. (The old names remain recognized, of course.)
3189 PH/44 There was no timeout on the connect() call when using a Unix domain
3190 socket in the ${readsocket expansion. There now is.
3192 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3193 be meaningful with "accept".
3195 SC/01 Eximstats V1.43
3196 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3198 SC/02 Eximstats V1.44
3199 Use a glob alias rather than an array ref in the generated
3200 parser. This improves both readability and performance.
3202 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3203 Collect SpamAssassin and rejection statistics.
3204 Don't display local sender or destination tables unless
3205 there is data to show.
3206 Added average volumes into the top table text output.
3208 SC/04 Eximstats V1.46
3209 Collect data on the number of addresses (recipients)
3210 as well as the number of messages.
3212 SC/05 Eximstats V1.47
3213 Added 'Message too big' to the list of mail rejection
3214 reasons (thanks to Marco Gaiarin).
3216 SC/06 Eximstats V1.48
3217 Mainlog lines which have GMT offsets and are too short to
3218 have a flag are now skipped.
3220 SC/07 Eximstats V1.49 (Alain Williams)
3221 Added the -emptyok flag.
3223 SC/08 Eximstats V1.50
3224 Fixes for obtaining the IP address from reject messages.
3226 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3227 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3228 whitespace changes from 4.64-PH/27
3230 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3233 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3234 are found, allow negative numbers in numeric criteria)
3236 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3238 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3239 to match changes made in 4.64-PH/43
3241 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3243 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3244 "domain" is actually an IP address, and force a failure. This locks out
3245 those revolvers/nameservers that support "A-for-A" lookups, in
3246 contravention of the specifications.
3248 PH/48 When a host name was looked up from an IP address, and the subsequent
3249 forward lookup of the name timed out, the host name was left in
3250 $sender_host_name, contrary to the specification.
3252 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3253 restricted to single-key lookups, Exim was not diagnosing an error if
3254 * or *@ was used with a query-style lookup.
3256 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3258 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3259 long ago, but noone interested enough thought of it. Let's just say that
3260 the "1.1" means that there are some new functions that weren't there at
3261 some point in the past.
3263 PH/51 Error processing for expansion failure of helo_data from an smtp
3264 transport during callout processing was broken.
3266 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3267 tested/used via the -bh/-bhc/-bs options.
3269 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3270 bug, fixed in subsequent PCRE releases).
3272 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3273 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3275 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3281 SC/01 Use a glob alias rather than an array ref in eximstats generated
3282 parser. This improves both readability and performance.
3284 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3285 Don't display local sender or destination tables in eximstats unless
3286 there is data to show.
3287 Added average volumes into the eximstats top table text output.
3289 SC/03 Collect data on the number of addresses (recipients) as well
3290 as the number of messages in eximstats.
3292 TF/01 Correct an error in the documentation for the redirect router. Exim
3293 does (usually) call initgroups() when daemonizing.
3295 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3296 with consistent privilege compared to when running as a daemon.
3298 TF/03 Note in the spec that $authenticated_id is not set for local
3299 submissions from trusted users.
3301 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3302 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3304 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3305 by adding some example configuration directives to the default
3306 configuration file. A little bit of work is required to uncomment the
3307 directives and define how usernames and passwords are checked, but
3308 there is now a framework to start from.
3310 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3311 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3312 without this. I don't know how relevant this is to other LDAP libraries.
3314 PH/02 Add the verb name to the "unknown ACL verb" error.
3316 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3318 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3320 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3321 directory not expanded when it should be if an expanded home directory
3322 was set for the address (which is overridden by the transport).
3324 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3327 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3328 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3329 because it is too late at that time, and has no effect.
3331 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3332 security issue with \' (bugzilla #107). I could not use the
3333 PQescapeStringConn() function, because it needs a PGconn value as one of
3336 PH/08 When testing addresses using -bt, indicate those final addresses that
3337 are duplicates that would not cause an additional delivery. At least one
3338 person was confused, thinking that -bt output corresponded to deliveries.
3339 (Suppressing duplicates isn't a good idea as you lose the information
3340 about possibly different redirections that led to the duplicates.)
3342 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3343 systems where poll() doesn't work, in particular OS X.
3345 PH/10 Added more information to debugging output for retry time not reached.
3347 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3348 operations in malware.c.
3350 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3353 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3354 syslog_duplication set false, log lines that would normally be written
3355 both the the main log and to the reject log were not written to syslog at
3358 PH/14 In the default configuration, change the use of "message" in ACL warn
3359 statements to "add_header".
3361 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3362 not followed by a command (e.g. "seen endif").
3364 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3365 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3368 PH/17 Added extra conditions to the default value of delay_warning_condition
3372 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3373 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3374 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3377 The Auto-Submitted: and various List- headers are standardised, whereas I
3378 don't think Precedence: ever was.
3380 PH/18 Refactored debugging code in route_finduser() to show more information,
3381 in particular, the error code if getpwnam() issues one.
3383 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3384 This is apparently needed in addition to the PH/07 change above to avoid
3385 any possible encoding problems.
3387 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3388 but not after initializing Perl.
3390 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3391 output them only if debugging. By default they are written stderr,
3392 apparently, which is not desirable.
3394 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3397 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3400 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3402 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3403 authenticated or an ident call has been made. Suppress the default
3404 values for $authenticated_id and $authenticated_sender (but permit -oMai
3405 and -oMas) when testing with -bh.
3407 PH/24 Re-jigged the order of the tests in the default configuration so that the
3408 tests for valid domains and recipients precede the DNS black list and CSA
3409 tests, on the grounds that those ones are more expensive.
3411 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3412 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3413 This bug exists in every version of Exim that I still have, right back to
3416 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3417 However, an attempt to turn on case-sensitivity in a regex key by
3418 including (?-i) didn't work because the subject string was already
3419 lowercased, and the effects were non-intuitive. It turns out that a
3420 one-line patch can be used to allow (?-i) to work as expected.
3426 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3427 other effects) broke the use of negated acl sub-conditions.
3429 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3432 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3433 "Deny" causes Exim to reject the incoming connection with a 554 error.
3434 Unfortunately, if there is a major crisis, such as a disk failure,
3435 tcp-wrappers gives "deny", whereas what one would like would be some
3436 kind of temporary error. A kludge has been added to help with this.
3437 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3438 554 error is used if errno is still zero or contains ENOENT (which occurs
3439 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3442 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3444 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3445 errors. Otherwise a message that provokes a temporary error (when other
3446 messages do not) can cause a whole host to time out.
3448 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3449 addresses were routed directly to files or pipes from a redirect router.
3450 File deliveries just didn't batch; pipe deliveries might have suffered
3453 PH/06 A failure to get a lock for a hints database would erroneously always say
3454 "Failed to get write lock", even when it was really a read lock.
3456 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3457 of 0600. This has been changed to use the value of the lockfile_mode
3458 option (which defaults to 0600).
3460 PH/08 Applied small patch from the Sieve maintainer.
3462 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3463 folder from quota calculations, a direct delivery into this folder messed
3464 up the contents of the maildirsize file. This was because the regex was
3465 used only to exclude .Trash (or whatever) when the size of the mailbox
3466 was calculated. There was no check that a delivery was happening into an
3467 excluded directory. This bug has been fixed by ignoring all quota
3468 processing for deliveries into excluded directories.
3470 PH/10 Added the maildirfolder_create_regex option to appendfile.
3476 PH/01 The code for finding all the local interface addresses on a FreeBSD
3477 system running IPv6 was broken. This may well have applied to all BSD
3478 systems, as well as to others that have similar system calls. The broken
3479 code found IPv4 interfaces correctly, but gave incorrect values for the
3480 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3481 that it would not match correctly against @[] and not recognize the IPv6
3484 PH/02 The ipliteral router was not recognizing addresses of the form user@
3485 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3487 PH/03 Added disable_ipv6.
3489 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3490 decoded form, because it is most often used to construct To: headers
3491 lines in autoreplies, and the decoded form may well be syntactically
3492 invalid. However, $reply_address has leading white space removed, and all
3493 newlines turned into spaces so that the autoreply transport does not
3496 PH/05 If group was specified without a user on a router, and no group or user
3497 was specified on a transport, the group from the router was ignored.
3499 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3500 for visible compile-time settings that can be used to change these
3501 numbers, for those that want even more. Backwards compatibility with old
3502 spool files has been maintained. However, going back to a previous Exim
3503 release will lost any variables that are in spool files.
3505 PH/07 Two small changes when running in the test harness: increase delay when
3506 passing a TCP/IP connection to a new process, in case the original
3507 process has to generate a bounce, and remove special handling of
3508 127.0.0.2 (sic), which is no longer necessary.
3510 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3511 be the same on different OS.
3513 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3516 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3517 whether --show-vars was specified or not
3519 JJ/02 exipick: Added support for new ACL variable spool format introduced
3522 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3523 syntactically invalid From: or Reply-to: line, and a filter used this to
3524 generate an autoreply, and therefore failed to obtain an address for the
3525 autoreply, Exim could try to deliver to a non-existent relative file
3526 name, causing unrelated and misleading errors. What now happens is that
3527 it logs this as a hard delivery error, but does not attempt to create a
3530 PH/11 The exinext utility has a -C option for testing purposes, but although
3531 the given file was scanned by exinext itself; it wasn't being passed on
3532 when Exim was called.
3534 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3535 an end-of-file indication when reading a command response.
3537 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3538 compiled. In many other places in Exim, IPv6 addresses are always
3539 recognized, so I have changed this. It also means that IPv4 domain
3540 literals of the form [IPV4:n.n.n.n] are now always recognized.
3542 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3543 used if the router is not running as root, for example, when verifying at
3544 ACL time, or when using -bh. The debugging output from this situation was
3545 non-existent - all you got was a failure to exec. I have made two
3548 (a) Failures to set uid/gid, the current directory, or a process leader
3549 in a subprocess such as that created by queryprogram now generate
3550 suitable debugging output when -d is set.
3552 (b) The queryprogram router detects when it is not running as root,
3553 outputs suitable debugging information if -d is set, and then runs
3554 the subprocess without attempting to change uid/gid.
3556 PH/15 Minor change to Makefile for building test_host (undocumented testing
3559 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3560 additional section of a DNS packet that returns MX or SRV records.
3561 Instead, it always explicitly searches for A/AAAA records. This avoids
3562 major problems that occur when a DNS server includes only records of one
3563 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3564 fixed another bug: if SRV records were looked up and the corresponding
3565 address records were *not* found in the additional section, the port
3566 values from the SRV records were lost.
3568 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3569 using the correct key (the original address) when searching the retry
3570 rules in order to find which one to use for generating the retry hint.
3572 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3573 adding the default one. Similarly, if it contains a Reply-To: header, the
3574 errors_reply_to option, if set, is not used.
3576 PH/19 When calculating a retry time, Exim used to measure the "time since
3577 failure" by looking at the "first failed" field in the retry record. Now
3578 it does not use this if it is later than than the arrival time of the
3579 message. Instead it uses the arrival time. This makes for better
3580 behaviour in cases where some deliveries succeed, thus re-setting the
3581 "first failed" field. An example is a quota failure for a huge message
3582 when small messages continue to be delivered. Without this change, the
3583 "time since failure" will always be short, possible causing more frequent
3584 delivery attempts for the huge message than are intended.
3585 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3587 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3588 $1, $2, $3) because the numerical variables can be reset during some
3589 expansion items (e.g. "match"), thereby losing the authentication data.
3591 PH/21 Make -bV show the size of off_t variables so that the test suite can
3592 decide whether to run tests for quotas > 2G.
3594 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3595 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3596 filecount value is greater than 2G or if a quota value is greater than 2G
3597 on a system where the size of off_t is not greater than 4, a panic error
3600 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3601 never match. The debug and -bh output now contains an explicit error
3602 message indicating a malformed IPv4 address or mask.
3604 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3605 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3606 PH/23 above applies.
3608 PH/25 Do not write to syslog when running in the test harness. The only
3609 occasion when this arises is a failure to open the main or panic logs
3610 (for which there is an explicit test).
3612 PH/26 Added the /no_tell option to "control=freeze".
3614 PH/27 If a host name lookup failed very early in a connection, for example, if
3615 the IP address matched host_lookup and the reverse lookup yielded a name
3616 that did not have a forward lookup, an error message of the form "no IP
3617 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3618 could be logged. Now it outputs the IP address instead of "NULL".
3620 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3621 allows the sender and the authenticated sender to be set when
3622 submitting a message from within Exim. Since child_open_exim() is
3623 documented for local_scan(), the new function should be too.
3625 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3626 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3627 results in an empty string is now treated as unset.
3629 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3631 PH/31 Added sender_verify_fail logging option.
3633 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3634 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3635 tidied the source and removed it altogether.
3637 PH/33 When a queue run was abandoned because the load average was too high, a
3638 log line was always written; now it is written only if the queue_run log
3639 selector is set. In addition, the log line for abandonment now contains
3640 information about the queue run such as the pid. This is always present
3641 in "start" and "stop" lines but was omitted from the "abandon" line.
3643 PH/34 Omit spaces between a header name and the colon in the error message that
3644 is given when verify = headers_syntax fails (if there are lots of them,
3645 the message gets confusing).
3647 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3648 names, as there are now some PTR records that contain slashes. This check
3649 is only to protect against broken name servers that fall over on strange
3650 characters, so the fact that it applies to all lookups doesn't matter.
3652 PH/36 Now that the new test suite is complete, we can remove some of the
3653 special code in Exim that was needed for the old test suite. For example,
3654 sorting DNS records because real resolvers return them in an arbitrary
3655 order. The new test suite's fake resolver always returns records in the
3658 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3659 bounces) except when queue_only is set, to avoid logging races between
3660 the different processes.
3662 PH/38 Panic-die if .include specifies a non-absolute path.
3664 PH/39 A tweak to the "H" retry rule from its user.
3666 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3667 a label. They prevented compilation on older perls.
3669 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3670 a warning to be raised on newish perls.
3672 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3673 on queue. Changes to match documented behaviour of showing count of
3674 messages matching specified criteria.
3676 PH/40 Changed the default ident timeout from 30s to 5s.
3678 PH/41 Added support for the use of login_cap features, on those BSD systems
3679 that have them, for controlling the resources used by pipe deliveries.
3681 PH/42 The content-scanning code uses fopen() to create files in which to put
3682 message data. Previously it was not paying any attention to the mode of
3683 the files. Exim runs with umask(0) because the rest of the code creates
3684 files with open(), and sets the required mode explicitly. Thus, these
3685 files were ending up world-writeable. This was not a big issue, because,
3686 being within the spool directory, they were not world-accessible. I have
3687 created a function called modefopen, which takes an additional mode
3688 argument. It sets umask(777), creates the file, chmods it to the required
3689 mode, then resets the umask. All the relevant calls to fopen() in the
3690 content scanning code have been changed to use this function.
3692 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3693 to 24 hours. This avoids potential overflow problems when processing G
3694 and H retry rules. I suspect nobody ever tinkers with this value.
3696 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3698 PH/45 When the plaintext authenticator is running as a client, the server's
3699 challenges are checked to ensure they are valid base64 strings. By
3700 default, the authentication attempt is cancelled if an invalid string is
3701 received. Setting client_ignore_invalid_base64 true ignores these errors.
3702 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3703 they are received. Thus, the responses can be made to depend on the
3704 challenges. If an invalid string is ignored, an empty string is placed in
3707 PH/46 Messages that are created by the autoreply transport now contains a
3708 References: header, in accordance with RFCs 2822 and 3834.
3710 PH/47 Added authenticated_sender_force to the smtp transport.
3712 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3714 PH/49 Installed latest patch from the Sieve maintainer.
3716 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3717 was also set, the appendfile transport was unnecessarily scanning a
3718 directory of message files (e.g. for maildir delivery) to find the count
3719 of files (along with the size), even though it did not need this
3720 information. It now does the scan only if it needs to find either the
3721 size of the count of files.
3723 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3725 PH/52 Two bugs concerned with error handling when the smtp transport is
3728 (i) Exim was not creating retry information for temporary errors given
3729 for individual recipients after the DATA command when the smtp transport
3730 was used in LMTP mode. This meant that they could be retried too
3731 frequently, and not timed out correctly.
3733 (ii) Exim was setting the flag that allows error details to be returned
3734 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3735 recipients that were returned after the DATA command.
3737 PH/53 This is related to PH/52, but is more general: for any failing address,
3738 when detailed error information was permitted to be returned to the
3739 sender, but the error was temporary, then after the final timeout, only
3740 "retry timeout exceeded" was returned. Now it returns the full error as
3741 well as "retry timeout exceeded".
3743 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3744 do this, and (what is worse) MTAs that accept it.
3746 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3747 will now be deprecated.
3749 PH/56 New os.c-cygwin from the Cygwin maintainer.
3751 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3752 formats (previously only available in exim formats via -bpr, -bpru,
3753 and -bpra. Now also available in native and exiqgrep formats)
3755 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3756 with very large, slow to parse queues
3758 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3760 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3762 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3763 responses to authentication challenges, though it was showing the
3764 challenges; (ii) I've removed the CR characters from the debug output for
3767 PH/58 Allow for the insertion of a newline as well as a space when a string
3768 is turned into more than one encoded-word during RFC 2047 encoding. The
3769 Sieve code now uses this.
3771 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3772 data_4xx, lost_connection, tls_required.
3774 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3775 message was being sent as an SMTP response.
3777 PH/61 Add -l and -k options to exicyclog.
3779 PH/62 When verifying, if an address was redirected to one new address, so that
3780 verification continued, and the new address failed or deferred after
3781 having set something in $address_data, the value of $address_data was not
3782 passed back to the ACL. This was different to the case when no
3783 redirection occurred. The value is now passed back in both cases.
3785 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3786 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3787 use login_cap.h, so on its own it isn't the distinguishing feature. The
3788 new name refers directly to the setclassresources() function.
3790 PH/65 Added configuration files for NetBSD3.
3792 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3794 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3795 is preferred over IPv4.
3797 PH/68 The bounce_return_message and bounce_return_body options were not being
3798 honoured for bounces generated during the reception of non-SMTP messages.
3799 In particular, this applied to messages rejected by the ACL. This bug has
3800 been fixed. However, if bounce_return_message is true and bounce_return_
3801 body is false, the headers that are returned for a non-SMTP message
3802 include only those that have been read before the error was detected.
3803 (In the case of an ACL rejection, they have all been read.)
3805 PH/69 The HTML version of the specification is now built in a directory called
3806 spec_html instead of spec.html, because the latter looks like a path with
3807 a MIME-type, and this confuses some software.
3809 PH/70 Catch two compiler warnings in sieve.c.
3811 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3812 function verify_get_ident() calls ip_connect() to connect a socket, but
3813 if the "connect()" function timed out, ip_connect() used to close the
3814 socket. However, verify_get_ident() also closes the socket later, and in
3815 between Exim writes to the log, which may get opened at this point. When
3816 the socket was closed in ip_connect(), the log could get the same file
3817 descriptor number as the socket. This naturally causes chaos. The fix is
3818 not to close the socket in ip_connect(); the socket should be closed by
3819 the function that creates it. There was only one place in the code where
3820 this was missing, in the iplookup router, which I don't think anybody now
3821 uses, but I've fixed it anyway.
3823 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3824 well as to direct DNS lookups. Otherwise the handling of names in host
3825 lists is inconsistent and therefore confusing.
3831 PH/01 Two changes to the default runtime configuration:
3833 (1) Move the checks for relay_from_hosts and authenticated clients from
3834 after to before the (commented out) DNS black list checks.
3836 (2) Add control=submission to the relay_from_hosts and authenticated
3837 clients checks, on the grounds that messages accepted by these
3838 statements are most likely to be submissions.
3840 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3842 (1) Generate an error if the third argument for the ${prvs expansion is
3845 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3848 (3) Reset the variables that are obtained from the first argument of
3849 ${prvscheck and used in the second argument before leaving the code,
3850 because their memory is reclaimed, so using them afterwards may do
3853 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3854 one (it's much easier than Tom thought :-).
3856 (5) Because of (4), we can now allow for the use of $prvscheck_result
3857 inside the third argument.
3859 PH/03 For some reason, the default setting of PATH when running a command from
3860 a pipe transport was just "/usr/bin". I have changed it to
3863 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3864 anything to be listed in the output from -bV.
3866 PH/05 When a filter generated an autoreply, the entire To: header line was
3867 quoted in the delivery log line, like this:
3869 => >A.N.Other <ano@some.domain> <original@ddress> ...
3871 This has been changed so that it extracts the operative address. There
3872 may be more than one such address. If so, they are comma-separated, like
3875 => >ano@some.domain,ona@other.domain <original@ddress> ...
3877 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3878 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3879 not being looked up in the rDNS to get a host name, Exim was showing the
3880 IP address twice in Received: lines, even though the IP addresses were
3881 identical. For example:
3883 Received: from [1.2.3.4] (helo=[1.2.3.4])
3885 However, if the real host name was known, it was omitting the HELO data
3886 if it matched the actual IP address. This has been tidied up so that it
3887 doesn't show the same IP address twice.
3889 PH/07 When both +timestamp and +memory debugging was on, the value given by
3890 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3891 called by the debug printing, thereby overwriting the timestamp buffer.
3892 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3894 PH/08 When the original message was included in an autoreply transport, it
3895 always said "this is a copy of the message, including all the headers",
3896 even if body_only or headers_only was set. It now gives an appropriate
3899 PH/09 Applied a patch from the Sieve maintainer which:
3901 o fixes some comments
3902 o adds the (disabled) notify extension core
3903 o adds some debug output for the result of if/elsif tests
3904 o points to the current vacation draft in the documentation
3905 and documents the missing references header update
3909 o fixes a bug in processing the envelope test (when testing
3910 multiple envelope elements, the last element determined the
3913 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3914 Electronic Mail") by including:
3916 Auto-submitted: auto-generated
3918 in the messages that it generates (bounce messages and others, such as
3919 warnings). In the case of bounce messages for non-SMTP messages, there was
3920 also a typo: it was using "Auto_submitted" (underscore instead of
3921 hyphen). Since every message generated by Exim is necessarily in response
3922 to another message, thes have all been changed to:
3924 Auto-Submitted: auto-replied
3926 in accordance with these statements in the RFC:
3928 The auto-replied keyword:
3930 - SHOULD be used on messages sent in direct response to another
3931 message by an automatic process,
3933 - MUST NOT be used on manually-generated messages,
3935 - MAY be used on Delivery Status Notifications (DSNs) and Message
3936 Disposition Notifications (MDNs),
3938 - MUST NOT be used on messages generated by automatic or periodic
3939 processes, except for messages which are automatic responses to
3942 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3943 to the default Received: header definition.
3945 PH/12 Added log selector acl_warn_skipped (default on).
3947 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3948 variables because (a) they are in the wrong storage pool and (b) even if
3949 they were copied, it wouldn't work properly because of the caching.
3951 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3952 checking when decoding. Apparently there are clients that generate
3953 overlong encoded strings. Why am I not surprised?
3955 PH/15 If the first argument of "${if match_address" was not empty, but did not
3956 contain an "@" character, Exim crashed. Now it writes a panic log message
3957 and treats the condition as false.
3959 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3961 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3962 extension "envelope-auth". The code is finished and in agreement with
3963 other implementations, but there is no documentation so far and in fact,
3964 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3965 not changing the active code.
3967 Print executed "if" and "elsif" statements when debugging is used. This
3968 helps a great deal to understand what a filter does.
3970 Document more things not specified clearly in RFC3028. I had all this
3971 sorted out, when out of a sudden new issues came to my mind. Oops."
3973 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3976 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3977 canonical form (as documented). However, after a host name lookup from
3978 the IP address, check_host() was doing a simple string comparison with
3979 addresses acquired from the DNS when checking that the found name did
3980 have the original IP as one of its addresses. Since any found IPv6
3981 addresses are likely to be in abbreviated form, the comparison could
3982 fail. Luckily, there already exists a function for doing the comparison
3983 by converting both addresses to binary, so now that is used instead of
3984 the text comparison.
3986 PH/20 There was another similar case to PH/19, when a complete host name was
3987 given in a host list; looking up its IP address could give an abbreviated
3988 form, whereas the current host's name might or might not be abbreviated.
3989 The same fix has been applied.
3995 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3996 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3999 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4000 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4002 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4004 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4005 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4006 Also, for "Auto-Submitted", treat anything other than "no" as
4007 non-personal, in accordance with RFC 3834. (Previously it treated
4008 anything starting "auto-" as non-personal.)
4010 TF/01 The control=submission/name=... option had a problem with syntax
4011 errors if the name included a slash character. The /name= option
4012 now slurps the rest of the string, so it can include any characters
4013 but it must come last in the list of options (after /sender_retain
4016 PH/05 Some modifications to the interface to the fake nameserver for the new
4024 TK/01 Added the "success_on_redirect" address verification option. See
4025 NewStuff for rationale and an example.
4027 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4029 PH/02 Patch to exigrep to allow it to work on syslog lines.
4031 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4032 fread() to skip over the body file's header line, because in Cygwin the
4033 header line is locked and is inaccessible.
4035 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4036 co-exist for some time) to make it clear that it is the Exim ID that is
4037 referenced, not the Message-ID: header line.
4039 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4040 string_format() function, because snprintf() does not exist on all
4043 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4044 ${stat: expansion item.
4046 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4047 protocol synchronization error", to keep the pedants happy.
4049 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4050 well as for IRIX systems, when gcc is being used. See the host.c source
4053 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4055 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4058 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4059 transport and to the smtp transport in LMTP mode.
4061 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4063 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4064 run a filter in a subprocess. This could lead to confusion in subsequent
4065 lookups in the parent process. There should also be a search_tidyup() at
4066 the end of the subprocess.
4068 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4069 only if the host matched helo_try_verify_hosts, which caused the
4070 verification to occur when the EHLO/HELO command was issued. The ACL just
4071 tested the remembered result. Now, if a previous verification attempt has
4072 not happened, "verify = helo" does it there and then.
4074 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4076 TK/03 Fix log output including CR from clamd.
4078 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4079 exist provoked a memory error which could cause a segfault.
4081 PH/15 Installed PCRE 6.2
4083 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4085 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4086 of the problem. Specifically, suggested +O2 rather than +O1 for the
4089 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4091 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4092 record, the port was not correctly logged when the outgoing_port log
4093 selector was set (it logged the transort's default port).
4095 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4096 fallback_hosts, and "hosts" in the smtp transport.
4098 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4099 host errors such as "Connection refused".
4101 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4102 authentication with radiusclient 0.4.9:
4104 - Error returned from rc_read_config was caught wrongly
4105 - Username/password not passed on to radius server due to wrong length.
4107 The presumption is that some radiusclient API changes for 4.51/PH/17
4108 were not taken care of correctly. The code is still untested by me (my
4109 Linux distribution still has 0.3.2 of radiusclient), but it was
4110 contributed by a Radius user.
4112 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4113 expanding the "port" option of the smtp transport.
4115 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4116 while reading a MIME header. Thanks to Tom Hughes for a patch.
4118 PH/24 Include config.h inside local_scan.h so that configuration settings are
4121 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4122 that in an ACL for RCPT (for example), you can examine exactly what was
4125 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4126 commands, but it was not correctly comparing the address with the actual
4127 client host address. Thus, it would show the EHLO address in Received:
4128 header lines when this was not necessary.
4130 PH/27 Added the % operator to ${eval:}.
4132 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4133 it should be ignoring failures (because with -C, for example, it has lost
4134 privilege). It wasn't ignoring creation failures other than "already
4137 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4140 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4141 because an input line was too long, either on its own, or by virtue of
4142 too many continuations, the temporary file was not being removed, and the
4143 return code was incorrect.
4145 PH/31 Missing "BOOL" in function definition in filtertest.c.
4147 PH/32 Applied Sieve patches from the maintainer.
4149 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4151 PH/33 Added "verify = not_blind".
4153 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4154 Local/Makefile (with some defaults set). These are used in built scripts
4155 such as exicyclog, but they have never been used in the exim_install
4156 script (though there are many overriding facilities there). I have
4157 arranged that the exim_install script now takes note of these two
4160 PH/35 Installed configuration files for Dragonfly.
4162 PH/36 When a locally submitted message by a trusted user did not contain a
4163 From: header, and the sender address was obtained from -f or from an SMTP
4164 MAIL command, and the trusted user did not use -F to supply a sender
4165 name, $originator_name was incorrectly used when constructing a From:
4166 header. Furthermore, $originator_name was used for submission mode
4167 messages from external hosts without From: headers in a similar way,
4168 which is clearly wrong.
4170 PH/37 Added control=suppress_local_fixups.
4172 PH/38 When log_selector = +received_sender was set, and the addition of the
4173 sender made the log line's construction buffer exactly full, or one byte
4174 less than full, an overflow happened when the terminating "\n" was
4177 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4178 when the result of a list match is failure because a DNS lookup failed.
4180 PH/40 RM_COMMAND is now used in the building process.
4182 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4183 the "build-* directories that it finds.
4185 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4186 address in a domain literal was a prefix of an interface address.
4188 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4189 when verifying a sender address, unless rewrite_headers is false.
4191 PH/44 Wrote a long comment about why errors_to addresses are verified as
4192 recipients, not senders.
4194 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4195 the ratelimit ACL was added.
4197 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4199 PH/46 Added extra information about PostgreSQL errors to the error string.
4201 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4202 avoiding the need to install special zones in a real server. This is
4203 backwards compatible; if it can't find the fake resolver, it drops back.
4204 Thus, both old and new test suites can be run.
4206 TF/02 Added util/ratelimit.pl
4208 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4209 clock is set back in time.
4211 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4212 Candler <B.Candler@pobox.com>.
4214 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4215 OK for addresses that are the result of redirections.
4217 PH/48 A number of further additions for the benefit of the new test suite,
4218 including a fake gethostbyname() that interfaces to the fake DNS resolver
4221 TF/06 The fix for widen_domains has also been applied to qualify_single and
4222 search_parents which are the other dnslookup options that can cause
4225 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4228 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4230 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4231 service" if it encountered an MX record with an empty target hostname.
4232 The message is now "an MX or SRV record indicated no SMTP service".
4234 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4235 if the DNS of the sending site is misconfigured. This is quite a
4236 common situation. This change restores the behaviour of treating a
4237 helo verification defer as a failure.
4239 PH/51 If self=fail was set on a router, the bounce message did not include the
4240 actual error message.
4246 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4248 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4249 command itself ended in error, the underlying message about the transport
4250 filter timeout was being overwritten with the pipe command error. Now the
4251 underlying error message should be appended to the second error message.
4253 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4255 PH/02 Reduce the amount of output that "make" produces by default. Full output
4256 can still be requested.
4258 PH/03 The warning log line about a condition test deferring for a "warn" verb
4259 was being output only once per connection, rather than after each
4260 occurrence (because it was using the same function as for successful
4261 "warn" verbs). This seems wrong, so I have changed it.
4263 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4264 it should not have, which might have caused a crash in the right
4265 circumstances, but probably never did.
4267 PH/04 Installed a modified version of Tony Finch's patch to make submission
4268 mode fix the return path as well as the Sender: header line, and to
4269 add a /name= option so that you can make the user's friendly name appear
4272 TF/03 Added the control = fakedefer ACL modifier.
4274 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4275 Mark Lowes for thorough testing.
4277 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4279 TK/03 Merged latest SRS patch from Miles Wilton.
4281 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4282 with the definition in sysexits.h (which is #included earlier).
4283 Fortunately, Exim does not actually use EX_OK. The code used to try to
4284 preserve the sysexits.h value, by assuming that macro definitions were
4285 scanned for macro replacements. I have been disabused of this notion,
4286 so now the code just undefines EX_OK before #including unistd.h.
4288 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4289 in the smtp transport. When a block could not be written in a single
4290 write() function, the timeout was being re-applied to each part-write.
4291 This seems wrong - if the receiver was accepting one byte at a time it
4292 would take for ever. The timeout is now adjusted when this happens. It
4293 doesn't have to be particularly precise.
4295 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4296 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4298 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4299 without a domain if the check to <postmaster@domain> fails.
4301 SC/01 Eximstats: added -xls and the ability to specify output files
4302 (patch written by Frank Heydlauf).
4304 SC/02 Eximstats: use FileHandles for outputting results.
4306 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4308 SC/04 Eximstats: fixed display of large numbers with -nvr option
4310 SC/05 Eximstats: fixed merging of reports with empty tables.
4312 SC/06 Eximstats: added the -include_original_destination flag
4314 SC/07 Eximstats: removed tabs and trailing whitespace.
4316 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4318 TK/06 MBOX spool code: Add real "From " MBOX separator line
4319 so the .eml file is really in mbox format (even though
4320 most programs do not really care). Patch from Alex Miller.
4322 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4323 The latter is generated from $received_to and is only set if the
4324 message has one envelope recipient. SA can use these headers,
4325 obviously out-of-the-box. Patch from Alex Miller.
4327 PH/08 The ${def test on a variable was returning false if the variable's
4328 value was "0", contrary to what the specification has always said!
4329 The result should be true unless the variable is empty.
4331 PH/09 The syntax error of a character other than { following "${if
4332 def:variable_name" (after optional whitespace) was not being diagnosed.
4333 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4334 accidental colon was present, for example, could give incorrect results.
4336 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4337 result is used (not including appendfile, where other changes are about
4340 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4341 This involved changing a lot of size variables from int to off_t. It
4342 should work with maildirs and everything.
4344 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4345 spamd dying while we are connected to it.
4347 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4350 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4351 function for BDB 4.3.
4353 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4355 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4356 the sources for PCRE have been re-arranged and more files are now
4359 PH/15 The code I had for printing potentially long long variables in PH/11
4360 above was not the best (it lost precision). The length of off_t variables
4361 is now inspected at build time, and an appropriate printing format (%ld
4362 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4363 to be "long long int" or "long int". This is needed for the internal
4364 formatting function string_vformat().
4366 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4367 the configuration file to be ":syslog", then the script "guesses" where
4368 the logs files are, rather than using the compiled in default. In our
4369 case the guess is not the same as the compiled default, so the script
4370 suddenly stopped working when I started to use syslog. The patch checks
4371 to see if log_file_path is "". If so, it attempts to read it from exim
4372 with no configuration file to get the compiled in version, before it
4373 falls back to the previous guessing code."
4375 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4376 implementing BATV in an Exim configuration. See NewStuff for the gory
4379 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4380 Makefile that are specific to HP-UX.
4382 PH/18 If the "use_postmaster" option was set for a recipient callout together
4383 with the "random" option, the postmaster address was used as the MAIL
4384 FROM address for the random test, but not for the subsequent recipient
4385 test. It is now used for both.
4387 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4388 patch removes a few documentation additions to RFC 3028, because the
4389 latest draft now contains them. It adds the new en;ascii-case comparator
4390 and a new error check for 8bit text in MIME parts. Comparator and
4391 require names are now matched exactly. I enabled the subaddress
4392 extension, but it is not well tested yet (read: it works for me)."
4394 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4395 rework some of the code of TK/09 above to avoid the hardwired use of
4396 "%lld" and "long long". Replaced the call to snprintf() with a call to
4399 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4400 records point to non-existent hosts", "retry timeout exceeded", and
4401 "retry time not reached for any host after a long failure period".
4403 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4404 experimental DomainKeys support:
4406 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4407 (2) On an error such as an illegally used "control", the wrong name for
4408 the control was given.
4410 These problems did NOT occur unless DomainKeys support was compiled.
4412 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4414 PH/24 Added ${if match_ip condition.
4416 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4417 databases so that it will be absolutely obvious if a crash occurs in the
4418 DB library. This is a regular occurrence (often caused by mis-matched
4421 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4422 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4423 picked up on a user's system that detects such things. There doesn't seem
4424 to be a gcc warning option for this - only an attribute that has to be
4425 put on the function's prototype. It seems that in Fedora Core 4 they have
4426 set this on a number of new functions. No doubt there will be more in due
4429 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4430 specify a transport. However, if an address that was verified by such a
4431 router was the subject of a callout, Exim crashed because it tried to
4432 read the rcpt_include_affixes from the non-existent transport. Now it
4433 just assumes that the setting of that option is false. This bug was
4434 introduced by 4.51/PH/31.
4436 PH/28 Changed -d+all to exclude +memory, because that information is very
4437 rarely of interest, but it makes the output a lot bigger. People tend to
4438 do -d+all out of habit.
4440 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4441 code in os-type was giving problems when libc.so lives in lib64, like on
4444 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4445 aren't the modern standard, and it seems that some systems' include files
4446 don't always have them. Exim was already checking for some of the newer
4447 ones like T_AAAA, and defining it itself. I've added checks for all the
4448 record types that Exim uses.
4450 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4451 not automatically generating a new one, as it is supposed to. This
4452 prevented TLS from working. If the file did exist, but contained invalid
4453 data, a new version was generated, as expected. It was only the case of a
4454 non-existent file that was broken.
4456 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4457 with a change in libdomainkeys > 0.64.
4459 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4460 from DNS. If the selector record carries the flag, it now has
4461 precedence over the domain-wide flag.
4463 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4465 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4466 the use of a port name that isn't defined in /etc/services) occurred, the
4467 message was deferred as in a normal delivery, and thus remained on the
4468 spool, instead of being failed because of the mua_wrapper setting. This
4469 is now fixed, and I tidied up some of the mua_wrapper messages at the
4472 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4473 the messages in a hash of arrays rather than using individual hashes.
4474 This is a bit cleaner and results in dramatic memory savings, albeit
4475 at a slight CPU cost.
4477 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4478 as requested by Marc Sherman.
4480 SC/10 Eximstats: added histograms for user specified patterns as requested
4483 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4485 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4486 fopen() in the content-scanning modules that did not already have it.
4492 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4493 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4495 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4497 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4499 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4500 to test Sieve filters that use "vacation".
4502 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4503 that changes the way the GnuTLS parameters are stored in the cache file.
4504 The new format can be generated externally. For backward compatibility,
4505 if the data in the cache doesn't make sense, Exim assumes it has read an
4506 old-format file, and it generates new data and writes a new file. This
4507 means that you can't go back to an older release without removing the
4510 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4511 work if there are any delivery delays because "one_time" forces the
4512 parent to be marked "delivered", so its unseen clone is never tried
4513 again. For this reason, Exim now forbids the simultaneous setting of
4516 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4517 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4518 entry if you want to know the details. The fix, however, neglected to
4519 consider the case when local delivery batching is involved. The test for
4520 "previously delivered" was not happening when checking to see if an
4521 address could be batched with a previous (undelivered) one; under
4522 certain circumstances this could lead to multiple deliveries to the same
4525 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4526 in its include files, and this causes problems building Exim.
4528 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4529 header_syntax) but Exim was just ignoring anything given after a slash.
4530 In particular, this caused confusion with an attempt to use "verify =
4531 reverse_host_lookup/defer_ok". An error is now given when options are
4532 supplied for verify items that do not have them. (Maybe reverse_host_
4533 lookup should have a defer_ok option, but that's a different point.)
4535 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4536 defined by RFC 821) to 2048, because there were problems with some AUTH
4537 commands, and RFC 1869 says the size should be increased for extended
4538 SMTP commands that take arguments.
4540 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4543 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4544 "unknown" error; now it says that the functionality isn't in the binary.
4546 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4547 an address' error message when a string expansion fails (syntax or
4548 whatever). Otherwise the password may appear in the log. Following change
4549 PH/42 below, there is no longer a chance of it appearing in a bounce
4552 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4554 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4555 of its addresses was ever tried. (Bugzilla bug #2.)
4557 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4558 the result incorrectly in the debug output. (It correctly added a newline
4559 to what was transported.)
4561 TF/01 Added $received_time.
4563 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4564 commented out examples of how to interface to a virus scanner and to
4565 SpamAssassin. Also added commented examples of av_scanner and
4566 spamd_address settings.
4568 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4569 and controls are allowed in which ACLs. There were a couple of minor
4570 errors. Some of the entries in the conditions table (which is a table of
4571 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4572 negation of where the condition IS allowed.
4574 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4576 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4577 header file does not have a version number, so I've had to invent a new
4578 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4579 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4580 radiusclient), but it was contributed by a Radius user.
4582 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4583 files or directories, for OpenSSL.
4585 PH/19 When an Exim process that is running as root has to create an Exim log
4586 file, it does so in a subprocess that runs as exim:exim so as to get the
4587 ownership right at creation (otherwise, other Exim processes might see
4588 the file with the wrong ownership). There was no test for failure of this
4589 fork() call, which would lead to the process getting stuck as it waited
4590 for a non-existent subprocess. Forks do occasionally fail when resources
4591 run out. I reviewed all the other calls to fork(); they all seem to check
4594 PH/20 When checking for unexpected SMTP input at connect time (before writing
4595 the banner), Exim was not dealing correctly with a non-positive return
4596 from the read() function. If the client had disconnected by this time,
4597 the result was a log entry for a synchronization error with an empty
4598 string after "input=" when read() returned zero. If read() returned -1
4599 (an event I could not check), uninitialized data bytes were printed.
4600 There were reports of junk text (parts of files, etc) appearing after
4603 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4605 PH/22 Added support for macro redefinition, and (re)definition in between
4606 driver and ACL definitions.
4608 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4609 forgetting to use the resulting value; it was using the unexpanded value.
4611 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4612 hadn't been configured. The fix is from Juergen Kreileder, who
4613 understands it better than I do:
4615 "Here's what I see happening with three configured cyrus_sasl
4616 authenticators configured (plain, login, cram-md5):
4618 On startup auth_cyrus_sasl_init() gets called for each of these.
4619 This means three calls to sasl_listmech() without a specified mech_list.
4620 => SASL tests which mechs of all available mechs actually work
4621 => three warnings about OTP not working
4622 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4624 With the patch, sasl_listmech() also gets called three times. But now
4625 SASL's mech_list option is set to the server_mech specified in the the
4626 authenticator. Or in other words, the answer from sasl_listmech()
4627 gets limited to just the mech you're testing for (which is different
4629 => the return list contains just 'plain' or 'login', 'cram-md5' or
4630 nothing depending on the value of ob->server_mech.
4632 I've just tested the patch: Authentication still works fine,
4633 unavailable mechs specified in the exim configuration are still
4634 caught, and the auth.log warnings about OTP are gone."
4636 PH/25 When debugging is enabled, the contents of the command line are added
4637 to the debugging output, even when log_selector=+arguments is not
4640 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4641 answer is "GNU", and only if the return is "GNU/something" is the answer
4644 PH/27 $acl_verify_message is now set immediately after the failure of a
4645 verification in an ACL, and so is available in subsequent modifiers. In
4646 particular, the message can be preserved by coding like this:
4648 warn !verify = sender
4649 set acl_m0 = $acl_verify_message
4651 Previously, $acl_verify_message was set only while expanding "message"
4652 and "log_message" when a very denied access.
4654 PH/28 Modified OS/os.c-Linux with
4656 -#ifndef OS_LOAD_AVERAGE
4657 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4659 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4660 nomenclature these days.)
4662 PH/29 Installed patch from the Sieve maintainer that adds the options
4663 sieve_useraddress and sieve_subaddress to the redirect router.
4665 PH/30 In these circumstances:
4666 . Two addresses routed to the same list of hosts;
4667 . First host does not offer TLS;
4668 . First host accepts first address;
4669 . First host gives temporary error to second address;
4670 . Second host offers TLS and a TLS session is established;
4671 . Second host accepts second address.
4672 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4673 and peerdn, if requested) that were in fact used only for the second
4676 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4677 attention to any local part prefix or suffix that was matched by the
4678 router that accepted the address. It now behaves in the same way as it
4679 does for delivery: the affixes are removed from the local part unless
4680 rcpt_include_affixes is set on the transport.
4682 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4683 timeout during the DATA phase of an incoming message.
4685 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4686 applied a patch sanctioned by the Sieve maintainer.
4688 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4689 the uid or gid is negative. A case of a negative gid caused this to be
4690 noticed. The fix allows for either to be negative.
4692 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4693 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4696 PH/36 The change PH/12 above was broken. Fixed it.
4698 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4699 the grounds that routing the same address twice would always produce the
4700 same answer. This might have been true once, but it is certainly no
4701 longer true now. Routing a child address may depend on the previous
4702 routing that produced that child. Some complicated redirection strategies
4703 went wrong when messages had multiple recipients, and made Exim's
4704 behaviour dependent on the order in which the addresses were given.
4706 I have moved the duplicate checking until after the routing is complete.
4707 Exim scans the addresses that are assigned to local and remote
4708 transports, and removes any duplicates. This means that more work will be
4709 done, as duplicates will always all be routed, but duplicates are
4710 presumably rare, so I don't expect this is of any significance.
4712 For deliveries to pipes, files, and autoreplies, the duplicate checking
4713 still happens during the routing process, since they are not going to be
4716 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4717 It corrects a timeout issue with spamd. This is Ian's comment: "The
4718 background is that sometimes spamd either never reads data from a
4719 connection it has accepted, or it never writes response data. The exiscan
4720 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4721 blindly assumes that writes won't block so it may never time out."
4723 PH/39 Allow G after quota size as well as K and M.
4725 PH/40 The value set for $authenticated_id in an authenticator may not contain
4726 binary zeroes or newlines because the value is written to log lines and
4727 to spool files. There was no check on this. Now the value is run through
4728 the string_printing() function so that such characters are converted to
4729 printable escape sequences.
4731 PH/41 $message_linecount is a new variable that contains the total number of
4732 lines in the message. Compare $body_linecount, which is the count for the
4735 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4736 bounce and delay warning messages, except in certain special cases, which
4739 (a) An SMTP error message from a remote host;
4740 (b) A message specified in a :fail: redirection;
4741 (c) A message specified in a "fail" command in a system filter;
4742 (d) A message specified in a FAIL return from the queryprogram router;
4743 (e) A message specified by the cannot_route_message router option.
4745 In these cases only, Exim does include the error details in bounce and
4746 warning messages. There are also a few cases where bland messages such
4747 as "unrouteable address" or "local delivery error" are given.
4749 PH/43 $value is now also set for the "else" part of a ${run expansion.
4751 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4752 being worked on, but at least Exim now implements the latest version to
4755 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4756 process to complete was treated as a delivery failure, a timeout while
4757 writing the message to the pipe was logged, but erroneously treated as a
4758 successful delivery. Such timeouts include transport filter timeouts. For
4759 consistency with the overall process timeout, these timeouts are now
4760 treated as errors, giving rise to delivery failures by default. However,
4761 there is now a new Boolean option for the pipe transport called
4762 timeout_defer, which, if set TRUE, converts the failures into defers for
4763 both kinds of timeout. A transport filter timeout is now identified in
4766 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4767 systems where "make" and "gmake" are different, calling "gmake" at top
4768 level broke things. I've arranged for the value of $(MAKE) to be passed
4769 from the Makefile to this script so that it can call the same version of
4773 A note about Exim versions 4.44 and 4.50
4774 ----------------------------------------
4776 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4777 changes of various kinds. As a consequence, a big documentation update was
4778 needed. This delayed the release for rather longer than seemed good, especially
4779 in the light of a couple of (minor) security issues. Therefore, the changes
4780 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4781 release. So 4.44 and 4.50 are in effect two different branches that both start
4784 I have left the 4.50 change log unchanged; it contains all the changes since
4785 4.43. The change log for 4.44 is below; many of its items are identical to
4786 those for 4.50. This seems to be the most sensible way to preserve the
4787 historical information.
4793 1. Minor wording change to the doc/README.SIEVE file.
4795 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4796 computation of the current number of files was incorrect.
4798 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4799 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4802 4. Give more explanation in the error message when the command for a transport
4803 filter fails to execute.
4805 5. There are several places where Exim runs a non-Exim command in a
4806 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4807 was being done only for the command run by the queryprogram router. It is
4808 now done for all such subprocesses. The other cases are: ${run, transport
4809 filters, and the commands run by the lmtp and pipe transports.
4811 6. Added CONFIGURE_GROUP build-time option.
4813 7. Some older OS have a limit of 256 on the maximum number of file
4814 descriptors. Exim was using setrlimit() to set 1000 as a large value
4815 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4816 systems. I've change it so that if it can't get 1000, it tries for 256.
4818 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4819 was an oversight, and furthermore, ever since the addition of extra
4820 controls (e.g. 4.43/32), the checks on when to allow different forms of
4821 "control" were broken. There should now be diagnostics for all cases when a
4822 control that does not make sense is encountered.
4824 9. Added the /retain_sender option to "control=submission".
4826 10. $recipients is now available in the predata ACL (oversight).
4828 11. Tidy the search cache before the fork to do a delivery from a message
4829 received from the command line. Otherwise the child will trigger a lookup
4830 failure and thereby defer the delivery if it tries to use (for example) a
4831 cached ldap connection that the parent has called unbind on.
4833 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4834 of $address_data from the recipient verification was clobbered by the
4835 sender verification.
4837 13. The value of address_data from a sender verification is now available in
4838 $sender_address_data in subsequent conditions in the ACL statement.
4840 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4842 15. Added a new option "connect=<time>" to callout options, to set a different
4845 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4846 was its contents. (It was OK if the option was not defined at all.)
4848 17. A "Completed" log line is now written for messages that are removed from
4849 the spool by the -Mrm option.
4851 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4852 information about exactly what failed.
4854 19. Added -dd to debug only the daemon process.
4856 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4857 handles timeouts, both on the server side and network timeouts. Renamed the
4858 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4860 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4861 It is now set to "smtps".
4863 22. $host_address is now set to the target address during the checking of
4864 ignore_target_hosts.
4866 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4867 being passed; this would have caused $sender_host_name to have been used if
4868 matching the list had actually called for a host name (not very likely,
4869 since this list is usually IP addresses). A host name is now passed as
4872 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4873 code that specifies a non-restarting handler (typically sigaction() in
4874 modern systems) in an attempt to fix a rare and obscure crash bug.
4876 25. Narrowed the window for a race in the daemon that could cause it to ignore
4877 SIGCHLD signals. This is not a major problem, because they are used only to
4878 wake it up if nothing else does.
4880 26. A malformed maildirsize file could cause Exim to calculate negative values
4881 for the mailbox size or file count. Odd effects could occur as a result.
4882 The maildirsize information is now recalculated if the size or filecount
4885 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4886 support for a long time. Removed HAVE_SYS_VFS_H.
4888 28. Installed the latest version of exipick from John Jetmore.
4890 29. In an address list, if the pattern was not a regular expression, an empty
4891 subject address (from a bounce message) matched only if the pattern was an
4892 empty string. Non-empty patterns were not even tested. This was the wrong
4893 because it is perfectly reasonable to use an empty address as part of a
4894 database query. An empty address is now tested by patterns that are
4895 lookups. However, all the other forms of pattern expect the subject to
4896 contain a local part and a domain, and therefore, for them, an empty
4897 address still always fails if the pattern is not itself empty.
4899 30. Exim went into a mad DNS loop when attempting to do a callout where the
4900 host was specified on an smtp transport, and looking it up yielded more
4901 than one IP address.
4903 31. Re-factored the code for checking spool and log partition space into a
4904 function that finds that data and another that does the check. The former
4905 is then used to implement four new variables: $spool_space, $log_space,
4906 $spool_inodes, and $log_inodes.
4908 32. The RFC2047 encoding function was originally intended for short strings
4909 such as real names; it was not keeping to the 75-character limit for
4910 encoded words that the RFC imposes. It now respects the limit, and
4911 generates multiple encoded words if necessary. To be on the safe side, I
4912 have increased the buffer size for the ${rfc2047: expansion operator from
4915 33. It is now permitted to omit both strings after an "if" condition; if the
4916 condition is true, the result is "true". As before, when the second string
4917 is omitted, a false condition yields an empty string. This makes it less
4918 cumbersome to write custom ACL and router conditions.
4920 34. Failure to deliver a bounce message always caused it to be frozen, even if
4921 there was an errors_to setting on the router. The errors_to setting is now
4924 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4925 canonical form (fully expanded) before being placed in
4926 $sender_host_address.
4928 36. The table in the code that translates DNS record types into text (T_A to
4929 "A" for instance) was missing entries for NS and CNAME. It is just possible
4930 that this could have caused confusion if both these types were looked up
4931 for the same domain, because the text type is used as part of Exim's
4932 per-process caching. But the chance of anyone hitting this buglet seems
4935 37. The dnsdb lookup has been extended in a number of ways.
4937 (1) There is a new type, "zns", which walks up the domain tree until it
4938 finds some nameserver records. It should be used with care.
4940 (2) There is a new type, "mxh", which is like "mx" except that it returns
4941 just the host names, not the priorities.
4943 (3) It is now possible to give a list of domains (or IP addresses) to be
4944 looked up. The behaviour when one of the lookups defers can be
4945 controlled by a keyword.
4947 (4) It is now possible to specify the separator character for use when
4948 multiple records are returned.
4950 38. The dnslists ACL condition has been extended: it is now possible to supply
4951 a list of IP addresses and/or domains to be looked up in a particular DNS
4954 39. Added log_selector=+queue_time_overall.
4956 40. When running the queue in the test harness, wait just a tad after forking a
4957 delivery process, to get repeatability of debugging output.
4959 41. Include certificate and key file names in error message when GnuTLS fails
4960 to set them up, because the GnuTLS error message doesn't include the name
4961 of the failing file when there is a problem reading it.
4963 42. Allow both -bf and -bF in the same test run.
4965 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4967 44. The "Exiscan patch" is now merged into the mainline Exim source.
4969 45. Sometimes the final signoff response after QUIT could fail to get
4970 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4971 < 0 before doing a fflush(). This bug looks as though it goes back to the
4972 introduction of TLS in release 3.20, but "sometimes" must have been rare
4973 because the tests only now provoked it.
4975 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4976 (this can affect the format of dates).
4978 47. exim_tidydb, when checking for the continued existence of a message for
4979 which it has found a message-specific retry record, was not finding
4980 messages that were in split spool directories. Consequently, it was
4981 deleting retry records that should have stayed in existence.
4983 48. Steve fixed some bugs in eximstats.
4985 49. The SPA authentication driver was not abandoning authentication and moving
4986 on to the next authenticator when an expansion was forced to fail,
4987 contradicting the general specification for all authenticators. Instead it
4988 was generating a temporary error. It now behaves as specified.
4990 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4991 (the order specifies the preference for clients). The order is now AES256,
4992 AES128, 3DES, ARCFOUR128.
4994 51. Small patch to Sieve code - explicitly set From: when generating an
4997 52. Exim crashed if a remote delivery caused a very long error message to be
4998 recorded - for instance if somebody sent an entire SpamAssassin report back
4999 as a large number of 550 error lines. This bug was coincidentally fixed by
5000 increasing the size of one of Exim's internal buffers (big_buffer) that
5001 happened as part of the Exiscan merge. However, to be on the safe side, I
5002 have made the code more robust (and fixed the comments that describe what
5005 53. Now that there can be additional text after "Completed" in log lines (if
5006 the queue_time_overall log selector is set), a one-byte patch to exigrep
5007 was needed to allow it to recognize "Completed" as not the last thing in
5010 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5011 patch that reportedly fixes this has been added. I am not expert enough to
5012 create a test for it. This is what the patch creator wrote:
5014 "I found a little strange behaviour of ldap code when working with
5015 Windows 2003 AD Domain, where users was placed in more than one
5016 Organization Units. When I tried to give exim partial DN, the exit code
5017 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5018 But simultaneously result of request was absolutely normal ldap result,
5019 so I produce this patch..."
5021 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5022 REFERENCE, so I have modified the code to exclude the patch when that macro
5025 55. Some experimental protocols are using DNS PTR records for new purposes. The
5026 keys for these records are domain names, not reversed IP addresses. The
5027 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5028 leaves it alone. Component reversal etc. now happens only for IP addresses.
5031 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5033 57. Double the size of the debug message buffer (to 2048) so that more of very
5034 long debug lines gets shown.
5036 58. The exicyclog utility now does better if the number of log files to keep
5037 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5039 59. Two changes related to the smtp_active_hostname option:
5041 (1) $smtp_active_hostname is now available as a variable.
5042 (2) The default for smtp_banner uses $smtp_active_hostname instead
5043 of $primary_hostname.
5045 60. The host_aton() function is supposed to be passed a string that is known
5046 to be a valid IP address. However, in the case of IPv6 addresses, it was
5047 not checking this. This is a hostage to fortune. Exim now panics and dies
5048 if the condition is not met. A case was found where this could be provoked
5049 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5050 components; fortuitously, this particular loophole had already been fixed
5051 by change 4.50/55 above.
5053 If there are any other similar loopholes, the new check in host_aton()
5054 itself should stop them being exploited. The report I received stated that
5055 data on the command line could provoke the exploit when Exim was running as
5056 exim, but did not say which command line option was involved. All I could
5057 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5058 running as the user.
5061 61. There was a buffer overflow vulnerability in the SPA authentication code
5062 (which came originally from the Samba project). I have added a test to the
5063 spa_base64_to_bits() function which I hope fixes it.
5066 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5067 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5069 63. The daemon start-up calls getloadavg() while still root for those OS that
5070 need the first call to be done as root, but it missed one case: when
5071 deliver_queue_load_max is set with deliver_drop_privilege. This is
5072 necessary for the benefit of the queue runner, because there is no re-exec
5073 when deliver_drop_privilege is set.
5075 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5076 This has been fixed.
5078 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5079 was in use, was not putting the data itself into the right store pool;
5080 consequently, it could be overwritten for a subsequent message in the same
5081 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5084 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5086 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5087 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5088 the function were treating the return as a boolean value, which happened to
5089 work because 0=false and not-0=true, but is not correct code.
5091 68. The host_aton() function was not handling scoped IPv6 addresses (those
5092 with, for example, "%eth0" on the end) correctly.
5094 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5095 negated items (that is, ~something) in unsigned ints. Some compilers
5096 apparently mutter when there is no cast.
5098 70. If an address verification called from an ACL failed, and did not produce a
5099 user-specific message (i.e. there was only a "system" message), nothing was
5100 put in $acl_verify_message. In this situation, it now puts the system
5103 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5104 session; change 4.31/43 added the unwanted input to the log line - except
5105 that it did not do this in the start of session case. It now does.
5107 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5108 This is wrong and can cause the other end to generate a synchronization
5109 error if it is another Exim or anything else that does the synchronization
5110 check. A QUIT command is no longer sent after a timeout.
5112 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5113 during host lookups.
5115 74. The defer_ok option of callout verification was not working if it was used
5116 when verifying addresses in header lines, that is, for this case:
5118 verify = header_sender/callout=defer_ok
5120 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5121 those file descriptors could be used for SMTP connections. If anything
5122 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5123 could be sent to the SMTP client, causing chaos. The daemon now opens
5124 stdin, stdout, and stderr to /dev/null when it puts itself into the
5127 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5128 log by default. The user can override this with suitable Perl magic.
5130 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5131 the log message when discard triggers, was not working for the DATA ACL or
5132 for the non-SMTP ACL.
5134 78. Error message wording change in sieve.c.
5136 79. If smtp_accept_max_per_host was set, the number of connections could be
5137 restricted to fewer than expected, because the daemon was trying to set up
5138 a new connection before checking whether the processes handling previous
5139 connections had finished. The check for completed processes is now done
5140 earlier. On busy systems, this bug wouldn't be noticed because something
5141 else would have woken the daemon, and it would have reaped the completed
5144 80. If a message was submitted locally by a user whose login name contained one
5145 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5146 It caused a spool format error. I have fixed the spool reading code. A
5147 related problem was that the "from" clause in the Received: line became
5148 illegal because of the space(s). It is now covered by ${quote_local_part.
5150 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5153 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5155 83. Patch from Sieve maintainer to fix unterminated string problem in
5156 "vacation" handling.
5158 84. Some minor changes to the Linux configuration files to help with other
5159 OS variants using glibc.
5161 85. One more patch for Sieve to update vacation handling to latest spec.
5164 ----------------------------------------------------
5165 See the note above about the 4.44 and 4.50 releases.
5166 ----------------------------------------------------
5172 1. Change 4.43/35 introduced a bug that caused file counts to be
5173 incorrectly computed when quota_filecount was set in an appendfile
5176 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5177 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5180 3. Give more explanation in the error message when the command for a transport
5181 filter fails to execute.
5183 4. There are several places where Exim runs a non-Exim command in a
5184 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5185 was being done only for the command run by the queryprogram router. It is
5186 now done for all such subprocesses. The other cases are: ${run, transport
5187 filters, and the commands run by the lmtp and pipe transports.
5189 5. Some older OS have a limit of 256 on the maximum number of file
5190 descriptors. Exim was using setrlimit() to set 1000 as a large value
5191 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5192 systems. I've change it so that if it can't get 1000, it tries for 256.
5194 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5195 was an oversight, and furthermore, ever since the addition of extra
5196 controls (e.g. 4.43/32), the checks on when to allow different forms of
5197 "control" were broken. There should now be diagnostics for all cases when a
5198 control that does not make sense is encountered.
5200 7. $recipients is now available in the predata ACL (oversight).
5202 8. Tidy the search cache before the fork to do a delivery from a message
5203 received from the command line. Otherwise the child will trigger a lookup
5204 failure and thereby defer the delivery if it tries to use (for example) a
5205 cached ldap connection that the parent has called unbind on.
5207 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5208 of $address_data from the recipient verification was clobbered by the
5209 sender verification.
5211 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5212 was its contents. (It was OK if the option was not defined at all.)
5214 11. A "Completed" log line is now written for messages that are removed from
5215 the spool by the -Mrm option.
5217 12. $host_address is now set to the target address during the checking of
5218 ignore_target_hosts.
5220 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5221 being passed; this would have caused $sender_host_name to have been used if
5222 matching the list had actually called for a host name (not very likely,
5223 since this list is usually IP addresses). A host name is now passed as
5226 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5227 code that specifies a non-restarting handler (typically sigaction() in
5228 modern systems) in an attempt to fix a rare and obscure crash bug.
5230 15. Narrowed the window for a race in the daemon that could cause it to ignore
5231 SIGCHLD signals. This is not a major problem, because they are used only to
5232 wake it up if nothing else does.
5234 16. A malformed maildirsize file could cause Exim to calculate negative values
5235 for the mailbox size or file count. Odd effects could occur as a result.
5236 The maildirsize information is now recalculated if the size or filecount
5239 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5240 support for a long time. Removed HAVE_SYS_VFS_H.
5242 18. Updated exipick to current release from John Jetmore.
5244 19. Allow an empty sender to be matched against a lookup in an address list.
5245 Previously the only cases considered were a regular expression, or an
5248 20. Exim went into a mad DNS lookup loop when doing a callout where the
5249 host was specified on the transport, if the DNS lookup yielded more than
5252 21. The RFC2047 encoding function was originally intended for short strings
5253 such as real names; it was not keeping to the 75-character limit for
5254 encoded words that the RFC imposes. It now respects the limit, and
5255 generates multiple encoded words if necessary. To be on the safe side, I
5256 have increased the buffer size for the ${rfc2047: expansion operator from
5259 22. Failure to deliver a bounce message always caused it to be frozen, even if
5260 there was an errors_to setting on the router. The errors_to setting is now
5263 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5264 canonical form (fully expanded) before being placed in
5265 $sender_host_address.
5267 24. Updated eximstats to version 1.33
5269 25. Include certificate and key file names in error message when GnuTLS fails
5270 to set them up, because the GnuTLS error message doesn't include the name
5271 of the failing file when there is a problem reading it.
5273 26. Expand error message when OpenSSL has problems setting up cert/key files.
5276 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5277 (this can affect the format of dates).
5279 28. exim_tidydb, when checking for the continued existence of a message for
5280 which it has found a message-specific retry record, was not finding
5281 messages that were in split spool directories. Consequently, it was
5282 deleting retry records that should have stayed in existence.
5284 29. eximstats updated to version 1.35
5285 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5286 1.35 - bugfix such that pie charts by volume are generated correctly
5288 30. The SPA authentication driver was not abandoning authentication and moving
5289 on to the next authenticator when an expansion was forced to fail,
5290 contradicting the general specification for all authenticators. Instead it
5291 was generating a temporary error. It now behaves as specified.
5293 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5294 (the order specifies the preference for clients). The order is now AES256,
5295 AES128, 3DES, ARCFOUR128.
5297 31. Small patch to Sieve code - explicitly set From: when generating an
5300 32. Exim crashed if a remote delivery caused a very long error message to be
5301 recorded - for instance if somebody sent an entire SpamAssassin report back
5302 as a large number of 550 error lines. This bug was coincidentally fixed by
5303 increasing the size of one of Exim's internal buffers (big_buffer) that
5304 happened as part of the Exiscan merge. However, to be on the safe side, I
5305 have made the code more robust (and fixed the comments that describe what
5308 33. Some experimental protocols are using DNS PTR records for new purposes. The
5309 keys for these records are domain names, not reversed IP addresses. The
5310 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5311 leaves it alone. Component reversal etc. now happens only for IP addresses.
5314 34. The host_aton() function is supposed to be passed a string that is known
5315 to be a valid IP address. However, in the case of IPv6 addresses, it was
5316 not checking this. This is a hostage to fortune. Exim now panics and dies
5317 if the condition is not met. A case was found where this could be provoked
5318 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5319 components; fortuitously, this particular loophole had already been fixed
5320 by change 4.50/55 or 4.44/33 above.
5322 If there are any other similar loopholes, the new check in host_aton()
5323 itself should stop them being exploited. The report I received stated that
5324 data on the command line could provoke the exploit when Exim was running as
5325 exim, but did not say which command line option was involved. All I could
5326 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5327 running as the user.
5330 35. There was a buffer overflow vulnerability in the SPA authentication code
5331 (which came originally from the Samba project). I have added a test to the
5332 spa_base64_to_bits() function which I hope fixes it.
5335 36. The daemon start-up calls getloadavg() while still root for those OS that
5336 need the first call to be done as root, but it missed one case: when
5337 deliver_queue_load_max is set with deliver_drop_privilege. This is
5338 necessary for the benefit of the queue runner, because there is no re-exec
5339 when deliver_drop_privilege is set.
5341 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5342 was in use, was not putting the data itself into the right store pool;
5343 consequently, it could be overwritten for a subsequent message in the same
5344 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5347 38. Sometimes the final signoff response after QUIT could fail to get
5348 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5349 < 0 before doing a fflush(). This bug looks as though it goes back to the
5350 introduction of TLS in release 3.20, but "sometimes" must have been rare
5351 because the tests only now provoked it.
5357 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5358 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5359 Now it returns an integer. A number of calls were still expecting a T/F
5360 return. Fortuitously, in all cases, the tests worked in OK situations,
5361 which is the norm. However, things would have gone wrong on any write
5362 failures on the smtp file descriptor. This function is used when sending
5363 messages over SMTP and also when doing verify callouts.
5365 2. When Exim is called to do synchronous delivery of a locally submitted
5366 message (the -odf or -odi options), it no longer closes stderr before doing
5369 3. Implemented the mua_wrapper option.
5371 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5373 5. Implemented the functions header_remove(), header_testname(),
5374 header_add_at_position(), and receive_remove_recipient(), and exported them
5377 6. If an ACL "warn" statement specified the addition of headers, Exim already
5378 inserted X-ACL-Warn: at the start if there was no header name. However, it
5379 was not making this test for the second and subsequent header lines if
5380 there were newlines in the string. This meant that an invalid header could
5381 be inserted if Exim was badly configured.
5383 7. Allow an ACL "warn" statement to add header lines at the start or after all
5384 the Received: headers, as well as at the end.
5386 8. Added the rcpt_4xx retry error code.
5388 9. Added postmaster_mailfrom=xxx to callout verification option.
5390 10. Added mailfrom=xxxx to the callout verification option, for verify=
5393 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5394 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5396 12. Inserted some casts to stop certain compilers warning when using pointer
5397 differences as field lengths or precisions in printf-type calls (mostly
5398 affecting debugging statements).
5400 13. Added optional readline() support for -be (dynamically loaded).
5402 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5403 same clock tick as a message's arrival, so that its received time was the
5404 same as the "first fail" time on the retry record, and that message
5405 remained on the queue past the ultimate address timeout, every queue runner
5406 would try a delivery (because it was past the ultimate address timeout) but
5407 after another failure, the ultimate address timeout, which should have then
5408 bounced the address, did not kick in. This was a "< instead of <=" error;
5409 in most cases the first failure would have been in the next clock tick
5410 after the received time, and all would be well.
5412 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5413 being recognized when the domain list was tested by the match_domain
5414 condition in an expansion string.
5416 16. Added the ${str2b64: operator.
5418 17. Exim was always calling setrlimit() to set a large limit for the number of
5419 processes, without checking whether the existing limit was already
5420 adequate. (It did check for the limit on file descriptors.) Furthermore,
5421 errors from getrlimit() and setrlimit() were being ignored. Now they are
5422 logged to the main and panic logs, but Exim does carry on, to try to do its
5423 job under whatever limits there are.
5425 18. Imported PCRE 5.0.
5427 19. Trivial typo in log message " temporarily refused connection" (the leading
5430 20. If the log selector return_path_on_delivery was set and an address was
5431 redirected to /dev/null, the delivery process crashed because it assumed
5432 that a return path would always be set for a "successful" delivery. In this
5433 case, the whole delivery is bypassed as an optimization, and therefore no
5436 21. Internal re-arrangement: the function for sending a challenge and reading
5437 a response while authentication was assuming a zero-terminated challenge
5438 string. It's now changed to take a pointer and a length, to allow for
5439 binary data in such strings.
5441 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5443 23. Exim was not respecting finduser_retries when seeking the login of the
5444 uid under which it was called; it was always trying 10 times. (The default
5445 setting of finduser_retries is zero.) Also, it was sleeping after the final
5446 failure, which is pointless.
5448 24. Implemented tls_on_connect_ports.
5450 25. Implemented acl_smtp_predata.
5452 26. If the domain in control=submission is set empty, Exim assumes that the
5453 authenticated id is a complete email address when it generates From: or
5454 Sender: header lines.
5456 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5457 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5458 chown and chgrp in /bin and hostname in /usr/bin.
5460 28. Exim was keeping the "process log" file open after each use, just as it
5461 does for the main log. This opens the possibility of it remaining open for
5462 long periods when the USR1 signal hits a daemon. Occasional processlog
5463 errors were reported, that could have been caused by this. Anyway, it seems
5464 much more sensible not to leave this file open at all, so that is what now
5467 29. The long-running daemon process does not normally write to the log once it
5468 has entered its main loop, and it closes the log before doing so. This is
5469 so that log files can straightforwardly be renamed and moved. However,
5470 there are a couple of unusual error situations where the daemon does write
5471 log entries, and I had neglected to close the log afterwards.
5473 30. The text of an SMTP error response that was received during a remote
5474 delivery was being truncated at 512 bytes. This is too short for some of
5475 the long messages that one sometimes sees. I've increased the limit to
5478 31. It is now possible to make retry rules that apply only when a message has a
5479 specific sender, in particular, an empty sender.
5481 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5482 it possible to be selective about when SMTP synchronization is enforced.
5484 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5486 32. Implemented hosts_connection_nolog.
5488 33. Added an ACL for QUIT.
5490 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5493 35. Added mailbox_size and mailbox_filecount to appendfile.
5495 36. Added control = no_multiline_responses to ACLs.
5497 37. There was a bug in the logic of the code that waits for the clock to tick
5498 in the case where the clock went backwards by a substantial amount such
5499 that the microsecond fraction of "now" was more than the microsecond
5500 fraction of "then" (but the whole seconds number was less).
5502 38. Added support for the libradius Radius client library this is found on
5503 FreeBSD (previously only the radiusclient library was supported).
5509 1. When certain lookups returned multiple values in the form name=value, the
5510 quoting of the values was not always being done properly. Specifically:
5511 (a) If the value started with a double quote, but contained no whitespace,
5513 (b) If the value contained whitespace other than a space character (i.e.
5514 tabs or newlines or carriage returns) it was not quoted.
5515 This fix has been applied to the mysql and pgsql lookups by writing a
5516 separate quoting function and calling it from the lookup code. The fix
5517 should probably also be applied to nisplus, ibase and oracle lookups, but
5518 since I cannot test any of those, I have not disturbed their existing code.
5520 2. A hit in the callout cache for a specific address caused a log line with no
5521 reason for rejecting RCPT. Now it says "Previous (cached) callout
5522 verification failure".
5524 3. There was an off-by-one bug in the queryprogram router. An over-long
5525 return line was truncated at 256 instead of 255 characters, thereby
5526 overflowing its buffer with the terminating zero. As well as fixing this, I
5527 have increased the buffer size to 1024 (and made a note to document this).
5529 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5530 when Exim is waiting for an SMTP response from a remote server, Exim
5531 restarts its select() call on the socket, thereby resetting its timeout.
5532 This is not a problem when such interrupts are rare. Somebody set up a cron
5533 job to run exiwhat every 2 minutes, which is less than the normal select()
5534 timeout (5 or 10 minutes). This meant that the select() timeout never
5535 kicked in because it was always reset. I have fixed this by comparing the
5536 time when an interrupt arrives with the time at the start of the first call
5537 to select(). If more time than the timeout has elapsed, the interrupt is
5538 treated as a timeout.
5540 5. Some internal re-factoring in preparation for the addition of Sieve
5541 extensions (by MH). In particular, the "personal" test is moved to a
5542 separate function, and given an option for scanning Cc: and Bcc: (which is
5543 not set for Exim filters).
5545 6. When Exim created an email address using the login of the caller as the
5546 local part (e.g. when creating a From: or Sender: header line), it was not
5547 quoting the local part when it contained special characters such as @.
5549 7. Installed new OpenBSD configuration files.
5551 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5552 try to make them clearer.
5554 9. Callout options, other than the timeout value, were being ignored when
5555 verifying sender addresses in header lines. For example, when using
5557 verify = header_sender/callout=no_cache
5559 the cache was (incorrectly) being used.
5561 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5562 only the Cygwin environment.
5564 11. When return_path_on_delivery was set as a log selector, if different remote
5565 addresses in the same message used different return paths and parallel
5566 remote delivery occurred, the wrong values would sometimes be logged.
5567 (Whenever a remote delivery process finished, the return path value from
5568 the most recently started remote delivery process was logged.)
5570 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5571 lines when AUTH and/or TLS are in use. This is the "received protocol"
5572 field. Exim used to use "asmtp" for authenticated SMTP, without any
5573 indication (in the protocol name) for TLS use. Now it follows the RFC and
5574 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5575 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5576 names appear in log lines as well as in Received: header lines.
5578 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5579 extensions, and comparison tests, and to fix some bugs.
5581 14. Changes to the "personal" filter test:
5583 (1) The test was buggy in that it was just doing the equivalent of
5584 "contains" tests on header lines. For example, if a user's address was
5585 anne@some.where, the "personal" test would incorrectly be true for
5587 To: susanne@some.where
5589 This test is now done by extracting each address from the header in turn,
5590 and checking the entire address. Other tests that are part of "personal"
5591 are now done using regular expressions (for example, to check local parts
5592 of addresses in From: header lines).
5594 (2) The list of non-personal local parts in From: addresses has been
5595 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5596 taken from the Sieve specification recommendations.
5598 (3) If the message contains any header line starting with "List-" it is
5599 treated as non-personal.
5601 (4) The test for "circular" in the Subject: header line has been removed
5602 because it now seems ill-conceived.
5604 15. Minor typos in src/EDITME comments corrected.
5606 16. Installed latest exipick from John Jetmore.
5608 17. If headers_add on a router specified a text string that was too long for
5609 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5610 of string_sprintf() is now avoided.
5612 18. $message_body_size was not set (it was always zero) when running the DATA
5613 ACL and the local_scan() function.
5615 19. For the "mail" command in an Exim filter, no default was being set for
5616 the once_repeat time, causing a random time value to be used if "once" was
5617 specified. (If the value happened to be <= 0, no repeat happened.) The
5618 default is now 0s, meaning "never repeat". The "vacation" command was OK
5619 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5620 (I found it when inspecting the code).
5622 20. There is now an overall timeout for performing a callout verification. It
5623 defaults to 4 times the callout timeout, which applies to individual SMTP
5624 commands during the callout. The overall timeout applies when there is more
5625 than one host that can be tried. The timeout is checked before trying the
5626 next host. This prevents very long delays if there are a large number of
5627 hosts and all are timing out (e.g. when the network connections are timing
5628 out). The value of the overall timeout can be changed by specifying an
5629 additional sub-option for "callout", called "maxwait". For example:
5631 verify = sender/callout=5s,maxwait=20s
5633 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5634 to the end before writing, but this should make it even safer).
5636 22. Exim was forgetting that it had advertised PIPELINING for the second and
5637 subsequent messages on an SMTP connection. It was also not resetting its
5638 memory on STARTTLS and an internal HELO.
5640 23. When Exim logs an SMTP synchronization error within a session, it now
5641 records whether PIPELINING has been advertised or not.
5643 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5644 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5645 rather than long int.
5647 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5649 26. Added the never_mail option to autoreply.
5655 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5656 crash if the getsockname() call failed; this can happen if a connection is
5657 closed very soon after it is established. The problem was simply in the
5658 order in which certain operations were done, causing Exim to try to write
5659 to the SMTP stream before it had set up the file descriptor. The bug has
5660 been fixed by making things happen in the correct order.
5666 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5667 before the connection was closed, thus losing the rejection response.
5669 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5670 some early Solaris releases, but causes trouble in current releases where
5671 socklen_t is defined.
5673 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5676 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5679 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5680 that releasing the top of it at the end releases what was used for sub-
5681 expansions (unless the block got too big). However, discard this block if
5682 the first thing is a variable or header, so that we can use its block when
5683 it is dynamic (useful for very large $message_headers, for example).
5685 6. Lookups now cache *every* query, not just the most recent. A new, separate
5686 store pool is used for this. It can be recovered when all lookup caches are
5687 flushed. Lookups now release memory at the end of their result strings.
5688 This has involved some general refactoring of the lookup sources.
5690 7. Some code has been added to the store_xxx() functions to reduce the amount
5691 of flapping under certain conditions.
5693 8. log_incoming_interface used to affect only the <= reception log lines. Now
5694 it causes the local interface and port to be added to several more SMTP log
5695 lines, for example "SMTP connection from", and rejection lines.
5697 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5699 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5701 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5702 could be overwritten at the end of the current message (or the start of a
5703 new message if it was set in a HELO ACL). The value is now preserved for
5704 the duration of the SMTP connection.
5706 12. If a transport had a headers_rewrite setting, and a matching header line
5707 contained an unqualified address, that address was qualified, even if it
5708 did not match any rewriting rules. The underlying bug was that the values
5709 of the flags that permit the existence of unqualified sender and recipient
5710 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5711 non-local messages, and by -bnq for local messages) were not being
5712 preserved with the message after it was received.
5714 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5715 "next input=" as part of the text comprising the host identity instead of
5716 the correct text. The code was using the same buffer for two different
5717 strings. However, depending on which order the printing function evaluated
5718 its arguments, the bug did not always show up. Under Linux, for example, my
5719 test suite worked just fine.
5721 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5722 doesn't work with some older versions of Perl. It has been changed to "my",
5723 which in any case is probably the better facility to use.
5725 15. A really picky compiler found some instances of statements for creating
5726 error messages that either had too many or two few arguments for the format
5729 16. The size of the buffer for calls to the DNS resolver has been increased
5730 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5731 for addresses that have a lot of PTR records. This alleviates a problem; it
5732 does not fully solve it.
5734 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5735 buffer now truncates the list and logs the incident, which is the same
5736 action as happens when Exim is looking up a host name and its aliases.
5737 Previously in this situation something unpredictable would happen;
5738 sometimes it was "internal error: store_reset failed".
5740 18. If a server dropped the connection unexpectedly when an Exim client was
5741 using GnuTLS and trying to read a response, the client delivery process
5742 crashed while trying to generate an error log message.
5744 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5745 string, for example:
5747 warn message = H1: something\nH2: something
5749 the text was added as a single header line from Exim's point of view
5750 though it ended up OK in the delivered message. However, searching for the
5751 second and subsequent header lines using $h_h2: did not work. This has been
5752 fixed. Similarly, if a system filter added multiple headers in this way,
5753 the routers could not see them.
5755 20. Expanded the error message when iplsearch is called with an invalid key to
5756 suggest using net-iplsearch in a host list.
5758 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5759 ACLs are no longer actually imposed (and a message to that effect is
5762 22. If a "gecos" field in a passwd entry contained escaped characters, in
5763 particular, if it contained a \" sequence, Exim got it wrong when building
5764 a From: or a Sender: header from that name. A second bug also caused
5765 incorrect handling when an unquoted " was present following a character
5766 that needed quoting.
5768 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5769 was not being matched caselessly.
5771 24. Arranged for all hyphens in the exim.8 source to be escaped with
5774 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5775 properly. Recipient callout cache records were still being keyed to include
5776 the sender, even when use_sender was set false. This led to far more
5777 callouts that were necessary. The sender is no longer included in the key
5778 when use_sender is false.
5780 26. Added "control = submission" modifier to ACLs.
5782 27. Added the ${base62d: operator to decode base 62 numbers.
5784 28. dnsdb lookups can now access SRV records.
5786 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5787 the configuration file.
5789 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5790 (-v) mode. This makes the output for a verbose queue run more intelligible.
5792 31. Added a use_postmaster feature to recipient callouts.
5794 32. Added the $body_zerocount variable, containing the number of binary zero
5795 bytes in the message body.
5797 33. The time of last modification of the "new" subdirectory is now used as the
5798 "mailbox time last read" when there is a quota error for a maildir
5801 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5803 35. Added +ignore_unknown as a special item in host lists.
5805 36. Code for decoding IPv6 addresses in host lists is now included, even if
5806 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5807 address was recognized as an IP address, but was then not correctly decoded
5808 into binary, causing unexpected and incorrect effects when compared with
5815 1. Very minor rewording of debugging text in manualroute to say "list of
5816 hosts" instead of "hostlist".
5818 2. If verify=header_syntax was set, and a header line with an unqualified
5819 address (no domain) and a large number of spaces between the end of the
5820 name and the colon was received, the reception process suffered a buffer
5821 overflow, and (when I tested it) crashed. This was caused by some obsolete
5822 code that should have been removed. The fix is to remove it!
5824 3. When running in the test harness, delay a bit after writing a bounce
5825 message to get a bit more predictability in the log output.
5827 4. Added a call to search_tidyup() just before forking a reception process. In
5828 theory, someone could use a lookup in the expansion of smtp_accept_max_
5829 per_host which, without the tidyup, could leave open a database connection.
5831 5. Added the variables $recipient_data and $sender_data which get set from a
5832 lookup success in an ACL "recipients" or "senders" condition, or a router
5833 "senders" option, similar to $domain_data and $local_part_data.
5835 6. Moved the writing of debug_print from before to after the "senders" test
5838 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5839 problems for message scanning, either using a data ACL, or using
5840 local_scan() because the Received: header was not generated till after they
5841 were called (in order to set the time as the time of reception completion).
5842 I have revised the way this works. The header is now generated after the
5843 body is received, but before the ACL or local_scan() are called. After they
5844 are run, the timestamp in the header is updated.
5850 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5851 before starting a queue runner without re-exec. This happened only when
5852 deliver_drop_privilege was set or when the Exim user was set to root. The
5853 effect of the bug was that timeouts during subsequent deliveries caused
5854 crashes instead of being properly handled. The handler is now left at its
5855 default (and expected) setting.
5857 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5858 message, again when deliver_drop_privilege is set or Exim is run as root.
5859 The bug described in (1) was not present in this case, but the tidying up
5860 of the other signals was missing. I have made the two cases consistent.
5862 3. The ignore_target_hosts setting on a manualroute router was being ignored
5863 for hosts that were looked up using the /MX notation.
5865 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5868 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5869 operated on the sender address. After changing the $sender_address to <>
5870 for the sender address verify, Exim was re-instated it as the original
5871 (before rewriting) address, but remembering that it had rewritten it, so it
5872 wasn't rewriting it again. This bug also had the effect of breaking the
5873 sender address verification caching when the sender address was rewritten.
5875 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5876 This has been changed so that if the ip literal address matches
5877 ignore_target_hosts, the router declines.
5879 7. Added expansion conditions match_domain, match_address, and match_local_
5880 part (NOT match_host).
5882 8. The placeholder for the Received: header didn't have a length field set.
5884 9. Added code to Exim itself and to exim_lock to test for a specific race
5885 condition that could lead to file corruption when using MBX delivery. The
5886 issue is with the lockfile that is created in /tmp. If this file is removed
5887 after a process has opened it but before that process has acquired a lock,
5888 there is the potential for a second process to recreate the file and also
5889 acquire a lock. This could lead to two Exim processes writing to the file
5890 at the same time. The added code performs the same test as UW imapd; it
5891 checks after acquiring the lock that its file descriptor still refers to
5892 the same named file.
5894 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5895 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5898 11. Added the smtp_active_hostname option. If used, this will typically be made
5899 to depend on the incoming interface address. Because $interface_address is
5900 not set up until the daemon has forked a reception process, error responses
5901 that can happen earlier (such as "too many connections") no longer contain
5904 12. If an expansion in a condition on a "warn" statement fails because a lookup
5905 defers, the "warn" statement is abandoned, and the next ACL statement is
5906 processed. Previously this caused the whole ACL to be aborted.
5908 13. Added the iplsearch lookup type.
5910 14. Added ident_timeout as a log selector.
5912 15. Added tls_certificate_verified as a log selector.
5914 16. Added a global option tls_require_ciphers (compare the smtp transport
5915 option of the same name). This controls incoming TLS connections.
5917 17. I finally figured out how to make tls_require_ciphers do a similar thing
5918 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5919 before starting the TLS session.
5921 18. Tabs are now shown as \t in -bP output.
5923 19. If the log selector return_path_on_delivery was set, Exim crashed when
5924 bouncing a message because it had too many Received: header lines.
5926 20. If two routers both had headers_remove settings, and the first one included
5927 a superfluous trailing colon, the final name in the first list and the
5928 first name in the second list were incorrectly joined into one item (with a
5929 colon in the middle).
5935 1. Added -C and -D options to the exinext utility, mainly to make it easier
5936 to include in the automated testing, but these could be helpful when
5937 multiple configurations are in use.
5939 2. The exinext utility was not formatting the output nicely when there was
5940 an alternate port involved in the retry record key, nor when there was a
5941 message id as well (for retries that were specific to a specific message
5942 and a specific host). It was also confused by IPv6 addresses, because of
5943 the additional colons they contain. I have fixed the IPv4 problem, and
5944 patched it up to do a reasonable job for IPv6.
5946 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5947 delivery, the log line now contains "pipelined" if PIPELINING was used.
5949 4. An SMTP transport process used to panic and die if the bind() call to set
5950 an explicit outgoing interface failed. This has been changed; it is now
5951 treated in the same way as a connect() failure.
5953 5. A reference to $sender_host_name in the part of a conditional expansion
5954 that was being skipped was still causing a DNS lookup. This no longer
5957 6. The def: expansion condition was not recognizing references to header lines
5958 that used bh_ and bheader_.
5960 7. Added the _cache feature to named lists.
5962 8. The code for checking quota_filecount in the appendfile transport was
5963 allowing one more file than it should have been.
5965 9. For compatibility with Sendmail, the command line option
5973 and sets the incoming protocol and host name (for trusted callers). The
5974 host name and its colon can be omitted when only the protocol is to be set.
5975 Note the Exim already has two private options, -pd and -ps, that refer to
5976 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5977 "s", but I don't think that's a major issue.
5979 10. A number of refactoring changes to the code, none of which should affect
5982 (a) The number of logging options was getting close to filling up the
5983 32-bit word that was used as a bit map. I have split them into two classes:
5984 those that are passed in the argument to log_write(), and those that are
5985 only ever tested independently outside of that function. These are now in
5986 separate 32-bit words, so there is plenty of room for expansion again.
5987 There is no change in the user interface or the logging behaviour.
5989 (b) When building, for example, log lines, the code previously used a
5990 macro that called string_cat() twice, in order to add two strings. This is
5991 not really sufficiently general. Furthermore, there was one instance where
5992 it was actually wrong because one of the argument was used twice, and in
5993 one call a function was used. (As it happened, calling the function twice
5994 did not affect the overall behaviour.) The macro has been replaced by a
5995 function that can join an arbitrary number of extra strings onto a growing
5998 (c) The code for expansion conditions now uses a table and a binary chop
5999 instead of a serial search (which was left over from when there were very
6000 few conditions). Also, it now recognizes conditions like "pam" even when
6001 the relevant support is not compiled in: a suitably worded error message is
6002 given if an attempt is made to use such a condition.
6004 11. Added ${time_interval:xxxxx}.
6006 12. A bug was causing one of the ddress fields not to be passed back correctly
6007 from remote delivery subprocesses. The field in question was not being
6008 subsequently used, so this caused to problems in practice.
6010 13. Added new log selectors queue_time and deliver_time.
6012 14. Might have fixed a bug in maildirsizefile handling that threw up
6013 "unexpected character" debug warnings, and recalculated the data
6014 unnecessarily. In any case, I expanded the warning message to give more
6017 15. Added the message "Restricted characters in address" to the statements in
6018 the default ACL that block characters like @ and % in local parts.
6020 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6021 Three changes have been made:
6023 (a) There was a serious bug; a negative response to MAIL caused the whole
6024 recipient domain to be cached as invalid, thereby blocking all messages
6025 to all local parts at the same domain, from all senders. This bug has
6026 been fixed. The domain is no longer cached after a negative response to
6027 MAIL if the sender used is not empty.
6029 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6032 (c) A new callout option, "use_sender" has been added for people who want
6033 the modified behaviour.
6039 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6042 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6043 indeed breaks things for older releases.
6045 3. Added additional logging to the case where there is a problem reading data
6046 from a filter that is running in a subprocess using a pipe, in order to
6047 try to track down a specific problem.
6049 4. Testing facility fudge: when running in the test harness and attempting
6050 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6051 getting "No route to host". Convert this to a timeout.
6053 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6056 6. Some OS don't have socklen_t but use size_t instead. This affects the
6057 fifth argument of getsockopt() amongst other things. This is now
6058 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6059 can be set for individual OS. I have set it for SunOS5, OSF1, and
6060 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6061 some earlier ones do not.
6063 7. Change 4.30/15 was not doing the test caselessly.
6065 8. The standard form for an IPv6 address literal was being rejected by address
6066 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6067 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6068 this, as well as the form without the "IPv6" on the front (but only when
6069 address literals are enabled, of course).
6071 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6073 10. Exim crashed if a message with an empty sender address specified by -f
6074 encountered a router with an errors_to setting. This could be provoked only
6075 by a command such as
6079 where an empty string was supplied; "<>" did not hit this bug.
6081 11. Installed PCRE release 4.5.
6083 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6084 remained set. It is now erased.
6086 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6087 times from message ids (which are base 36 rather than the normal 62).
6089 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6090 were being counted as actual protocol errors, and logged if the log
6091 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6092 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6093 and DATA following a set of rejected RCPTs do not count as protocol errors.
6094 In other words, Exim assumes they were pipelined, though this may not
6095 actually be the case. Of course, in all cases the client gets an
6096 appropriate error code.
6098 15. If a lookup fails in an ACL condition, a message about the failure may
6099 be available; it is used if testing the ACL cannot continue, because most
6100 such messages specify what the cause of the deferral is. However, some
6101 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6102 that caused an old message to be retained and used if a later statement
6103 caused a defer, replacing the real cause of the deferral.
6105 16. If an IP address had so many PTR records that the DNS lookup buffer
6106 was not large enough to hold them, Exim could crash while trying to process
6107 the truncated data. It now detects and logs this case.
6109 17. Further to 4.21/58, another change has been made: if (and only if) the
6110 first line of a message (the first header line) ends with CRLF, a bare LF
6111 in a subsequent header line has a space inserted after it, so as not to
6112 terminate the header.
6114 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6115 unnecessarily, used atoi() instead of strtol(), and didn't check the
6116 termination when getting file sizes from file names by regex.
6118 19. Completely re-implemented the support for maildirsize files, in the light
6119 of a number of problems with the previous contributed implementation
6120 (4.30/29). In particular:
6122 . If the quota is zero, the maildirsize file is maintained, but no quota is
6125 . If the maildir directory does not exist, it is created before any attempt
6126 to write a maildirsize file.
6128 . The quota value in the file is just a cache; if the quota is changed in
6129 the transport, the new value overrides.
6131 . A regular expression is available for excluding directories from the
6134 20. The autoreply transport checks the characters in options that define the
6135 message's headers; it allows continued headers, but it was checking with
6136 isspace() after an embedded newline instead of explicitly looking for a
6139 21. If all the "regular" hosts to which an address was routed had passed their
6140 expiry times, and had not reached their retry times, the address was
6141 bounced, even if fallback hosts were defined. Now Exim should go on to try
6144 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6145 equivalent code in the SMTP transport. Some hosts send humungous responses
6146 to HELO/EHLO, more than 1024 it seems.
6148 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6149 gives compiler warnings in some environments. I've now done it "properly",
6152 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6153 (because of problems with the built-in one) was declared to return uschar *
6154 instead of char *, causing compiler failure.
6156 25. Fixed a file descriptor leak when processing alias/forward files.
6158 26. Fixed a minor format string issue in dbfn.c.
6160 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6162 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6163 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6166 29. When a sender address is verified, it is cached, to save repeating the test
6167 when there is more than one recipient in a message. However, when the
6168 verification involves a callout, it is possible for different callout
6169 options to be set for different recipients. It is too complicated to keep
6170 track of this in the cache, so now Exim always runs a verification when a
6171 callout is required, relying on the callout cache for the optimization.
6172 The overhead is duplication of the address routing, but this should not be
6175 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6176 to be verified with callout=postmaster, and the main callout worked but the
6177 postmaster check failed, the verification correctly failed. However, if a
6178 subsequent RCPT command asked for sender verification *without* the
6179 postmaster check, incorrect caching caused this verification also to fail,
6182 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6183 it was not caching the DNS options (qualify_single, search_parents) that
6184 were used when the lookup failed. A subsequent lookup with different
6185 options therefore always gave the same answer, though there were cases
6186 where it should not have. (Example: a "domains = !$mx_any" option on a
6187 dnslookup router: the "domains" option is always processed without any
6188 widening, but the router might have qualify_single set.) Now Exim uses the
6189 cached value only when the same options are set.
6191 32. Added John Jetmore's "exipick" utility to the distribution.
6193 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6194 than a timeout (e.g. a certificate is required, and is not provided), an
6195 Exim server now closes the connection immediately. Previously it waited for
6196 the client to close - but if the client is SSL, it seems that they each
6197 wait for each other, leading to a delay before one of them times out.
6199 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6200 maintained 0.8.x compatibility because I don't think many are using it, and
6201 it is clearly obsolete.
6203 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6206 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6207 client certificate was expired. A simple patch fixes this, though I don't
6208 understand the full logic of why the verify callback is called multiple
6211 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6212 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6213 which causes problems with some clients (such as the Certicom SSL Plus
6214 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6215 disables the coutermeasure allowing Eudora to connect."
6217 38. Exim was not checking that a write() to a log file succeeded. This could
6218 lead to Bad Things if a log got too big, in particular if it hit a file
6219 size limit. Exim now panics and dies if it cannot write to a log file, just
6220 as it does if it cannot open a log file.
6222 39. Modified OS/Makefile-Linux so that it now contains
6224 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6226 The two -D definitions ensure that Exim is compiled with large file
6227 support, which makes it possible to handle log files that are bigger than
6230 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6231 instance) a domain was checked against a named list that involved a lookup,
6232 causing $domain_data to be set, then another domain was checked against the
6233 same list, then the first domain was re-checked, the value of $domain_data
6234 after the final check could be wrong. In particular, if the second check
6235 failed, it could be set empty. This bug probably also applied to
6238 41. The strip_trailing_dot option was not being applied to the address given
6239 with the -f command-line option.
6241 42. The code for reading a message's header from the spool was incrementing
6242 $received_count, but never initializing it. This meant that the value was
6243 incorrect (doubled) while delivering a message in the same process in which
6244 it was received. In the most common configuration of Exim, this never
6245 happens - a fresh exec is done - but it can happen when
6246 deliver_drop_privilege is set.
6248 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6249 it now includes up to 150 characters of the unexpected data in the log
6252 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6253 and building data strings. The size of both of these buffers was 10 000
6254 bytes - far larger than anybody would *ever* want, thought I. Needless to
6255 say, somebody hit the limit. I have increased the maximum line length to
6256 20 000 and the maximum data length of concatenated lines to 100 000. I have
6257 also fixed two bugs, because there was no checking on these buffers. Tsk,
6258 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6259 buffer is too small.
6261 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6262 lsearch lookups. Now it does.
6264 46. When parsing a route_list item in a manualroute router, a fixed-length
6265 buffer was used for the list of hosts. I made this 1024 bytes long,
6266 thinking that nobody would ever have a list of hosts that long. Wrong.
6267 Somebody had a whole pile of complicated expansion conditions, and the
6268 string was silently truncated, leading to an expansion error. It turns out
6269 that it is easier to change to an unlimited length (owing to other changes
6270 that have happened since this code was originally written) than to build
6271 structure for giving a limitation error. The length of the item that
6272 expands into the list of hosts is now unlimited.
6274 47. The lsearch lookup could not handle data where the length of text line was
6275 more than 4095 characters. Such lines were truncated, leading to shortened
6276 data being returned. It should now handle lines of any length.
6278 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6279 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6282 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6283 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6284 no longer happens in comment lines. A list of replacements is now placed
6285 at the head of all of the source files, except those whose only change is
6286 to replace PERL_COMMAND in the very first #! line.
6288 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6289 messages on the queue, with a bottom-up merge sort, using code contributed
6290 by Michael Haardt. This should make operations like -bp somewhat faster on
6291 large queues. It won't affect queue runners, except when queue_run_in_order
6294 51. Installed eximstats 1.31 in the distribution.
6296 52. Added support for SRV lookups to the dnslookup router.
6298 53. If an ACL referred to $message_body or $message_body_end, the value was not
6299 reset for any messages that followed in the same SMTP session.
6301 54. The store-handling optimization for building very long strings was not
6302 differentiating between the different store pools. I don't think this
6303 actually made any difference in practice, but I've tidied it.
6305 55. While running the routers to verify a sender address, $sender_address
6306 was still set to the sender address. This is wrong, because when routing to
6307 send a bounce to the sender, it would be empty. Therefore, I have changed
6308 it so that, while verifying a sender address, $sender_address is set to <>.
6309 (There is no change to what happens when verifying a recipient address.)
6311 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6312 target A or AAAA records (if not already returned) without resetting the
6313 qualify_single or search_parents options of the DNS resolver. These are
6314 inappropriate in this case because the targets of MX and SRV records must
6315 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6316 target that, when qualified, matched something in the local domain. These
6317 two options are now turned off when doing these lookups.
6319 57. It seems that at least some releases of Reiserfs (which does not have the
6320 concept of a fixed number of inodes) returns zero and not -1 for the
6321 number of available inodes. This interacted badly with check_spool_inodes,
6322 which assumed that -1 was the "no such thing" setting. What I have done is
6323 to check that the total number of inodes is greater than zero before doing
6324 the test of how many are available.
6326 58. When a "warn" ACL statement has a log_message modifier, the message is
6327 remembered, and not repeated. This is to avoid a lot of repetition when a
6328 message has many recipients that cause the same warning to be written.
6329 However, Exim was preserving the list of already written lines for an
6330 entire SMTP session, which doesn't seem right. The memory is now reset if a
6331 new message is started.
6333 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6334 debugging output unless log_rewrite was also set.
6336 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6337 of (int)(handle) when we know that handle contains (void *)(-1).
6339 61. The Exim daemon panic-logs an error return when it closes the incoming
6340 connection. However "connection reset by peer" seems to be common, and
6341 isn't really an error worthy of noting specially, so that particular error
6344 62. When Exim is trying to find all the local interfaces, it used to panic and
6345 die if the ioctl to get the interface flags failed. However, it seems that
6346 on at least one OS (Solaris 9) it is possible to have an interface that is
6347 included in the list of interfaces, but for which you get a failure error
6348 for this call. This happens when the interface is not "plumbed" into a
6349 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6350 failure of the "get flags" call assumes that the interface is down.
6352 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6353 makes life easier for people who are doing arithmetic on fields extracted
6354 from dates, where you often get leading zeros that should not be
6355 interpreted as octal.
6357 64. Added qualify_domain to the redirect router, to override the global
6360 65. If a pathologically long header line contained very many addresses (the
6361 report of this problem mentioned 10 000) and each of them was rewritten,
6362 Exim could use up a very large amount of memory. (It kept on making new
6363 copies of the header line as it rewrote, and never released the old ones.)
6364 At the expense of a bit more processing, the header rewriting function has
6365 been changed so that it no longer eats memory in this way.
6367 66. The generation of the Received: header has been moved from the time that a
6368 message starts to be received, to the time that it finishes. The timestamp
6369 in the Received: header should now be very close to that of the <= log
6370 line. There are two side-effects of this change:
6372 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6373 logged header lines no longer include the local Received: line, because
6374 it has not yet been created. The same applies to a copy of the message
6375 that is returned to a non-SMTP sender when a message is rejected.
6377 (b) When a filter file is tested using -bf, no additional Received: header
6378 is added to the test message. After some thought, I decided that this
6381 This change does not affect the value of $received_for. It is still set
6382 after address rewriting, but before local_scan() is called.
6384 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6386 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6387 gave an unhelpful panic error message, and a defer error. I have managed to
6388 change this behaviour so that it now rejects any supplied certificate,
6389 which seems right, as the list of acceptable certificates is empty.
6391 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6392 gave an unhelpful defer error. I have not managed to make this reject any
6393 supplied certificates, but the error message it gives is "no certificate
6394 supplied", which is not helpful.
6396 70. exigrep's output now also includes lines that are not associated with any
6397 message, but which match the given pattern. Implemented by a patch from
6398 Martin Sluka, which also tidied up the Perl a bit.
6400 71. Recipient callout verification, like sender verification, was using <> in
6401 the MAIL FROM command. This isn't really the right thing, since the actual
6402 sender may affect whether the remote host accepts the recipient or not. I
6403 have changed it to use the actual sender in the callout; this means that
6404 the cache record is now keyed on a recipient/sender pair, not just the
6405 recipient address. There doesn't seem to be a real danger of callout loops,
6406 since a callout by the remote host to check the sender would use <>.
6407 [SEE ABOVE: changed after hitting problems.]
6409 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6410 temporary errors. However, in the case of such a code being given after
6411 the end of a data transmission (i.e. after ".") Exim was failing to write
6412 a retry record for the message. (Yes, there was some broken host that was
6413 actually sending 8xx at this point.)
6415 73. An unknown lookup type in a host list could cause Exim to panic-die when
6416 the list was checked. (An example that provoked this was putting <; in the
6417 middle of a list instead of at the start.) If this happened during a DATA
6418 ACL check, a -D file could be left lying around. This kind of configuration
6419 error no longer causes Exim to die; instead it causes a defer error. The
6420 incident is still logged to the main and panic logs.
6422 74. Buglet left over from Exim 3 conversion. The message "too many messages
6423 in one connection" was written to the rejectlog but not the mainlog, except
6424 when address rewriting (yes!) was being logged.
6426 75. Added write_rejectlog option.
6428 76. When a system filter was run not as root (that is, when system_filter_user
6429 was set), the values of the $n variables were not being returned to the
6430 main process; thus, they were not subsequently available in the $sn
6433 77. Added +return_path_on_delivery log selector.
6435 78. A connection timeout was being treated differently from recipients deferred
6436 when testing hosts_max_try with a message that was older than the host's
6437 retry timeout. (The host should not be counted, thus allowing all hosts to
6438 be tried at least once before bouncing.) This may have been the cause of an
6439 occasionally reported bug whereby a message would remain on the queue
6440 longer than the retry timeout, but would be bounced if a delivery was
6441 forced. I say "may" because I never totally pinned down the problem;
6442 setting up timeout/retry tests is difficult. See also the next item.
6444 79. The ultimate address timeout was not being applied to errors that involved
6445 a combination of host plus message (for example, a timeout on a MAIL
6446 command). When an address resolved to a number of possible hosts, and they
6447 were not all tried for each delivery (e.g. because of hosts_max_try), a
6448 message could remain on the queue longer than the retry timeout.
6450 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6453 81. Fixed an obscure SMTP outgoing bug which required at least the following
6454 conditions: (a) there was another message waiting for the same server;
6455 (b) the server returned 5xx to all RCPT commands in the first message so
6456 that the message was not completed; (c) the server dropped the connection
6457 or gave a negative response to the RSET that Exim sends to abort the
6458 transaction. The observed case was a dropped connection after DATA that had
6459 been sent in pipelining mode. That is, the server had advertised PIPELINING
6460 but was not implementing it correctly. The effect of the bug was incorrect
6461 behaviour, such as trying another host, and this could lead to a crash.
6467 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6468 and daemon.c were passed as pointers to ints; they should have been
6469 pointers to socklen_t variables (which are typically unsigned ints).
6471 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6474 3. Fixed a really odd bug that affected only the testing scheme; patching a
6475 certain fixed string in the binary changed the value of another string that
6476 happened to be identical to the end of the original first string.
6478 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6479 name", it returns that address as the IP address. On some operating
6480 systems (e.g. Solaris), it also passes back the IP address string as the
6481 "host name". However, on others (e.g. Linux), it passes back an empty
6482 string. Exim wasn't checking for this, and was changing the host name to an
6483 empty string, assuming it had been canonicalized.
6485 5. Although rare, it is permitted to have more than one PTR record for a given
6486 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6487 all the names associated with an address, because they do in Solaris.
6488 However, it seems that they do not in Linux for data that comes from the
6489 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6490 I found this out when I moved to a new Linux workstation and tried to run
6491 the Exim test suite.
6493 To get round this problem I have changed the code so that it now does its
6494 own call to the DNS to look up PTR records when searching for a host name.
6495 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6496 addresses that are only in /etc/hosts are still found.
6498 This behaviour is, however, controlled by an option called host_lookup_
6499 order, which defaults to "bydns:byaddr". If people want to use the other
6500 order, or indeed, just use one or the other means of lookup, they can
6501 specify it in this variable.
6503 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6504 some operating systems, this comes back from gethostbyaddr() as an empty
6505 string, and this is what Exim used to test for. However, it seems that in
6506 other systems, "." is yielded. Exim now tests for this case too.
6508 7. The values of check_spool_space and check_log_space are now held internally
6509 as a number of kilobytes instead of an absolute number of bytes. If a
6510 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6511 kilobyte. This means that much larger values can be stored.
6513 8. Exim monitor: an attempt to get the action menu when not actually pointing
6514 at a message produces an empty menu entitled "No message selected". This
6515 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6516 no entries in it ("Shell widget menu has zero width and/or height"). So I
6517 have added a single, blank menu entry in this case.
6519 9. Added ${quote_local_part.
6521 10. MIME decoding is now applied to the contents of Subject: header lines when
6524 11. Now that a reference to $sender_host_address automatically causes a reverse
6525 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6526 host lookup before query-style lookups in lists that might use this
6527 variable. This has therefore been abolished, and the "net-" prefix is no
6528 longer necessary for query-style lookups.
6530 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6531 have been a typo for LFLAGS, so it has been changed.
6533 13. The install script calls Exim with "-C /dev/null" in order to find the
6534 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6535 to be output. However, since Exim outputs its version number before the
6536 error, it didn't break the script. It just looked ugly. I fixed this by
6537 always allowing "-C /dev/null" if the caller is root.
6539 14. Ignore overlarge ACL variable number when reading spool file - insurance
6540 against a later release with more variables having written the file.
6542 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6543 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6544 this, as well as the form without the "IPv6" on the front.
6546 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6547 OS/Makefile-Darwin file.
6549 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6550 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6552 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6553 variable. However, some OS use ints for inodes. I've added cast to long int
6554 to get rid of the compiler warning.
6556 19. I had forgotten to lock out "/../" in configuration file names when
6557 ALT_CONFIG_PREFIX was set.
6559 20. Routers used for verification do not need to specify transports. However,
6560 if such a router generated a host list, and callout was configured, Exim
6561 crashed, because it could not find a port number from the (non-existent)
6562 transport. It now assumes port 25 in this circumstance.
6564 21. Added the -t option to exigrep.
6566 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6567 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6568 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6569 them is set, LOOKUP_LSEARCH is forced.
6571 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6573 24. Added sender and host information to the "rejected by local_scan()" log
6574 line; previously there was no indication of these.
6576 25. Added .include_if_exists.
6578 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6579 Linux. It turns out that not all file systems support this. Apparently some
6580 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6581 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6582 not supported on the file descriptor, is now ignored when Exim is trying to
6583 sync a directory. This applies only to Linux.
6585 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6587 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6588 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6591 29. Added support for maildirsize files from supplied patch (modified a bit).
6593 30. The use of :fail: followed by an empty string could lead Exim to respond to
6594 sender verification failures with (e.g.):
6596 550 Verification failed for <xxx>
6597 550 Sender verify failed
6599 where the first response line was missing the '-' that indicates it is not
6600 the final line of the response.
6602 31. The loop for finding the name of the user that called Exim had a hardwired
6603 limit of 10; it now uses the value of finduser_retries, which is used for
6604 all other user lookups.
6606 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6609 33. Exim was neglecting to zero errno before one call of strtol() when
6610 expanding a string and expecting an integer value. On some systems this
6611 resulted in spurious "integer overflow" errors. Also, it was casting the
6612 result into an int without checking.
6614 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6615 did not work. The code looks as if it has *never* worked, though it appears
6616 to have been documented since at least release 1.62. I have made it work.
6618 35. The "timeout_DNS" error in retry rules, also documented since at least
6619 1.62, also never worked. As it isn't clear exactly what this means, and
6620 clearly it isn't a major issue, I have abolished the feature by treating it
6621 as "timeout", and writing a warning to the main and panic logs.
6623 36. The display of retry rules for -brt wasn't always showing the error code
6626 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6627 timeout_connect_A, timeout_connect_MX.
6629 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6630 to the empty sender.
6632 39. The daemon was not analysing the content of -oX till after it had closed
6633 stderr and disconnected from the controlling terminal. This meant that any
6634 syntax errors were only noted on the panic log, and the return code from
6635 the command was 0. By re-arranging the code a little, I've made the
6636 decoding happen first, so such errors now appear on stderr, and the return
6637 code is 1. However, the actual setting up of the sockets still happens in
6638 the disconnected process, so errors there are still only recorded on the
6641 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6642 connections (as happens on some IP stacks) was logged at start up time as
6643 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6644 it from "IPv6 and IPv4", which means that two separate sockets are being
6647 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6648 says whether AF_INET or AF_INET6 was passed as an argument.
6650 42. Exiwhat output was messed up when time zones were included in log
6653 43. Exiwhat now gives more information about the daemon's listening ports,
6654 and whether -tls-on-connect was used.
6656 44. The "port" option of the smtp transport is now expanded.
6658 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6659 silently ignored. Now an error message is written to the main and panic
6662 46. There's a new ACL modifier called "logwrite" which writes to a log file
6663 as soon as it is encountered.
6665 47. Added $local_user_uid and $local_user_gid at routing time.
6667 48. Exim crashed when trying to verify a sender address that was being
6670 49. Exim was recognizing only a space character after ".include". It now also
6671 recognizes a tab character.
6673 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6674 extracting the relevant information from the specification. The man page no
6675 longer contains scrambled data for the -d option, and I've added a section
6676 at the front about calling Exim under different names.
6678 51. Added "extra_headers" argument to the "mail" command in filter files.
6680 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6683 53. Installed eximstats 1.29.
6685 54. Added transport_filter_timeout as a generic transport option.
6687 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6688 Cc: header lines. This was required by RFC 822, but it not required by RFC
6691 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6692 incoming messages that did not have them. Now it does so only if the
6693 message originates locally, that is, if there is no associated remote host
6694 address. When Resent- header lines are present, this applies to the Resent-
6695 lines rather than the non-Resent- lines.
6697 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6698 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6700 58. Messages for configuration errors now include the name of the main
6701 configuration file - useful now that there may be more than one file in a
6702 list (.included file names were always shown).
6704 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6705 for those rare installations that do not start the daemon as root or run it
6706 setuid root. I've cut out the call to initgroups() if the daemon is not
6709 60. The Exim user and group can now be bound into the binary as text strings
6710 that are looked up at the start of Exim's processing.
6712 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6714 62. Added $mailstore_basename variable.
6716 63. Installed patch to sieve.c from Michael Haardt.
6718 64. When Exim failed to open the panic log after failing to open the main log,
6719 the original message it was trying to log was written to stderr and debug
6720 output, but if they were not available (the usual case in production), it
6721 was lost. Now it is written to syslog before the two lines that record the
6722 failures to open the logs.
6724 65. Users' Exim filters run in subprocesses under the user's uid. It is
6725 possible for a "deliver" command or an alias in a "personal" command to
6726 provoke an address rewrite. If logging of address rewriting is configured,
6727 this fails because the process is not running as root or exim. There may be
6728 a better way of dealing with this, but for the moment (because 4.30 needs
6729 to be released), I have disabled address rewrite logging when running a
6730 filter in a non-root, non-exim process.
6736 1. The buildconfig auxiliary program wasn't quoting the value set for
6737 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6738 not defined. This bug was masked in 4.22 by the effect that was fixed in
6741 2. Some messages that were rejected after a message id was allocated were
6742 shown as "incomplete" by exigrep. It no longer does this for messages that
6743 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6745 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6746 have allow_domain_literals set, the ID did not get logged in the <= line.
6747 Domain literals are now always recognized in Message-ID: header lines.
6749 4. The first argument for a ${extract expansion item is the key name or field
6750 number. Leading and trailing spaces in this item were not being ignored,
6751 causing some misleading effects.
6753 5. When deliver_drop_privilege was set, single queue runner processes started
6754 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6755 same command in the process it spins off) were not dropping privilege.
6757 6. When the daemon running as "exim" started a queue runner, it always
6758 re-executed Exim in the spun-off process. This is a waste of effort when
6759 deliver_drop_privilege is set. The new process now just calls the
6760 queue-runner function directly.
6766 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6769 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6770 ignored. Though the use of -oP was forcing the writing of a pid file, it
6771 was always written to the default place.
6773 3. If the message "no IP address found for host xxxx" is generated during
6774 incoming verification, it is now followed by identification of the incoming
6775 connection (so you can more easily find what provoked it).
6777 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6779 5. Added some features to "harden" Exim a bit more against certain attacks:
6781 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6782 be put in Local/Makefile. This is like the never_users runtime option,
6783 but it cannot be overridden. The default setting is "root".
6785 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6786 prefix string with which any file named in a -C command line option
6789 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6790 is retained for -C and -D only if the caller of Exim is root. Without
6791 it, the exim user may also use -C and -D and retain privilege.
6793 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6794 command line option is disabled.
6796 6. Macro names set by the -D option must start with an upper case letter, just
6797 like macro names defined in the configuration file.
6799 7. Added "dereference=" facility to LDAP.
6801 8. Two instances of the typo "uknown" in the source files are fixed.
6803 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6804 the Configure-Makefile script screwed up while processing it.
6806 10. Incorporated PCRE 4.4.
6808 11. The SMTP synchronization check was not operating right at the start of an
6809 SMTP session. For example, it could not catch a HELO sent before the client
6810 waited for the greeting. There is now a check for outstanding input at the
6811 point when the greeting is written. Because of the duplex, asynchronous
6812 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6813 way, but not yet received, when the check is performed.
6815 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6816 on TCP/IP sockets, because this apparently causes some broken clients to
6819 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6820 unchanged) from the Cygwin maintainer.
6822 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6823 was supported instead of testing for mbx. Effectively a typo.
6825 15. The spa authenticator server code was not checking that the input it
6826 received was valid base64.
6828 16. The debug output line for the "set" modifier in ACLs was not showing the
6829 name of the variable that was being set.
6831 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6833 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6834 lookup on its own. Something else was needed to trigger the lookup. For
6835 example, a match in host_lookup or the need for a host name in a host list.
6836 Now, if $sender_host_name is referenced and the host name has not yet been
6837 looked up, a lookup is performed. If the lookup fails, the variable remains
6838 empty, and $host_lookup_failed is set to "1".
6840 19. Added "eqi" as a case-independent comparison operator.
6842 20. The saslauthd authentication condition could segfault if neither service
6843 nor realm was specified.
6845 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6846 error message that was logged was misleading, and incoming SMTP
6847 connections were dropped. The message is now more accurate, and temporary
6848 errors are given to SMTP connections.
6850 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6851 (except RSET) with a 421 error, until QUIT is received. However, it was
6852 failing to send a response to QUIT.
6854 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6855 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6856 ACL is now run after the helo_try_verify_hosts code.
6858 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6859 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6860 case-independent, so other case variants are also recognized). Apparently
6861 some systems use these upper case variants.
6863 25. If more than two messages were waiting for the same host, and a transport
6864 filter was specified for the transport, Exim sent two messages over the
6865 same TCP/IP connection, and then failed with "socket operation on non-
6866 socket" when it tried to send the third.
6868 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6870 27. The extern definition of crypt16() in expand.c was not being excluded when
6871 the OS had its own crypt16() function.
6873 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6874 as a preferred synonym for return_size_limit, both as an option and as an
6877 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6879 30. Changed the default configuration ACL to relax the local part checking rule
6880 for addresses that are not in any local domains. For these addresses,
6881 slashes and pipe symbols are allowed within local parts, but the sequence
6882 /../ is explicitly forbidden.
6884 31. SPA server authentication was not clearing the challenge buffer before
6887 32. log_message in a "warn" ACL statement was writing to the reject log as
6888 well as to the main log, which contradicts the documentation and doesn't
6889 seem right (because no rejection is happening). So I have stopped it.
6891 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6892 However, I am unable to do any testing of this.
6894 34. Fixed an infelicity in the appendfile transport. When checking directories
6895 for a mailbox, to see if any needed to be created, it was accidentally
6896 using path names with one or more superfluous leading slashes; tracing
6897 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6899 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6900 contents are added to the log line that is written for every discarded
6901 recipient. (Previously a log_message setting was ignored.)
6903 36. The ${quote: operator now quotes the string if it is empty.
6905 37. The install script runs exim in order to find its version number. If for
6906 some reason other than non-existence or emptiness, which it checks, it
6907 could not run './exim', it was installing it with an empty version number,
6908 i.e. as "exim-". This error state is now caught, and the installation is
6911 38. An argument was missing from the function that creates an error message
6912 when Exim fails to connect to the socket for saslauthd authentication.
6913 This could cause Exim to crash, or give a corrupted message.
6915 39. Added isip, isip4, and isip6 to ${if conditions.
6917 40. The ACL variables $acl_xx are now saved with the message, and can be
6918 accessed later in routers, transports, and filters.
6920 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6921 strings in the file are not string-expanded.
6923 42. If a MAIL command specified a SIZE value that was too large to fit into an
6924 int variable, the check against message_size_limit failed. Such values are
6925 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6926 day this will have to be increased, but I don't think I want to be around
6927 when emails are that large.
6934 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6935 iconv() is not standard in FreeBSD.
6937 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6938 IPv6 enabled. The observed symptom was a segmentation fault on return from
6939 the function os_common_find_running_interfaces() in src/os.c.
6941 3. In the check_special_case() function in daemon.c I had used "errno" as an
6942 argument name, which causes warnings on some systems. This was basically a
6943 typo, since it was named "eno" in the comments!
6945 4. The code that waits for the clock to tick (at a resolution of some fraction
6946 of a second) so as to ensure message-id uniqueness was always waiting for
6947 at least one whole tick, when it could have waited for less. [This is
6948 almost certainly not relevant at current processor speeds, where it is
6949 unlikely to ever wait at all. But we try to future-proof.]
6951 5. The function that sleeps for a time interval that includes fractions of a
6952 second contained a race. It did not block SIGALRM between setting the
6953 timer, and suspending (a couple of lines later). If the interval was short
6954 and the sigsuspend() was delayed until after it had expired, the suspension
6955 never ended. On busy systems this could lead to processes getting stuck for
6958 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6959 process, before it forks any delivery processes. The open lookup caching
6960 mechanism meant that the open file or database connection was passed into
6961 the delivery process. The problem was that delivery processes always tidy
6962 up cached lookup data. This could cause a problem for the next delivery
6963 process started by the queue runner, because the external queue runner
6964 process does not know about the closure. So the next delivery process
6965 still has data in the lookup cache. In the case of a file lookup, there was
6966 no problem because closing a file descriptor in a subprocess doesn't affect
6967 the parent. However, if the lookup was caching a connection to a database,
6968 the connection was closed, and the second delivery process was likely to
6969 see errors such as "PGSQL: query failed: server closed the connection
6970 unexpectedly". The problem has been fixed by closing all cached lookups
6971 in a queue runner before running a delivery process.
6973 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6974 seem to have the "const" qualifier which it has on other OS. I've
6977 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6978 *of the same type* (client or server) with the same public name that an
6979 error should be diagnosed.
6981 9. When Exim looked up a host name for an IP address, but failed to find the
6982 original IP address when looking up the host name (a safety check), it
6983 output the message "<ip address> does not match any IP for NULL", which was
6984 confusing, to say the least. The bug was that the host name should have
6985 appeared instead of "NULL".
6987 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6988 user that is built into the binary, and the -C or -D options is used, root
6989 privilege is dropped before the configuration file is read. In addition,
6990 logging is switched to stderr instead of the normal log files. If the
6991 configuration then re-defines the Exim user, the unprivileged environment
6992 is probably not what is expected, so Exim logs a panic warning message (but
6995 However, if deliver_drop_privilege is set, the unprivileged state may well
6996 be exactly what is intended, so the warning has been cut out in that case,
6997 and Exim is allowed to try to write to its normal log files.
7003 1. smtp_return_error_details was not giving details for temporary sender
7004 or receiver verification errors.
7006 2. Diagnose a configuration error if two authenticators have the same public
7009 3. Exim used not to create the message log file for a message until the first
7010 delivery attempt. This could be confusing when incoming messages were held
7011 for policy or load reasons. The message log file is now created at the time
7012 the message is received, and an initial "Received" line is written to it.
7014 4. The automatically generated man page for command line options had a minor
7015 bug that caused no ill effects; however, a more serious problem was that
7016 the procedure for building the man page automatically didn't always
7017 operate. Consequently, release 4.20 contains an out-of-date version. This
7018 shouldn't happen again.
7020 5. When building Exim with embedded Perl support, the script that builds the
7021 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7022 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7024 6. The freeze_tell option was not being used for messages that were frozen on
7025 arrival, either by an ACL or by local_scan().
7027 7. Added the smtp_incomplete_transaction log selector.
7029 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7030 was accepting AUTH without a new EHLO.
7032 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7033 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7036 10. Logging of TCP/IP connections (when configured) now happens in the main
7037 daemon process instead of the child process, so that the TCP/IP connection
7038 count is more accurate (but it can never be perfect).
7040 11. The use of "drop" in a nested ACL was not being handled correctly in the
7041 outer ACL. Now, if condition failure induced by the nested "drop" causes
7042 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7043 or "require"), the connection is dropped.
7045 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7046 that yield "discard" can now be used with an "accept" or a "discard" verb,
7047 but an error is generated for any others (because I can't see a useful way
7048 to define what should happen).
7050 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7051 are now processed in the same way as lines in the Exim configuration file.
7052 In particular, continuation lines are supported.
7054 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7056 15. Added -ti meaning -t -i.
7058 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7059 domains, and warn by logging if others are found.
7061 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7062 returned from GIFCONF when Exim is trying to find the list of interfaces on
7063 a host. The code in os.c has been modified to copy each ifreq to an aligned
7064 structure in all cases.
7066 Also, in some cases, the returned ifreq's were being copied to a 'struct
7067 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7068 means the last couple of bytes of an IPv6 address could be chopped if the
7069 ifreq contained only a normal sockaddr (14 bytes storage).
7071 18. Named domain lists were not supported in the hosts_treat_as_local option.
7072 An entry such as +xxxx was not recognized, and was treated as a literal
7075 19. Ensure that header lines added by a DATA ACL are included in the reject log
7076 if the ACL subsequently rejects the message.
7078 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7079 MD5 (which is deprecated).
7081 21. When testing a filter file using -bf, Exim was writing a message when it
7082 took the sender from a "From " line in the message, but it was not doing so
7083 when it took $return_path from a Return-Path: header line. It now does.
7085 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7086 with a valid header line field name (a series of printing characters
7087 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7089 23. Changed "disc" in the source to "disk" to conform to the documentation and
7090 the book and for uniformity.
7092 24. Ignore Sendmail's -Ooption=value command line item.
7094 25. When execve() failed while trying to run a command in a pipe transport,
7095 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7096 could be confused with a return value of 69 from the command itself. This
7097 has been changed to 127, the value the shell returns if it is asked to run
7098 a non-existent command. The wording for the related log line suggests a
7099 non-existent command as the problem.
7101 26. If received_header_text expands to an empty string, do not add a Received:
7102 header line to the message. (Well, it adds a token one on the spool, but
7103 marks it "old" so that it doesn't get used or transmitted.)
7105 27. Installed eximstats 1.28 (addition of -nt option).
7107 28. There was no check for failure on the call to getsockname() in the daemon
7108 code. This can fail if there is a shortage of resources on the system, with
7109 ENOMEM, for example. A temporary error is now given on failure.
7111 29. Contrary to the C standard, it seems that in some environments, the
7112 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7113 program. Exim now does this explicitly; it affects the formatting of
7114 timestamps using strftime().
7116 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7117 complaints. I've now initialized all the variables, to avoid this.
7119 32. Header lines added by a system filter were not being "seen" during
7120 transport-time rewrites.
7122 33. The info_callback() function passed to OpenSSL is set up with type void
7123 (*)(SSL *, int, int), as described somewhere. However, when calling the
7124 function (actually a macro) that sets it up, the type void(*)() is
7125 expected. I've put in a cast to prevent warnings from picky compilers.
7127 34. If a DNS black list lookup found a CNAME record, but there were no A
7128 records associated with the domain it pointed at, Exim crashed.
7130 35. If a DNS black list lookup returned more than one A record, Exim ignored
7131 all but the first. It now scans all returned addresses if a particular IP
7132 value is being sought. In this situation, the contents of the
7133 $dnslist_value variable are a list of all the addresses, separated by a
7136 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7137 to accept a host name and all its aliases if the forward lookup for any of
7138 them yielded the IP address of the incoming connection. Now it accepts only
7139 those names whose forward lookup yields the correct IP address. Any other
7140 names are discarded. This closes a loophole whereby a rogue DNS
7141 administrator could create reverse DNS records to break through a
7142 wildcarded host restriction in an ACL.
7144 37. If a user filter or a system filter that ran in a subprocess used any of
7145 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7146 the wrong values were passed to the pipe command ($thisaddress had the
7147 value of $0, $0 had the value of $1, etc). This bug was introduced by
7148 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7150 38. Improved the line breaking for long SMTP error messages from ACLs.
7151 Previously, if there was no break point between 40 and 75 characters, Exim
7152 left the rest of the message alone. Two changes have been made: (a) I've
7153 reduced the minimum length to 35 characters; (b) if it can't find a break
7154 point between 35 and 75 characters, it looks ahead and uses the first one
7155 that it finds. This may give the occasional overlong line, but at least the
7156 remaining text gets split now.
7158 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7159 file descriptors might be low, and that setting 1000 would always raise it.
7160 It turns out that in some environments, the limit is already over 1000 and
7161 that lowering it causes trouble. So now Exim takes care not to decrease it.
7163 40. When delivering a message, the value of $return_path is set to $sender_
7164 address at the start of routing (routers may change the value). By an
7165 oversight, this default was not being set up when an address was tested by
7166 -bt or -bv, which affected the outcome if any router or filter referred to
7169 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7170 log only when "message" or "log_message" are set. However, if one of the
7171 conditions was an address verification, or a call to a nested ACL, the
7172 messages generated by the underlying test were being passed through. This
7173 no longer happens. The underlying message is available in $acl_verify_
7174 message for both "message" and "log_message" expansions, so it can be
7175 passed through if needed.
7177 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7178 new expansion $bh_ to give the encoded byte string without charset
7179 translation. Translation happens only if iconv() is available; HAVE_ICONV
7180 indicates this at build time. HEADERS_CHARSET gives the charset to
7181 translate to; headers_charset can change it in the configuration, and
7182 "headers charset" can change it in an individual filter file.
7184 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7185 that creates RFC 2047 encoded "words" labels them as that charset instead
7186 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7187 expansion operator; (ii) when Exim creates a From: line for a local
7188 message; (iii) when a header line is rewritten to include a "phrase" part.
7190 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7191 buggy, causing it to skip the first lines of messages whose message ID
7192 ended in 'D'. This would not have bitten before Exim release 4.14, because
7193 message IDs were unlikely to end in 'D' before then. The effect was to have
7194 incorrect size information for certain domains.
7196 45. #include "config.h" was missing at the start of the crypt16.c module. This
7197 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7200 46. If there was a timeout during a "random" callout check, Exim treated it as
7201 a failure of the random address, and carried on sending RSET and the real
7202 address. If the delay was just some slowness somewhere, the response to the
7203 original RCPT would be taken as a response to RSET and so on, causing
7204 mayhem of various kinds.
7206 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7207 when I implemented it. It didn't allow for the fact that some option values
7208 may legitimately be negative (e.g. size_addition), and it didn't even do
7209 the right test for positive values.
7211 48. Domain names in DNS records are case-independent. Exim always looks them up
7212 in lower case. Some resolvers return domain names in exactly the case they
7213 appear in the zone file, that is, they may contain uppercase letters. Not
7214 all resolvers do this - some return always lower case. Exim was treating a
7215 change of case by a resolver as a change of domain, similar to a widening
7216 of a domain abbreviation. This triggered its re-routing code and so it was
7217 trying to route what was effectively the same domain again. This normally
7218 caused routing to fail (because the router wouldn't handle the domain
7219 twice). Now Exim checks for this case specially, and just changes the
7220 casing of the domain that it ultimately uses when it transmits the message
7223 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7226 50. If a filter generated a file delivery with a non-absolute name (possible if
7227 no home directory exists for the router), the forbid_file option was not
7230 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7231 the existing equality matching.
7233 52. Exim was using ints instead of ino_t variables in some places where it was
7234 dealing with inode numbers.
7236 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7237 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7238 called TMPDIR, and if it finds it is different, it changes its value.
7240 54. The smtp_printf() function is now made available to local_scan() so
7241 additional output lines can be written before returning. There is also an
7242 smtp_fflush() function to enable the detection of a dropped connection.
7243 The variables smtp_input and smtp_batched_input are exported to
7246 55. Changed the default runtime configuration: the message "Unknown user"
7247 has been removed from the ACL, and instead placed on the localuser router,
7248 using the cannot_route_message feature. This means that any verification
7249 failures that generate their own messages won't get overridden. Similarly,
7250 the "Unrouteable address" message that was in the ACL for unverifiable
7251 relay addresses has also been removed.
7253 56. Added hosts_avoid_esmtp to the smtp transport.
7255 57. The exicyclog script was not checking for the esoteric option
7256 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7257 will work only if exicyclog is run under the appropriate euid.
7259 58. Following a discussion on the list, the rules by which Exim recognises line
7260 endings on incoming messages have been changed. The -dropcr and drop_cr
7261 options are now no-ops, retained only for backwards compatibility. The
7262 following line terminators are recognized: LF CRLF CR. However, special
7263 processing applies to CR:
7265 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7266 nor a local message in the state where . is a terminator.
7268 (ii) If a bare CR is encountered in a header line, an extra space is added
7269 after the line terminator so as not to end the header. The reasoning
7270 behind this is that bare CRs in header lines are most likely either
7271 to be mistakes, or people trying to play silly games.
7273 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7274 was being incorrectly given as 18 bytes larger than it should have been.
7275 This is a VOB (very old bug).
7277 60. This may never have affected anything current, but just in case it has:
7278 When the local host is found other than at the start of a list of hosts,
7279 the local host, those with the same MX, and any that follow, are discarded.
7280 When the list in question was part of a longer list of hosts, the following
7281 hosts (not currently being processed) were also being discarded. This no
7282 longer happens. I'm not sure if this situation could ever has previously
7285 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7288 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7289 header. This is something that is recommended in a new Internet Draft, and
7290 is something that is documented as being done by Sendmail. There are two
7291 possible values. For messages generated by the autoreply transport, Exim
7294 Auto-Submitted: auto-replied
7296 whereas for all other generated messages (e.g. bounces) it adds
7298 Auto-Submitted: auto-generated
7300 63. The "personal" condition in filters now includes a test for the
7301 Auto-Submitted: header. If it contains the string "auto-" the message it
7302 not considered personal.
7304 64. Added rcpt_include_affixes as a generic transport option.
7306 65. Added queue_only_override (default true).
7308 66. Added the syslog_duplication option.
7310 67. If what should have been the first header line of a message consisted of
7311 a space followed by a colon, Exim was mis-interpreting it as a header line.
7312 It isn't of course - it is syntactically invalid and should therefore be
7313 treated as the start of the message body. The misbehaviour could have
7314 caused a number of strange effects, including loss of data in subsequent
7315 header lines, and spool format errors.
7317 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7318 client host had authenticated. This control can now be exercised by an ACL
7319 for more flexibility.
7321 69. By default, callouts do not happen when testing with -bh. There is now a
7322 variant, -bhc, which does actually run the callout code, including
7323 consulting and updating the callout cache.
7325 70. Added support for saslauthd authentication, courtesy of Alexander
7328 71. If statvfs() failed on the spool or log directories while checking their
7329 size for availability, Exim confusingly gave the error "space shortage".
7330 Furthermore, in debugging mode it crashed with a floating point exception.
7331 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7332 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7333 serious problem, Exim now writes to the main and panic logs when this
7334 happens, with details of the failure. It then refuses to accept the
7335 incoming message, giving the message "spool directory problem" or "log
7336 directory problem" with a 421 code for SMTP messages.
7338 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7339 0, 1, and 2 exist, because some OS complain for execs without them (see
7340 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7341 descriptors. However, the code omitted to check that the open succeeded,
7342 causing mysterious errors if for some reason the permissions on /dev/null
7343 got screwed. Now Exim writes a message to the main and panic logs, and
7344 bombs out if it can't open /dev/null.
7346 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7347 interact so that it is all more flexible. It is supposed to remain
7348 backwards compatible. Also added extra_local_interfaces.
7350 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7351 to bomb out with an assertion failure - to the client this appears as a
7352 connection drop. This problem occurs in the part of the code that was taken
7353 from the Samba project. Fortunately, the assertion is in a very simple
7354 function, so I have fixed this by reproducing the function inline in the
7355 one place where it is called, and arranging for authentication to fail
7356 instead of killing the process with assert().
7358 75. The SPA client code was not working when the server requested OEM rather
7359 than Unicode encoding.
7361 76. Added code to make require_files with a specific uid setting more usable in
7362 the case where statting the file as root fails - usually a non-root-mounted
7363 NFS file system. When this happens and the failure is EACCES, Exim now
7364 forks a subprocess and does the per-uid checking as the relevant uid.
7366 77. Added process_log_path.
7368 78. If log_file_path was not explicitly set, a setting of check_log_space or
7369 check_log_inodes was ignored.
7371 79. If a space check for the spool or log partitions fails, the incident is now
7372 logged. Of course, in the latter case the data may get lost...
7374 80. Added the %p formatting code to string_format() so that it can be used to
7375 print addresses in debug_print(). Adjusted all the address printing in the
7376 debugging in store.c to use %p rather than %d.
7378 81. There was a concern that a line of code in smtp_in.c could overflow a
7379 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7380 initially expressed, the concern was not well-founded, because trailing
7381 spaces are removed early. However, if the trailing spaces were followed by
7382 a NULL, they did not get removed, so the overflow was possible. Two fixes
7385 (a) I re-wrote the offending code in a cleaner fashion.
7386 (b) If an incoming SMTP command contains a NULL character, it is rejected
7389 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7390 runs initgroups(), so that if the Exim user is in any additional groups,
7391 they will be used during message reception.
7397 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob