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 Bug 2375: fix expansions of 822 addresses having comments in local-part
36 and/or domain. Found and fixed by Jason Betts.
38 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
39 configuration). If a CNAME target was not a wellformed name pattern, a
46 JH/01 Remove code calling the customisable local_scan function, unless a new
47 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
49 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
50 non-signal-safe functions being used.
52 JH/03 Bug 2269: When presented with a received message having a stupidly large
53 number of DKIM-Signature headers, disable DKIM verification to avoid
54 a resource-consumption attack. The limit is set at twenty.
56 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
57 report of oldest_pass in ${authres } in consequence, and separate out
58 some descriptions of reasons for verification fail.
60 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
61 files in the spool were present and unlocked. A queue-runner could spot
62 them, resulting in a duplicate delivery. Fix that by doing the unlock
63 after the unlink. Investigation by Tim Stewart. Take the opportunity to
64 add more error-checking on spoolfile handling while that code is being
67 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
68 No known attacks, no CVE, this is defensive hardening.
70 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
71 a queue-runner could start a delivery while other operations were ongoing.
72 Cutthrough delivery was a common victim, resulting in duplicate delivery.
73 Found and investigated by Tim Stewart. Fix by using the open message data
74 file handle rather than opening another, and not locally closing it (which
75 releases a lock) for that case, while creating the temporary .eml format
76 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
78 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
79 $sender_verify_failure/$recipient_verify_failure to "random".
81 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
84 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
85 Previously this would segfault.
87 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
90 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
91 like zero, since the resolver should be doing this for us, But we need one
92 as a CNAME but no MX presence gets the CNAME returned; we need to check
93 that doesn't point to an MX to declare it "no MX returned" rather than
94 "error, loop". A new main option is added so the older capability of
95 following some limited number of chain links is maintained.
97 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
99 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
100 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
101 now-deprecated earlier definitions used only specified the range up to TLS
102 1.2 (in the older-version library docs).
104 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
106 JH/15 Rework TLS client-side context management. Stop using a global, and
107 explicitly pass a context around. This enables future use of TLS for
108 connections to service-daemons (eg. malware scanning) while a client smtp
109 connection is using TLS; with cutthrough connections this is quite likely.
111 JH/16 Fix ARC verification to do AS checks in reverse order.
113 JH/17 Support a "tls" option on the ${readsocket } expansion item.
115 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
116 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
117 Previously the "utf8" would be re-prepended for every additional message.
119 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
120 Previously thery were accepted, resulting in issues when attempting to
121 forward messages to a non-supporting MTA.
123 PP/02 Let -n work with printing macros too, not just options.
125 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
126 one parent address was copied, and bogus data was used at delivery-logging
127 time. Either a crash (after delivery) or bogus log data could result.
128 Discovery and analysis by Tim Stewart.
130 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
131 Previously if the string ended mid-character, we did not insert the
132 promised '?' replacement.
134 PP/04 Documentation: current string operators work on bytes, not codepoints.
136 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
137 should pack well giving a smaller memory footprint so better caching and
138 therefore performance. Group the declarations where this can't be done so
139 that the byte-sized flag variables are not interspersed among pointer
140 variables, giving a better chance of good packing by the compiler.
142 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
143 non-null, to avoid issues with sites running BATV. Previously reports were
144 sent with an empty envelope sender so looked like bounces.
146 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
147 The ignore_error flag wasn't being returned from the filter subprocess so
148 was not set for later routers. Investigation and fix by Matthias Kurz.
150 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
151 and a msg:complete for the whole, when a message is manually removed using
152 -Mrm. Developement by Matthias Kurz, hacked on by JH.
154 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
155 a "Gnu special" function, asprintf() in the DB utility binary builds; I
156 hope that is portable enough.
158 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
159 requiring a known-CA anchor certificate; make it now rely entirely on the
160 TLSA as an anchor. Checking the name on the leaf cert against the name
161 on the A-record for the host is still done for TA (but not for EE mode).
163 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
164 included in delivery lines for non-proxied connections, when compiled with
165 SUPPORT_SOCKS and running with proxy logging enabled.
167 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
168 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
169 move the existing event to fire before the normal logging of message
170 failure so that custom logging is bracketed by normal logging.
172 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
173 msg:fail:internal event. Developement by Matthias Kurz.
175 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
176 far too small for todays use of crypto signatures stored there. Go all
177 the way to the max DNS message size of 64kB, even though this might be
178 overmuch for IOT constrained device use.
180 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
181 copy a string over itself. The library routine is documented as not
182 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
184 JH/32 For main options check_spool_space and check_inode_space, where the
185 platform supports 64b integers, support more than the previous 2^31 kB
186 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
187 the previous G, M, k.
189 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
190 $authenticated_fail_id variable on authentication failure. Previously
193 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
194 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
195 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
196 GNUTLS_SEC_PARAM_MEDIUM.
198 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
199 side. Previously we would continue as if no SNI had been received.
201 JH/36 Harden the handling of string-lists. When a list consisted of a sole
202 "<" character, which should be a list-separator specification, we walked
203 off past the nul-terimation.
205 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
206 causes) even when the retry time is not yet met. Previously they were
207 not, meaning that when (say) an account was over-quota and temp-rejecting,
208 and multiple senders' messages were queued, only one sender would get
209 notified on each configured delay_warning cycle.
211 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
213 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
214 cipher-suites, an error can be left on the stack even for a succeeding
215 accept; this results in impossible error messages when a later operation
218 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
219 return error codes indicating retry. Under TLS1.3 this becomes required.
221 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
222 it only wrote the new authenticators, resulting in a lack of tracking of
223 peer changes of ESMTP extensions until the next cache flush.
225 JH/41 Fix the loop reading a message header line to check for integer overflow,
226 and more-often against header_maxsize. Previously a crafted message could
227 induce a crash of the recive process; now the message is cleanly rejected.
229 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
230 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
236 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
237 When redis_servers is set to a list of > 1 element, and the Redis servers
238 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
239 case of MOVED into a DEFER case instead, thus moving the query onto the
240 next server in the list. For a cluster of N elements, all N servers must
241 be defined in redis_servers.
243 GF/02 Catch and remove uninitialized value warning in exiqsumm
244 Check for existence of @ARGV before looking at $ARGV[0]
246 JH/01 Replace the store_release() internal interface with store_newblock(),
247 which internalises the check required to safely use the old one, plus
248 the allocate and data copy operations duplicated in both (!) of the
249 extant use locations.
251 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
252 modifier. This matches the restriction on the commandline.
254 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
255 Previously only the last row was returned.
257 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
258 we assumed that tags in the header were well-formed, and parsed the
259 element content after inspecting only the first char of the tag.
260 Assumptions at that stage could crash the receive process on malformed
263 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
264 While running the DKIM ACL we operate on the Permanent memory pool so that
265 variables created with "set" persist to the DATA ACL. Also (at any time)
266 DNS lookups that fail create cache records using the Permanent pool. But
267 expansions release any allocations made on the current pool - so a dnsdb
268 lookup expansion done in the DKIM ACL releases the memory used for the
269 DNS negative-cache, and bad things result. Solution is to switch to the
270 Main pool for expansions.
271 While we're in that code, add checks on the DNS cache during store_reset,
272 active in the testsuite.
273 Problem spotted, and debugging aided, by Wolfgang Breyha.
275 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
276 When none of the hosts presented to a transport match an already-open
277 connection, close it and proceed with the list. Previously we would
278 queue the message. Spotted by Lena with Yahoo, probably involving
281 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
282 Previously a spurious "250 OK id=" response was appended to the proper
285 JH/08 The "support for" informational output now, which built with Content
286 Scanning support, has a line for the malware scanner interfaces compiled
287 in. Interface can be individually included or not at build time.
289 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
290 by the template makefile "src/EDITME". The "STREAM" support for an older
291 ClamAV interface method is removed.
293 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
294 rows affected is given instead).
296 JH/11 The runtime Berkeley DB library version is now additionally output by
297 "exim -d -bV". Previously only the compile-time version was shown.
299 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
300 SMTP connection. Previously, when one had more receipients than the
301 first, an abortive onward connection was made. Move to full support for
302 multiple onward connections in sequence, handling cutthrough connection
303 for all multi-message initiating connections.
305 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
306 routers. Previously, a multi-recipient message would fail to match the
307 onward-connection opened for the first recipient, and cause its closure.
309 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
310 a timeout on read on a GnuTLS initiating connection, resulting in the
311 initiating connection being dropped. This mattered most when the callout
312 was marked defer_ok. Fix to keep the two timeout-detection methods
315 JH/15 Relax results from ACL control request to enable cutthrough, in
316 unsupported situations, from error to silently (except under debug)
317 ignoring. This covers use with PRDR, frozen messages, queue-only and
320 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
322 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
323 metadata, resulting in a crash in free().
325 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
326 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
327 Broken also in d185889f4, with init system revamp.
329 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
330 Previously we abruptly closed the connection after reading a malware-
331 found indication; now we go on to read the "scan ok" response line,
334 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
335 ACL. Previously, a crash would result.
337 JH/19 Speed up macro lookups during configuration file read, by skipping non-
338 macro text after a replacement (previously it was only once per line) and
339 by skipping builtin macros when searching for an uppercase lead character.
341 JH/20 DANE support moved from Experimental to mainline. The Makefile control
342 for the build is renamed.
344 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
345 was allocated for every new TLS startup, meaning one per message. Fix
346 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
348 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
349 reported the original. Fix to report (as far as possible) the ACL
350 result replacing the original.
352 JH/23 Fix memory leak during multi-message connections using STARTTLS under
353 OpenSSL. Certificate information is loaded for every new TLS startup,
354 and the resources needed to be freed.
356 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
358 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
361 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
362 DATA response info to the (existing) per-recipient response info for
363 the "C=" log element. It can have useful tracking info from the
364 destination system. Patch from Simon Arlott.
366 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
367 length value. Previously this would segfault.
369 HS/02 Support Avast multiline protoocol, this allows passing flags to
370 newer versions of the scanner.
372 JH/28 Ensure that variables possibly set during message acceptance are marked
373 dead before release of memory in the daemon loop. This stops complaints
374 about them when the debug_store option is enabled. Discovered specifically
375 for sender_rate_period, but applies to a whole set of variables.
376 Do the same for the queue-runner and queue-list loops, for variables set
377 from spool message files. Do the same for the SMTP per-message loop, for
378 certain variables indirectly set in ACL operations.
380 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
381 as a multi-recipient message from a mailinglist manager). The coding had
382 an arbitrary cutoff number of characters while checking for more input;
383 enforced by writing a NUL into the buffer. This corrupted long / fast
384 input. The problem was exposed more widely when more pipelineing of SMTP
385 responses was introduced, and one Exim system was feeding another.
386 The symptom is log complaints of SMTP syntax error (NUL chars) on the
387 receiving system, and refused recipients seen by the sending system
388 (propating to people being dropped from mailing lists).
389 Discovered and pinpointed by David Carter.
391 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
392 replaced by the ${authresults } expansion.
394 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
396 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
397 allows proper process termination in container environments.
399 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
400 Previously the "final dot" had a newline after it; ensure it is CR,LF.
402 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
403 and "err_perm", deprecated since 4.83 when the RFC-defined words
404 "temperror" and "permerror" were introduced.
406 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
407 transport-filters or DKIM-signing. The restriction was lost in the
408 consolidation of verify-callout and delivery SMTP handling.
409 Extend the restriction to also cover ARC-signing.
411 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
412 in defer=pass mode supply a 450 to the initiator. Previously the message
415 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
416 tls_require_ciphers is used as before.
418 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
419 "pass_unscanned". Only tmpfails from the scanner are written to
420 the paniclog, as they may require admin intervention (permission
421 denied, license issues). Other scanner errors (like decompression
422 bombs) do not cause a paniclog entry.
424 JH/36 Fix reinitialisation of DKIM logging variable between messages.
425 Previously it was possible to log spurious information in receive log
428 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
429 triggered odd behaviour from Outlook Express clients.
431 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
434 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
435 since the IETF WG has not yet settled on that versus the original
436 "bare" representation.
438 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
439 Previously the millisecond value corrupted the output.
440 Fix also for syslog_pid=no and log_selector +pid, for which the pid
441 corrupted the output.
447 JH/01 Rework error string handling in TLS interface so that the caller in
448 more cases is responsible for logging. This permits library-sourced
449 string to be attached to addresses during delivery, and collapses
450 pairs of long lines into single ones.
452 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
453 during configuration. Wildcards are allowed and expanded.
455 JH/02 Rework error string handling in DKIM to pass more info back to callers.
456 This permits better logging.
458 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
459 do not close it down and have the child transport start it up again on
460 the passed-on TCP connection. Instead, proxy the child (and any
461 subsequent ones) for TLS via a unix-domain socket channel. Logging is
462 affected: the continued delivery log lines do not have any DNSSEC, TLS
463 Certificate or OCSP information. TLS cipher information is still logged.
465 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
466 identical IP addresses on different listening ports. Will also affect
469 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
470 add noisy ifdef guards to special-case this sillyness.
471 Patch from Bernd Kuhls.
473 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
474 than 255 are no longer allowed.
476 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
477 Disable the session-cache too, which might reduce our load. Since we
478 currrectly use a new context for every connection, both as server and
479 client, there is no benefit for these.
480 GnuTLS appears to not support tickets server-side by default (we don't
481 call gnutls_session_ticket_enable_server()) but client side is enabled
482 by default on recent versions (3.1.3 +) unless the PFS priority string
485 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
486 <https://reproducible-builds.org/specs/source-date-epoch/>.
488 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
489 the check for any unsuccessful recipients did not notice the limit, and
490 erroneously found still-pending ones.
492 JH/08 Pipeline CHUNKING command and data together, on kernels that support
493 MSG_MORE. Only in-clear (not on TLS connections).
495 JH/09 Avoid using a temporary file during transport using dkim. Unless a
496 transport-filter is involved we can buffer the headers in memory for
497 creating the signature, and read the spool data file once for the
498 signature and again for transmission.
500 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
501 4.77 as the kernel support then wasn't solid, having issues in 64bit
502 mode. Now, it's been long enough. Add support for FreeBSD also.
504 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
505 case where the routing stage had gathered several addresses to send to
506 a host before calling the transport for the first, we previously failed
507 to close down TLS in the old transport process before passing the TCP
508 connection to the new process. The new one sent a STARTTLS command
509 which naturally failed, giving a failed delivery and bloating the retry
510 database. Investigation and fix prototype from Wolfgang Breyha.
512 JH/12 Fix check on SMTP command input synchronisation. Previously there were
513 false-negatives in the check that the sender had not preempted a response
514 or prompt from Exim (running as a server), due to that code's lack of
515 awareness of the SMTP input buffering.
517 PP/04 Add commandline_checks_require_admin option.
518 Exim drops privileges sanely, various checks such as -be aren't a
519 security problem, as long as you trust local users with access to their
520 own account. When invoked by services which pass untrusted data to
521 Exim, this might be an issue. Set this option in main configuration
522 AND make fixes to the calling application, such as using `--` to stop
525 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
526 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
527 responses to those, into a single TLS record each way (this usually means
528 a single packet). As a side issue, smtp_enforce_sync now works on TLS
531 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
532 affects you only if you're dancing at the edge of the param size limits.
533 If you are, and this message makes sense to you, then: raise the
534 configured limit or use OpenSSL 1.1. Nothing we can do for older
537 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
538 cmdline element to get the documented default one. Previously it was
541 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
542 are unsuable for non-message-specific reasons - eg. connection timeout,
543 banner-time rejection.
545 JH/16 Fix logging of delivery remote port, when specified by router, under
548 PP/06 Repair manualroute's ability to take options in any order, even if one
549 is the name of a transport.
552 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
554 JH/17 Change the list-building routines interface to use the expanding-string
555 triplet model, for better allocation and copying behaviour.
557 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
558 Previously it was constructed the first time a possibly-matching string
559 was met in the configuration file input during startup; now it is done
562 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
563 compatible one, to avoid the (poorly documented) possibility of a config
564 file in the working directory redirecting the DB files, possibly correpting
565 some existing file. CVE-2017-10140 assigned for BDB.
567 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
568 cache-hot. Previously, although the result was properly cached, the
569 initial verify call returned a defer.
571 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
572 the main verify for receipient in uncached-mode.
574 JH/22 Retire historical build files to an "unsupported" subdir. These are
575 defined as "ones for which we have no current evidence of testing".
577 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
578 if present. Previously it was ignored.
580 JH/24 Start using specified-initialisers in C structure init coding. This is
581 a C99 feature (it's 2017, so now considered safe).
583 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
584 if was a fixed-sized field and bitmask ops via macros; it is now more
587 PP/07 GitHub PR 56: Apply MariaDB build fix.
588 Patch provided by Jaroslav Škarvada.
590 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
591 during Coverity cleanups [4.87 JH/47]
592 Diagnosis and fix provided by Michael Fischer v. Mollard.
594 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
595 the right size to place the terminating semicolon on its own folded
596 line, the header hash was calculated to an incorrect value thanks to
597 the (relaxed) space the fold became.
599 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
600 and confused the parent.
602 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
603 which could crash as a result. This could lead to undeliverable messages.
605 JH/28 Logging: "next input sent too soon" now shows where input was truncated
608 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
609 matters on fast-turnover and PID-randomising systems, which were getting
610 out-of-order delivery.
612 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
613 a possibly-overlapping copy. The symptom was that "Remote host closed
614 connection in response to HELO" was logged instead of the actual 4xx
617 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
618 Previously only that bufferd was discarded, resulting in SYMTP command
621 JH/32 DKIM: when a message has multiple signatures matching an identity given
622 in dkim_verify_signers, run the dkim acl once for each. Previously only
623 one run was done. Bug 2189.
625 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
626 "panic the current process" to "deliberately defer". The panic log is
627 still written with the problem list name; the mail and reject logs now
628 get a temp-reject line for the message that was being handled, saying
629 something like "domains check lookup or other defer". The SMTP 451
630 message is still "Temporary local problem".
632 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
633 A crafted sequence of BDAT commands could result in in-use memory beeing
634 freed. CVE-2017-16943.
636 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
637 from SMTP input. Previously it was always done; now only done for DATA
638 and not BDAT commands. CVE-2017-16944.
640 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
641 to the message (such as an overlong header line). Previously this was
642 not done and we did not exit BDAT mode. Followon from the previous item
643 though a different problem.
649 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
650 than -2003 did; needs libidn2 in addition to libidn.
652 JH/02 The path option on a pipe transport is now expanded before use.
654 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
655 Patch provided by "Björn", documentation fix added too.
657 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
658 missing a wire-to-host endian conversion.
660 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
661 close after a BDAT command line could be taken as a following command,
662 giving a synch failure. Fix by only checking for synch immediately
663 before acknowledging the chunk.
665 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
666 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
667 macro. Patches provided by Josh Soref.
669 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
670 Previously we did not; the RFC seems ambiguous and VRFY is not listed
671 by IANA as a service extension. However, John Klensin suggests that we
674 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
675 the dkim code may be unix-mode line endings rather than smtp wire-format
676 CRLF, so prepend a CR to any bare LF.
678 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
679 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
681 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
682 the first were themselves being wrongly included in the feed into dkim
683 processing; with most chunk sizes in use this resulted in an incorrect
684 body hash calculated value.
686 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
687 DKIM signature block, for verification. Although advised against by
688 standards it is specifically not ruled illegal.
690 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
692 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
693 missing a body hash (the bh= tag).
695 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
696 It seems that HAProxy sends the Proxy Protocol information in clear and
697 only then does a TLS startup, so do the same.
699 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
700 TCP connections (such as for Spamd) unless the daemon successfully set
701 Fast Open mode on its listening sockets. This fixes breakage seen on
702 too-old kernels or those not configured for Fast Open, at the cost of
703 requiring both directions being enabled for TFO, and TFO never being used
704 by non-daemon-related Exim processes.
706 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
707 endings, at least on the first header line. Try to canonify any that get
708 past that check, despite the cost.
710 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
711 now limited to an arbitrary five deep, while parsing addresses with the
712 strip_excess_angle_brackets option enabled.
714 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
715 instead leave the unprompted TLS handshake in socket buffer for the
716 TLS library to consume.
718 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
720 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
722 JH/16 Drop variables when they go out of scope. Memory management drops a whole
723 region in one operation, for speed, and this leaves assigned pointers
724 dangling. Add checks run only under the testsuite which checks all
725 variables at a store-reset and panics on a dangling pointer; add code
726 explicitly nulling out all the variables discovered. Fixes one known
727 bug: a transport crash, where a dangling pointer for $sending_ip_address
728 originally assigned in a verify callout, is re-used.
730 PP/06 Drop '.' from @INC in various Perl scripts.
732 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
734 PP/08 Reduce a number of compilation warnings under clang; building with
735 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
736 should be warning-free.
738 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
740 HS/01 Fix portability problems introduced by PP/08 for platforms where
741 realloc(NULL) is not equivalent to malloc() [SunOS et al].
743 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
744 chunk. This allows us to accept broken chunked messages. We need a more
745 general solution here.
747 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
748 already-broken messages in the queue.
750 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
752 JH/19 Fix reference counting bug in routing-generated-address tracking.
758 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
759 supports it and a size is available (ie. the sending peer gave us one).
761 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
762 years of being deprecated). The replacements are the ACLs
763 acl_smtp_mime and acl_not_smtp_mime.
765 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
766 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
767 or even an in-clear connection were permitted. Now, if the host lookup
768 was dnssec and dane was requested then the host is only used if the
769 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
770 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
771 if one fails this test.
772 This means that a poorly-configured remote DNS will make it incommunicado;
773 but it protects against a DNS-interception attack on it.
775 JH/04 Bug 1810: make continued-use of an open smtp transport connection
776 non-noisy when a race steals the message being considered.
778 JH/05 If main configuration option tls_certificate is unset, generate a
779 self-signed certificate for inbound TLS connections.
781 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
782 in rewrites and routers.
784 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
785 and logged a warning sing 4.83; now they are a configuration file error.
787 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
788 (lacking @domain). Apply the same qualification processing as RCPT.
790 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
792 JH/10 Support ${sha256:} applied to a string (as well as the previous
795 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
796 a cutthrough deliver is pending, as we always want to make a connection.
797 This also avoids re-routing the message when later placing the cutthrough
798 connection after a verify cache hit.
799 Do not update it with the verify result either.
801 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
802 when routing results in more than one destination address.
804 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
805 signing (which inhibits the cutthrough capability). Previously only
806 the presence of an option was tested; now an expansion evaluating as
807 empty is permissible (obviously it should depend only on data available
808 when the cutthrough connection is made).
810 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
811 the relevant preceding SMTP command did not note the pipelining mode.
813 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
814 Previously they were not counted.
816 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
817 as one having no matching records. Previously we deferred the message
818 that needed the lookup.
820 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
821 distinguished as "(=".
823 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
824 for missing MX records. Previously it only worked for missing A records.
826 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
828 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
829 after the data-go-ahead and data-ack. Patch from Jason Betts.
831 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
832 even for a "none" policy. Patch from Tony Meyer.
834 JH/22 Fix continued use of a connection for further deliveries. If a port was
835 specified by a router, it must also match for the delivery to be
838 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
839 When one of the recipients of a message was unsuitable for the connection
840 (has no matching addresses), we lost track of needing to mark it
841 deferred. As a result mail would be lost.
843 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
845 JH/25 Decoding ACL controls is now done using a binary search; the source code
846 takes up less space and should be simpler to maintain. Merge the ACL
847 condition decode tables also, with similar effect.
849 JH/26 Fix problem with one_time used on a redirect router which returned the
850 parent address unchanged. A retry would see the parent address marked as
851 delivered, so not attempt the (identical) child. As a result mail would
854 JH/27 Fix a possible security hole, wherein a process operating with the Exim
855 UID can gain a root shell. Credit to http://www.halfdog.net/ for
856 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
859 JH/28 Enable {spool,log} filesystem space and inode checks as default.
860 Main config options check_{log,spool}_{inodes,space} are now
861 100 inodes, 10MB unless set otherwise in the configuration.
863 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
864 Previously it only applied to the main-section connection policy
867 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
869 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
870 by me. Added RFC7919 DH primes as an alternative.
872 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
873 are not in the system include path.
875 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
876 GnuTLS, when a session startup failed (eg because the client disconnected)
877 Exim did stdio operations after fclose. This was exposed by a recent
878 change which nulled out the file handle after the fclose.
880 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
881 signed directly by the cert-signing cert, rather than an intermediate
882 OCSP-signing cert. This is the model used by LetsEncrypt.
884 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
886 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
887 an incoming connection.
889 HS/02 Bug 1802: Do not half-close the connection after sending a request
892 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
893 fallback to "prime256v1".
895 JH/34 SECURITY: Use proper copy of DATA command in error message.
896 Could leak key material. Remotely exploitable. CVE-2016-9963.
902 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
903 and 3.4.4 - once the server is enabled to respond to an OCSP request
904 it does even when not requested, resulting in a stapling non-aware
905 client dropping the TLS connection.
907 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
908 support variable-length bit vectors. No functional change.
910 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
911 The I= interface field on outgoing lines is now after the H= remote
912 host field, same as incoming lines. There is a separate
913 outgoing_interface log selector which allows you to disable the
916 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
917 If not running log_selector +smtp_connection the mainlog would be held
918 open indefinitely after a "too many connections" event, including to a
919 deleted file after a log rotate. Leave the per net connection logging
920 leaving it open for efficiency as that will be quickly detected by the
921 check on the next write.
923 HS/01 Bug 1671: Fix post transport crash.
924 Processing the wait-<transport> messages could crash the delivery
925 process if the message IDs didn't exist for some reason. When
926 using 'split_spool_directory=yes' the construction of the spool
927 file name failed already, exposing the same netto behaviour.
929 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
930 mime_regex ACL conditions.
932 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
933 to DSN fail messages (bounces): remote IP, remote greeting, remote response
934 to HELO, local diagnostic string.
936 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
937 log line to debug. Even when configured with a tls authenticator many
938 client connections are expected to not authenticate in this way, so
939 an authenticate fail is not an error.
941 HS/02 Add the Exim version string to the process info. This way exiwhat
942 gives some more detail about the running daemon.
944 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
945 matter for fast-change records such as DNSBLs.
947 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
948 retry record, even if constant. There may be multiple transports with
949 different interface settings and the retry behaviour needs to be kept
952 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
954 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
956 JH/10 Bug 840: fix log_defer_output option of pipe transport
958 JH/11 Bug 830: use same host for all RCPTS of a message, even under
959 hosts_randomize. This matters a lot when combined with mua_wrapper.
961 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
962 ${quote_pgsql:<string>} operator.
964 JH/13 Bug 1708: avoid misaligned access in cached lookup.
966 JH/14 Change header file name for freeradius-client. Relevant if compiling
967 with Radius support; from the Gentoo tree and checked under Fedora.
969 JH/15 Bug 1712: Introduce $prdr_requested flag variable
971 JH/16 Bug 1714: Permit an empty string as expansion result for transport
972 option transport_filter, meaning no filtering.
974 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
976 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
977 defaults to "*" (all hosts). The variable is now available when not built
978 with TLS, default unset, mainly to enable keeping the testsuite sane.
979 If a server certificate is not supplied (via tls_certificate) an error is
980 logged, and clients will find TLS connections fail on startup. Presumably
981 they will retry in-clear.
982 Packagers of Exim are strongly encouraged to create a server certificate
983 at installation time.
985 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
986 with the $config_file variable.
988 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
989 in transport context, after the attempt, and per-recipient. The latter type
990 is per host attempted. The event data is the error message, and the errno
991 information encodes the lookup type (A vs. MX) used for the (first) host,
992 and the trailing two digits of the smtp 4xx response.
994 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
995 to write to mainlog (or rejectlog, paniclog) in the window between file
996 creation and permissions/ownership being changed. Particularly affects
997 installations where exicyclog is run as root, rather than exim user;
998 result is that the running daemon panics and dies.
1000 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1002 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1003 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1004 "pri" and "weight". Note that the previous implicit priority given by the
1005 list order is no longer honoured.
1007 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1008 for DKIM processing.
1010 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1011 by defining SUPPORT_SOCKS.
1013 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1014 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1015 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1016 variables are renamed to proxy_{local,external}_{address,port}.
1018 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1019 for a build by defining SUPPORT_I18N
1021 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1022 of the query string, and make ${quote_redis:} do that quoting.
1024 JH/29 Move Events support from Experimental to mainline, enabled by default
1025 and removable for a build by defining DISABLE_EVENT.
1027 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1029 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1030 cached by the daemon.
1032 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1033 by defining LOOKUP_REDIS. The libhiredis library is required.
1035 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1036 keys are given for lookup.
1038 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1039 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1040 only supported when built with TLS support. The PolarSSL SHA routines
1041 are still used when the TLS library is too old for convenient support.
1043 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1044 openssl_options), for security. OpenSSL forces this from version 1.1.0
1045 server-side so match that on older versions.
1047 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1048 allocation for $value could be released as the expansion processing
1049 concluded, but leaving the global pointer active for it.
1051 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1052 and to use the domains and local_parts ACL conditions.
1054 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1055 incorrectly not doubled on cutthrough transmission, hence seen as a
1056 body-termination at the receiving system - resulting in truncated mails.
1057 Commonly the sender saw a TCP-level error, and retransmitted the message
1058 via the normal store-and-forward channel. This could result in duplicates
1059 received - but deduplicating mailstores were liable to retain only the
1060 initial truncated version.
1062 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1064 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1066 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1067 we're in there, support oversigning also; bug 1309.
1069 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1071 HS/04 Add support for keep_environment and add_environment options.
1073 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1074 either intentional arithmetic overflow during PRNG, or testing config-
1077 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1078 delivery resulted in actual delivery. Cancel cutthrough before DATA
1081 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1082 rejects a recipient: pass the reject to the originator.
1084 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1085 Many were false-positives and ignorable, but it's worth fixing the
1088 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1089 for the new environment-manipulation done at startup. Move the routines
1090 from being local to tls.c to being global via the os.c file.
1092 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1093 an extract embedded as result-arg for a map, the first arg for extract
1094 is unavailable so we cannot tell if this is a numbered or keyed
1095 extraction. Accept either.
1101 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1104 JH/02 The smtp transport option "multi_domain" is now expanded.
1106 JH/03 The smtp transport now requests PRDR by default, if the server offers
1109 JH/04 Certificate name checking on server certificates, when exim is a client,
1110 is now done by default. The transport option tls_verify_cert_hostnames
1111 can be used to disable this per-host. The build option
1112 EXPERIMENTAL_CERTNAMES is withdrawn.
1114 JH/05 The value of the tls_verify_certificates smtp transport and main options
1115 default to the word "system" to access the system default CA bundle.
1116 For GnuTLS, only version 3.0.20 or later.
1118 JH/06 Verification of the server certificate for a TLS connection is now tried
1119 (but not required) by default. The verification status is now logged by
1120 default, for both outbound TLS and client-certificate supplying inbound
1123 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1126 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1127 Status Notification (bounce) messages are now MIME format per RFC 3464.
1128 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1129 under the control of the dsn_advertise_hosts option, and routers may
1130 have a dsn_lasthop option.
1132 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1133 default, modifiable by a malware= option. The list separator for
1134 the options can now be changed in the usual way. Bug 68.
1136 JH/10 The smtp_receive_timeout main option is now expanded before use.
1138 JH/11 The incoming_interface log option now also enables logging of the
1139 local interface on delivery outgoing connections.
1141 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1142 if the interface and destination host and port all match.
1144 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1147 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1148 Patch from Andrew Lewis.
1150 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1151 now supports optional time-restrictions, weighting, and priority
1152 modifiers per server. Patch originally by <rommer@active.by>.
1154 JH/16 The spamd_address main option now supports a mixed list of local
1155 and remote servers. Remote servers can be IPv6 addresses, and
1156 specify a port-range.
1158 JH/17 Bug 68: The spamd_address main option now supports an optional
1159 timeout value per server.
1161 JH/18 Bug 1581: Router and transport options headers_add/remove can
1162 now have the list separator specified.
1164 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1167 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1170 JH/21 Support for the A6 type of dns record is withdrawn.
1172 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1173 rather than the verbs used.
1175 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1176 from 255 to 1024 chars.
1178 JH/24 Verification callouts now attempt to use TLS by default.
1180 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1181 are generic router options now. The defaults didn't change.
1183 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1184 Original patch from Alexander Shikoff, worked over by JH.
1186 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1187 files if ZCAT_COMMAND is not executable.
1189 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1191 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1193 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1194 Normally benign, it bites when the pair was led to by a CNAME;
1195 modern usage is to not canonicalize the domain to a CNAME target
1196 (and we were inconsistent anyway for A-only vs AAAA+A).
1198 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1200 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1201 when evaluating $sender_host_dnssec.
1203 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1204 $sender_helo_dnssec variable.
1206 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1208 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1210 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1212 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1213 documented as working, but never had. Support all but $spam_report.
1215 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1216 added for tls authenticator.
1218 HS/03 Add perl_taintmode main config option
1224 TL/01 When running the test suite, the README says that variables such as
1225 no_msglog_check are global and can be placed anywhere in a specific
1226 test's script, however it was observed that placement needed to be near
1227 the beginning for it to behave that way. Changed the runtest perl
1228 script to read through the entire script once to detect and set these
1229 variables, reset to the beginning of the script, and then run through
1230 the script parsing/test process like normal.
1232 TL/02 The BSD's have an arc4random API. One of the functions to induce
1233 adding randomness was arc4random_stir(), but it has been removed in
1234 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1235 function when detected.
1237 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1238 cause callback expansion.
1240 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1241 syntax errors in an expansion can be treated as a string instead of
1242 logging or causing an error, due to the internal use of bool_lax
1243 instead of bool when processing it.
1245 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1246 server certificates when making smtp deliveries.
1248 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1250 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1252 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1254 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1255 Merged patch from Sebastian Wiedenroth.
1257 JH/05 Fix results-pipe from transport process. Several recipients, combined
1258 with certificate use, exposed issues where response data items split
1259 over buffer boundaries were not parsed properly. This eventually
1260 resulted in duplicates being sent. This issue only became common enough
1261 to notice due to the introduction of connection certificate information,
1262 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1264 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1265 size buffer was used, resulting in syntax errors when an expansion
1268 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1269 version 3.3.6 or later.
1271 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1272 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1273 both become "event_action", the variables become $event_name, $event_data
1274 and $event_defer_errno. There is a new variable $verify_mode, usable in
1275 routers, transports and related events. The tls:cert event is now also
1276 raised for inbound connections, if the main configuration event_action
1279 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1280 early OCSP support, but no stapling (appears to be less than 1.0.0).
1282 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1283 server certificate names available under the smtp transport option
1284 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1287 JH/10 Time-related extraction expansions from certificates now use the main
1288 option "timezone" setting for output formatting, and are consistent
1289 between OpenSSL and GnuTLS compilations. Bug 1541.
1291 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1292 encoded parameter in the incoming message. Bug 1558.
1294 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1295 include certificate info, eximon was claiming there were spoolfile
1298 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1300 JH/14 Log delivery-related information more consistently, using the sequence
1301 "H=<name> [<ip>]" wherever possible.
1303 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1304 are problematic for Debian distribution, omit them from the release
1307 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1309 JH/16 Fix string representation of time values on 64bit time_t architectures.
1312 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1313 output list separator was used.
1318 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1319 checkers that were complaining about end of non-void function with no
1322 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1323 This was a regression introduced in 4.83 by another bugfix.
1325 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1327 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1328 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1334 TF/01 Correctly close the server side of TLS when forking for delivery.
1336 When a message was received over SMTP with TLS, Exim failed to clear up
1337 the incoming connection properly after forking off the child process to
1338 deliver the message. In some situations the subsequent outgoing
1339 delivery connection happened to have the same fd number as the incoming
1340 connection previously had. Exim would try to use TLS and fail, logging
1341 a "Bad file descriptor" error.
1343 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1344 utilities have not been installed.
1346 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1347 temporary space as the ACL may create new global variables.
1349 TL/01 LDAP support uses per connection or global context settings, depending
1350 upon the detected version of the libraries at build time.
1352 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1353 to extract and use the src ip:port in logging and expansions as if it
1354 were a direct connection from the outside internet. PPv2 support was
1355 updated based on HAProxy spec change in May 2014.
1357 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1359 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1360 Properly escape header and check for NULL return.
1362 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1365 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1367 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1368 characters in header names, implemented as a verify condition.
1369 Contributed by Michael Fischer v. Mollard.
1371 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1372 results permerror and temperror. Previous values are deprecated but
1373 still accepted. In a future release, err_perm and err_temp will be
1374 completely removed, which will be a backward incompatibility if the
1375 ACL tests for either of these two old results. Patch contributed by
1376 user bes-internal on the mailing list.
1378 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1380 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1381 selectors, in both main and reject logs.
1383 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1386 JH/07 Add malware type "sock" for talking to simple daemon.
1388 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1390 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1391 routers/transports under cutthrough routing.
1393 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1394 numbers. Touch up "bool" conditional to keep the same definition.
1396 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1398 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1400 JH/12 Expand items in router/transport headers_add or headers_remove lists
1401 individually rather than the list as a whole. Bug 1452.
1403 Required for reasonable handling of multiple headers_ options when
1404 they may be empty; requires that headers_remove items with embedded
1405 colons must have them doubled (or the list-separator changed).
1407 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1408 view the policy declared in the DMARC record. Currently, $dmarc_status
1409 is a combined value of both the record presence and the result of the
1412 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1414 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1415 dnslookup router and the smtp transport (applying to the forward
1418 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1419 of ldap servers used for a specific lookup. Patch provided by Heiko
1422 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1423 New variable $lookup_dnssec_authenticated for observability.
1425 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1426 Patch submitted by Lars Timman.
1428 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1430 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1431 Requires trusted mode and valid format message id, aborts otherwise.
1432 Patch contributed by Heiko Schlichting.
1434 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1435 certextract with support for various fields. Bug 1358.
1437 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1438 is requested by default, modifiable by smtp transport option
1441 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1442 operate on certificate variables to give certificate fingerprints
1443 Also new ${sha256:cert_variable}.
1445 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1447 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1448 Christian Aistleitner.
1450 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1452 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1453 file. Patch from Wolfgang Breyha.
1455 JH/25 Expand the coverage of the delivery $host and $host_address to
1456 client authenticators run in verify callout. Bug 1476.
1458 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1459 align with daemon_smtp_ports. Bug 72.
1461 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1462 support and error reporting did not work properly.
1464 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1465 and is readable. Patch from Andrew Colin Kissa.
1467 TL/14 Enhance documentation of ${run expansion and how it parses the
1468 commandline after expansion, particularly in the case when an
1469 unquoted variable expansion results in an empty value.
1471 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1473 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1474 by renaming away from it. Addresses GH issue 15, reported by
1477 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1479 TL/15 SECURITY: prevent double expansion in math comparison functions
1480 (can expand unsanitized data). Not remotely exploitable.
1487 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1489 PP/02 Make -n do something, by making it not do something.
1490 When combined with -bP, the name of an option is not output.
1492 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1495 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1496 $sender_host_name and config options to manage this, and basic check
1499 PP/05 DSCP support for outbound connections and control modifier for inbound.
1501 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1502 (Only plugin which currently uses this is kerberos4, which nobody should
1503 be using, but we should make it available and other future plugins might
1504 conceivably use it, even though it would break NAT; stuff *should* be
1505 using channel bindings instead).
1507 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1508 name; added for Sendmail compatibility; requires admin caller.
1509 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1510 just ignore it); requires trusted caller.
1511 Also parse but ignore: -Ac -Am -X<logfile>
1514 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1516 TL/02 Add +smtp_confirmation as a default logging option.
1518 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1519 Patch by Magnus Holmgren from 2007-02-20.
1521 TL/04 Bugzilla 1281 - Spec typo.
1522 Bugzilla 1283 - Spec typo.
1523 Bugzilla 1290 - Spec grammar fixes.
1525 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1527 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1529 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1530 mailing list by by Dmitry Isaikin.
1532 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1534 JH/02 Support "G" suffix to numbers in ${if comparisons.
1536 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1538 NM/01 Bugzilla 1197 - Spec typo
1539 Bugzilla 1196 - Spec examples corrections
1541 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1543 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1544 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1547 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1548 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1550 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1551 "acl {{name}{arg}...}", and optional args on acl condition
1554 JH/05 Permit multiple router/transport headers_add/remove lines.
1556 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1558 JH/07 Avoid using a waiting database for a single-message-only transport.
1559 Performance patch from Paul Fisher. Bugzilla 1262.
1561 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1564 JH/09 Add $headers_added variable, with content from use of ACL modifier
1565 add_header (but not yet added to the message). Bugzilla 199.
1567 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1568 Pulled from Bugzilla 817 by Wolfgang Breyha.
1570 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1572 (nb: this is the same fix as in Exim 4.80.1)
1574 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1577 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1580 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1582 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1583 advertise SMTP AUTH mechanism to us, instead of a generic
1584 protocol violation error. Also, make Exim more robust to bad
1585 data from the Dovecot auth socket.
1587 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1589 When a queue runner is handling a message, Exim first routes the
1590 recipient addresses, during which it prunes them based on the retry
1591 hints database. After that it attempts to deliver the message to
1592 any remaining recipients. It then updates the hints database using
1595 So if a recipient address works intermittently, it can get repeatedly
1596 deferred at routing time. The retry hints record remains fresh so the
1597 address never reaches the final cutoff time.
1599 This is a fairly common occurrence when a user is bumping up against
1600 their storage quota. Exim had some logic in its local delivery code
1601 to deal with this. However it did not apply to per-recipient defers
1602 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1604 This change adds a proper retry rule check during routing so that the
1605 final cutoff time is checked against the message's age. We only do
1606 this check if there is an address retry record and there is not a
1607 domain retry record; this implies that previous attempts to handle
1608 the address had the retry_use_local_parts option turned on. We use
1609 this as an approximation for the destination being like a local
1610 delivery, as in LMTP.
1612 I suspect this new check makes the old local delivery cutoff check
1613 redundant, but I have not verified this so I left the code in place.
1615 TF/02 Correct gecos expansion when From: is a prefix of the username.
1617 Test 0254 submits a message to Exim with the header
1621 When I ran the test suite under the user fanf2, Exim expanded
1622 the header to contain my full name, whereas it should have added
1623 a Resent-Sender: header. It erroneously treats any prefix of the
1624 username as equal to the username.
1626 This change corrects that bug.
1628 GF/01 DCC debug and logging tidyup
1629 Error conditions log to paniclog rather than rejectlog.
1630 Debug lines prefixed by "DCC: " to remove any ambiguity.
1632 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1634 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1635 Bug spotted by Jeremy Harris; was flawed since initial commit.
1636 Would have resulted in OCSP responses post-SNI triggering an Exim
1637 NULL dereference and crash.
1639 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1641 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1642 Bug detection, analysis and fix by Samuel Thibault.
1643 Bugzilla 1331, Debian bug #698092.
1645 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1647 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1648 Server implementation by Todd Lyons, client by JH.
1649 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1650 config variable "prdr_enable" controls whether the server
1651 advertises the facility. If the client requests PRDR a new
1652 acl_data_smtp_prdr ACL is called once for each recipient, after
1653 the body content is received and before the acl_smtp_data ACL.
1654 The client is controlled by both of: a hosts_try_prdr option
1655 on the smtp transport, and the server advertisement.
1656 Default client logging of deliveries and rejections involving
1657 PRDR are flagged with the string "PRDR".
1659 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1660 fclose(). Diagnosis by Todd Lyons.
1662 PP/17 Update configure.default to handle IPv6 localhost better.
1663 Patch by Alain Williams (plus minor tweaks).
1666 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1667 This is now consistent with GnuTLS, and is now documented: the
1668 previous undocumented portable approach to treating the option as
1669 unset was to force an expansion failure. That still works, and
1670 an empty string is now equivalent.
1672 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1673 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1674 not performing validation itself.
1676 PP/20 Added force_command boolean option to pipe transport.
1677 Patch from Nick Koston, of cPanel Inc.
1679 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1682 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1684 PP/21 Fix eximon continuous updating with timestamped log-files.
1685 Broken in a format-string cleanup in 4.80, missed when I repaired the
1686 other false fix of the same issue.
1687 Report and fix from Heiko Schlichting.
1690 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1691 Report from Prashanth Katuri.
1693 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1694 It's SecureTransport, so affects any MacOS clients which use the
1695 system-integrated TLS libraries, including email clients.
1697 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1698 using a MIME ACL for non-SMTP local injection.
1699 Report and assistance in diagnosis by Warren Baker.
1701 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1703 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1705 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1706 last id that failed so it may be referenced in subsequent ACL's.
1708 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1711 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1712 ldap library initialization, allowing self-signed CA's to be
1713 used. Also properly sets require_cert option later in code by
1714 using NULL (global ldap config) instead of ldap handle (per
1715 session). Bug diagnosis and testing by alxgomz.
1717 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1718 the src/util/ subdirectory.
1720 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1721 renamed to Transport Post Delivery Action by Jeremy Harris, as
1724 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1725 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1726 redis_servers = needs to be configured which will be used by the redis
1727 lookup. Patch from Warren Baker, of The Packet Hub.
1729 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1731 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1732 hostname or reverse DNS when processing a host list. Used suggestions
1733 from multiple comments on this bug.
1735 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1737 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1738 Missed a few lines, added it to make the runtest require no keyboard
1741 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1742 contains upper case chars. Make router use caseful_local_part.
1744 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1745 support when GnuTLS has been built with p11-kit.
1751 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1753 This, or similar/improved, will also be change PP/11 of 4.82.
1759 PP/01 Handle short writes when writing local log-files.
1760 In practice, only affects FreeBSD (8 onwards).
1761 Bugzilla 1053, with thanks to Dmitry Isaikin.
1763 NM/01 Bugzilla 949 - Documentation tweak
1765 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1768 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1770 PP/02 Implemented gsasl authenticator.
1772 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1774 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1775 `pkg-config foo` for cflags/libs.
1777 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1778 with rest of GSASL and with heimdal_gssapi.
1780 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1781 `pkg-config foo` for cflags/libs for the TLS implementation.
1783 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1784 properties get this fed in as external SSF. A number of robustness
1785 and debugging improvements to the cyrus_sasl authenticator.
1787 PP/08 cyrus_sasl server now expands the server_realm option.
1789 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1790 Patch by Jeremy Harris.
1792 PP/10 Added dbmjz lookup type.
1794 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1796 PP/12 MAIL args handles TAB as well as SP, for better interop with
1797 non-compliant senders.
1798 Analysis and variant patch by Todd Lyons.
1800 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1801 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1802 Patch from Dirk Mueller <dmueller@suse.com>
1804 PP/13 tls_peerdn now print-escaped for spool files.
1805 Observed some $tls_peerdn in wild which contained \n, which resulted
1806 in spool file corruption.
1808 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1809 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1810 or write after TLS renegotiation, which otherwise led to messages
1813 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1814 as a tracking header (ie: a signed header comes before the signature).
1815 Patch from Wolfgang Breyha.
1817 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1818 comma-sep list; embedded commas doubled.
1820 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1822 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1824 Report and patch from Dmitry Banschikov.
1826 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1827 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1828 failures. We appear to now support TLS1.1+ with Exim.
1830 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1831 lets Exim select keys and certificates based upon TLS SNI from client.
1832 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1833 before an outbound SMTP session. New log_selector, +tls_sni.
1835 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1836 NULL dereference. Report and patch from Alun Jones.
1838 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1839 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1840 Not seeing resolver debug output on NetBSD, but suspect this is a
1841 resolver implementation change.
1843 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1844 Left warnings. Added "eximon gdb" invocation mode.
1846 PP/21 Defaulting "accept_8bitmime" to true, not false.
1848 PP/22 Added -bw for inetd wait mode support.
1850 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1851 locate the relevant includes and libraries. Made this the default.
1853 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1854 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1856 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1857 This may cause build issues on older platforms.
1859 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1860 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1861 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1862 Added SNI support via GnuTLS too.
1863 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1865 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1867 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1868 Applied second patch from Janne, implementing suggestion to default
1869 multiple-strings-in-record handling to match SPF spec.
1871 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1873 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1874 read-only, out of scope).
1875 Patch from Wolfgang Breyha, report from Stuart Northfield.
1877 PP/29 Fix three issues highlighted by clang analyser static analysis.
1878 Only crash-plausible issue would require the Cambridge-specific
1879 iplookup router and a misconfiguration.
1880 Report from Marcin Mirosław.
1882 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1884 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1885 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1886 As part of this, removing so much warning spew let me fix some minor
1887 real issues in debug logging.
1889 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1890 assignment on my part. Fixed.
1892 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1893 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1894 Janne Snabb (who went above and beyond: thank you).
1896 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1897 string otherwise requires a connection and a bunch more work and it's
1898 relatively easy to get wrong. Should also expose TLS library linkage
1901 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1902 64-bit ${eval} (JH/03).
1904 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1905 GNU libc to support some of the 64-bit stuff, should not lead to
1906 conflicts. Defined before os.h is pulled in, so if a given platform
1907 needs to override this, it can.
1909 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1910 protection layer was required, which is not implemented.
1911 Bugzilla 1254, patch from Wolfgang Breyha.
1913 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1914 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1915 tls_dhparam take prime identifiers. Also unbreak combination of
1916 OpenSSL+DH_params+TLSSNI.
1918 PP/39 Disable SSLv2 by default in OpenSSL support.
1924 PP/01 Solaris build fix for Oracle's LDAP libraries.
1925 Bugzilla 1109, patch from Stephen Usher.
1927 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1929 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1932 TF/02 Fix a couple more cases where we did not log the error message
1933 when unlink() failed. See also change 4.74-TF/03.
1935 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1936 lock up or crash if it happened to be inside a call to libc when it
1937 got a SIGUSR1 from exiwhat.
1939 The SIGUSR1 handler appends the current process status to the process
1940 log which is later printed by exiwhat. It used to use the general
1941 purpose logging code to do this, but several functions it calls are
1942 not safe for signals.
1944 The new output code in the SIGUSR1 handler is specific to the process
1945 log, and simple enough that it's easy to inspect for signal safety.
1946 Removing some special cases also simplifies the general logging code.
1947 Removing the spurious timestamps from the process log simplifies
1950 TF/04 Improved ratelimit ACL condition.
1952 The /noupdate option has been deprecated in favour of /readonly which
1953 has clearer semantics. The /leaky, /strict, and /readonly update modes
1954 are mutually exclusive. The update mode is no longer included in the
1955 database key; it just determines when the database is updated. (This
1956 means that when you upgrade Exim will forget old rate measurements.)
1958 Exim now checks that the per_* options are used with an update mode that
1959 makes sense for the current ACL. For example, when Exim is processing a
1960 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1961 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1962 must specify per_mail/readonly. If you omit the update mode it defaults to
1963 /leaky where that makes sense (as before) or /readonly where required.
1965 The /noupdate option is now undocumented but still supported for
1966 backwards compatibility. It is equivalent to /readonly except that in
1967 ACLs where /readonly is required you may specify /leaky/noupdate or
1968 /strict/noupdate which are treated the same as /readonly.
1970 A useful new feature is the /count= option. This is a generalization
1971 of the per_byte option, so that you can measure the throughput of other
1972 aggregate values. For example, the per_byte option is now equivalent
1973 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1975 The per_rcpt option has been generalized using the /count= mechanism
1976 (though it's more complicated than the per_byte equivalence). When it is
1977 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1978 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1979 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1980 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1981 Note that using per_rcpt with a non-readonly update mode in more than
1982 one ACL will cause the recipients to be double-counted. (The per_mail
1983 and per_byte options don't have this problem.)
1985 The handling of very low rates has changed slightly. If the computed rate
1986 is less than the event's count (usually one) then this event is the first
1987 after a long gap. In this case the rate is set to the same as this event's
1988 count, so that the first message of a spam run is counted properly.
1990 The major new feature is a mechanism for counting the rate of unique
1991 events. The new per_addr option counts the number of different
1992 recipients that someone has sent messages to in the last time period. It
1993 behaves like per_rcpt if all the recipient addresses are different, but
1994 duplicate recipient addresses do not increase the measured rate. Like
1995 the /count= option this is a general mechanism, so the per_addr option
1996 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1997 example, measure the rate that a client uses different sender addresses
1998 with the options per_mail/unique=$sender_address. There are further
1999 details in the main documentation.
2001 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2003 TF/06 Removed a few PCRE remnants.
2005 TF/07 Automatically extract Exim's version number from tags in the git
2006 repository when doing development or release builds.
2008 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2009 Bugzilla 879. Patch from Paul Fisher.
2011 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2012 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2015 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2017 PP/05 Variable $av_failed, true if the AV scanner deferred.
2018 Bugzilla 1078. Patch from John Horne.
2020 PP/06 Stop make process more reliably on build failure.
2021 Bugzilla 1087. Patch from Heiko Schlittermann.
2023 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2024 Bugzilla 1089. Patch from Heiko Schlittermann.
2026 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2027 Bugzilla 1131. Patch from Holger Weiß.
2029 PP/09 Handle IPv6 addresses with SPF.
2030 Bugzilla 860. Patch from Wolfgang Breyha.
2032 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2034 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2037 PP/11 match_* no longer expand right-hand-side by default.
2038 New compile-time build option, EXPAND_LISTMATCH_RHS.
2039 New expansion conditions, "inlist", "inlisti".
2041 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2043 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2045 PP/14 fix log_write() format string regression from TF/03.
2046 Bugzilla 1152. Patch from Dmitry Isaikin.
2052 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2054 PP/02 Harmonised TLS library version reporting; only show if debugging.
2055 Layout now matches that introduced for other libraries in 4.74 PP/03.
2057 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2059 PP/04 New "dns_use_edns0" global option.
2061 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2064 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2065 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2067 TK/01 Updated PolarSSL code to 0.14.2.
2068 Bugzilla 1097. Patch from Andreas Metzler.
2070 PP/07 Catch divide-by-zero in ${eval:...}.
2071 Fixes bugzilla 1102.
2073 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2076 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2077 format-string attack -- SECURITY: remote arbitrary code execution.
2079 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2080 time unintentionally subject to list matching rules, letting the header
2081 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2082 arbitrary string expansion). This allowed for information disclosure.
2084 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2085 INT_MIN/-1 -- value coerced to INT_MAX.
2091 NM/01 Workaround for PCRE version dependency in version reporting
2094 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2095 This fixes portability to compilers other than gcc, notably
2096 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2098 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2099 makefiles for portability to HP-UX and POSIX correctness.
2101 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2102 Also via indented variable definition in the Makefile.
2103 (Debugging by Oliver Heesakkers).
2105 PP/02 Restore caching of spamd results with expanded spamd_address.
2106 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2108 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2109 Improves build reliability. Fix from: Frank Elsner
2111 NM/02 Fix wide character breakage in the rfc2047 coding
2112 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2114 NM/03 Allow underscore in dnslist lookups
2115 Fixes bug 1026. Patch from Graeme Fowler
2117 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2118 Code patches from Adam Ciarcinski of NetBSD.
2120 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2123 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2124 is logged, to avoid truncation. Patch from John Horne.
2126 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2127 Patch from Jakob Hirsch.
2129 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2130 SQL string expansion failure details.
2131 Patch from Andrey Oktyabrski.
2133 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2134 Patch from Simon Arlott.
2136 PP/09 New lookups functionality failed to compile on old gcc which rejects
2137 extern declarations in function scope.
2138 Patch from Oliver Fleischmann
2140 PP/10 Use sig_atomic_t for flags set from signal handlers.
2141 Check getgroups() return and improve debugging.
2142 Fixed developed for diagnosis in bug 927 (which turned out to be
2145 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2146 Patch from Mark Zealey.
2148 PP/12 Bugzilla 1056: Improved spamd server selection.
2149 Patch from Mark Zealey.
2151 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2152 Based on patch from Heiko Schlittermann.
2154 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2155 Patch from Uwe Doering, sign-off by Michael Haardt.
2157 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2158 variable declaration deep within a block. Bug and patch from
2161 PP/15 lookups-Makefile IRIX compatibility coercion.
2163 PP/16 Make DISABLE_DKIM build knob functional.
2165 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2166 Patch by Simon Arlott
2168 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2169 variable argument macros. Our copy now differs from upstream.
2175 TF/01 Failure to get a lock on a hints database can have serious
2176 consequences so log it to the panic log.
2178 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2179 controlled using the smtp_confirmation log selector.
2181 TF/03 Include the error message when we fail to unlink a spool file.
2183 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2184 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2185 for maintaining out-of-tree patches for some time.
2187 PP/01 Bugzilla 139: Documentation and portability issues.
2188 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2189 Handle per-OS dynamic-module compilation flags.
2191 PP/02 Let /dev/null have normal permissions.
2192 The 4.73 fixes were a little too stringent and complained about the
2193 permissions on /dev/null. Exempt it from some checks.
2194 Reported by Andreas M. Kirchwitz.
2196 PP/03 Report version information for many libraries, including
2197 Exim version information for dynamically loaded libraries. Created
2198 version.h, now support a version extension string for distributors
2199 who patch heavily. Dynamic module ABI change.
2201 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2202 privilege escalation vulnerability whereby the Exim run-time user
2203 can cause root to append content of the attacker's choosing to
2206 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2209 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2210 If dropping privileges for untrusted macros, we disabled normal logging
2211 on the basis that it would fail; for the Exim run-time user, this is not
2212 the case, and it resulted in successful deliveries going unlogged.
2213 Fixed. Reported by Andreas Metzler.
2219 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2220 only prepend for the Resent-* case. Fixes regression introduced in
2221 Exim 4.70 by NM/22 for Bugzilla 607.
2223 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2224 increasing numbers of administrators be bitten by this.
2226 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2228 PP/03 Bugzilla 994: added openssl_options main configuration option.
2230 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2232 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2234 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2236 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2237 without a peer certificate, leading to a segfault because of an
2238 assumption that peers always have certificates. Be a little more
2239 paranoid. Problem reported by Martin Tscholak.
2241 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2242 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2243 NB: ClamAV planning to remove STREAM in "middle of 2010".
2244 CL also introduces -bmalware, various -d+acl logging additions and
2245 more caution in buffer sizes.
2247 PP/09 Implemented reverse_ip expansion operator.
2249 PP/10 Bugzilla 937: provide a "debug" ACL control.
2251 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2253 PP/12 Bugzilla 973: Implement --version.
2255 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2257 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2259 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2261 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2262 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2263 ignore trailing whitespace.
2265 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2267 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2270 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2271 Notification from Dr Andrew Aitchison.
2273 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2274 ExtendedDetectionInfo response format.
2275 Notification from John Horne.
2277 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2280 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2281 XSL and documented dependency on system catalogs, with examples of how
2284 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2287 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2288 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2289 configuration file which is writeable by the Exim user or group.
2291 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2292 of configuration files to cover files specified with the -C option if
2293 they are going to be used with root privileges, not just the default
2296 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2297 option (effectively making it always true).
2299 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2300 files to be used while preserving root privileges.
2302 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2303 that rogue child processes cannot use them.
2305 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2306 run-time user, instead of root.
2308 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2309 Exim run-time user without dropping privileges.
2311 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2312 result string, instead of calling string_vformat() twice with the same
2315 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2316 for other users. Others should always drop root privileges if they use
2317 -C on the command line, even for a whitelisted configure file.
2319 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2321 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2327 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2328 $data_path, and $header_path variables; fixed documentation bugs and
2331 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2332 exipick to access non-standard spools, including the "frozen" queue
2335 NM/01 Bugzilla 965: Support mysql stored procedures.
2336 Patch from Alain Williams
2338 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2340 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2341 Patch from Andreas Metzler
2343 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2344 Patch from Kirill Miazine
2346 NM/05 Bugzilla 671: Added umask to procmail example.
2348 JJ/03 installed exipick 20100323.0, fixing doc bug
2350 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2351 directory. Notification and patch from Dan Rosenberg.
2353 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2355 TK/02 Improve log output when DKIM signing operation fails.
2357 MH/01 Treat the transport option dkim_domain as a colon separated
2358 list, not as a single string, and sign the message with each element,
2359 omitting multiple occurences of the same signer.
2361 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2362 Bugzilla 985, 986. Patch by Simon Arlott
2364 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2365 Patch by Simon Arlott
2367 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2368 MBX locking. Notification from Dan Rosenberg.
2374 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2376 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2378 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2380 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2382 NM/04 Bugzilla 915: Flag broken perl installation during build.
2388 TK/01 Added patch by Johannes Berg that expands the main option
2389 "spamd_address" if it starts with a dollar sign.
2391 TK/02 Write list of recipients to X-Envelope-Sender header when building
2392 the mbox-format spool file for content scanning (suggested by Jakob
2395 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2396 (http://www.dcc-servers.net/) support via dccifd. Activated by
2397 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2399 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2400 by Mark Daniel Reidel <mr@df.eu>.
2402 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2403 When building exim an external PCRE library is now needed -
2404 PCRE is a system library on the majority of modern systems.
2405 See entry on PCRE_LIBS in EDITME file.
2407 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2408 conversation. Added nologin parameter to request.
2409 Patch contributed by Kirill Miazine.
2411 TF/01 Do not log submission mode rewrites if they do not change the address.
2413 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2415 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2416 log files in place. Contributed by Roberto Lima.
2418 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2420 TF/03 Bugzilla 615: When checking the local_parts router precondition
2421 after a local_part_suffix or local_part_prefix option, Exim now
2422 does not use the address's named list lookup cache, since this
2423 contains cached lookups for the whole local part.
2425 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2426 Robert Millan. Documentation is in experimental-spec.txt.
2428 TF/04 Bugzilla 668: Fix parallel build (make -j).
2430 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2432 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2433 Patch provided by Jan Srzednicki.
2435 TF/05 Leading white space used to be stripped from $spam_report which
2436 wrecked the formatting. Now it is preserved.
2438 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2439 that they are available at delivery time.
2441 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2443 TF/08 TLS error reporting now respects the incoming_interface and
2444 incoming_port log selectors.
2446 TF/09 Produce a more useful error message if an SMTP transport's hosts
2447 setting expands to an empty string.
2449 NM/06 Bugzilla 744: EXPN did not work under TLS.
2450 Patch provided by Phil Pennock.
2452 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2453 Patch provided by Richard Godbee.
2455 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2456 acl_smtp_notquit, added index entry.
2458 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2459 Patch provided by Eugene Bujak.
2461 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2462 accept(). Patch provided by Maxim Dounin.
2464 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2465 Patch provided by Phil Pennock.
2467 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2469 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2470 Patch provided by Brad "anomie" Jorsch.
2472 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2473 Patch provided by Dean Brooks.
2475 TK/05 Add native DKIM support (does not depend on external libraries).
2477 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2478 Patch provided by Graeme Fowler.
2480 NM/16 Bugzilla 851: Documentation example syntax fix.
2482 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2484 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2487 NM/19 Bugzilla 745: TLS version reporting.
2488 Patch provided by Phil Pennock.
2490 NM/20 Bugzilla 167: bool: condition support.
2491 Patch provided by Phil Pennock.
2493 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2494 clients. Patch provided by Phil Pennock.
2496 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2497 Patch provided by Brad "anomie" Jorsch.
2499 NM/23 Bugzilla 687: Fix misparses in eximstats.
2500 Patch provided by Heiko Schlittermann.
2502 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2503 Patch provided by Heiko Schlittermann.
2505 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2506 plus update to original patch.
2508 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2510 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2511 Patch provided by David Brownlee.
2513 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2515 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2517 NM/30 Bugzilla 888: TLS documentation bugfixes.
2519 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2521 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2522 Unlike the original bugzilla I have changed all shell scripts in src tree.
2524 NM/33 Bugzilla 898: Transport filter timeout fix.
2525 Patch by Todd Rinaldo.
2527 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2528 Patch by Serge Demonchaux.
2530 NM/35 Bugzilla 39: Base64 decode bug fixes.
2531 Patch by Jakob Hirsch.
2533 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2535 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2537 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2539 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2545 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2546 ALT-N's libdkim that I have put here:
2547 http://duncanthrax.net/exim-experimental/
2549 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2550 were called 'true' and it seems that C99 defines that as a reserved
2551 keyword to be used with 'bool' variable types. That means you could
2552 not include C99-style headers which use bools without triggering
2553 build errors in sieve.c.
2555 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2556 as mailq or other aliases. Changed the --help handling significantly
2557 to do whats expected. exim_usage() emits usage/help information.
2559 SC/01 Added the -bylocaldomain option to eximstats.
2561 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2563 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2565 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2571 PH/01 Another patch from the Sieve maintainer.
2573 PH/02 When an IPv6 address is converted to a string for single-key lookup
2574 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2575 dots are used instead of colons so that keys in lsearch files need not
2576 contain colons. This was done some time before quoting was made available
2577 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2578 (notated using the quote facility) so as to distinguish them from IPv4
2579 keys. This meant that lookups for IP addresses in host lists did not work
2580 for iplsearch lookups.
2582 This has been fixed by arranging for IPv6 addresses to be expressed with
2583 colons if the lookup type is iplsearch. This is not incompatible, because
2584 previously such lookups could never work.
2586 The situation is now rather anomalous, since one *can* have colons in
2587 ordinary lsearch keys. However, making the change in all cases is
2588 incompatible and would probably break a number of configurations.
2590 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2593 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2594 conversion specification without a maximum field width, thereby enabling
2595 a rogue spamd server to cause a buffer overflow. While nobody in their
2596 right mind would setup Exim to query an untrusted spamd server, an
2597 attacker that gains access to a server running spamd could potentially
2598 exploit this vulnerability to run arbitrary code as the Exim user.
2600 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2601 $primary_hostname instead of what libspf2 thinks the hosts name is.
2603 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2604 a directory entry by the name of the lookup key. Previously, if a
2605 symlink pointed to a non-existing file or a file in a directory that
2606 Exim lacked permissions to read, a lookup for a key matching that
2607 symlink would fail. Now it is enough that a matching directory entry
2608 exists, symlink or not. (Bugzilla 503.)
2610 PH/03 The body_linecount and body_zerocount variables are now exported in the
2613 PH/04 Added the $dnslist_matched variable.
2615 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2616 This means they are set thereafter only if the connection becomes
2619 PH/06 Added the client_condition to authenticators so that some can be skipped
2620 by clients under certain conditions.
2622 PH/07 The error message for a badly-placed control=no_multiline_responses left
2623 "_responses" off the end of the name.
2625 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2627 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2628 (without spaces) instead of just copying the configuration text.
2630 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2632 PH/11 Added $max_received_linelength.
2634 PH/12 Added +ignore_defer and +include_defer to host lists.
2636 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2637 way in which PCRE > 7.0 is built.
2639 PH/14 Implemented queue_only_load_latch.
2641 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2642 MAIL command. The effect was to mangle the value on 64-bit systems.
2644 PH/16 Another patch from the Sieve maintainer.
2646 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2648 PH/18 If a system quota error occurred while trying to create the file for
2649 a maildir delivery, the message "Mailbox is full" was not appended to the
2650 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2651 only to a quota excession during the actual writing of the file.
2653 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2654 characters?) which causes problems in log lines. The DN values are now
2655 passed through string_printing() before being added to log lines.
2657 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2658 and InterBase are left for another time.)
2660 PH/21 Added message_body_newlines option.
2662 PH/22 Guard against possible overflow in moan_check_errorcopy().
2664 PH/23 POSIX allows open() to be a macro; guard against that.
2666 PH/24 If the recipient of an error message contained an @ in the local part
2667 (suitably quoted, of course), incorrect values were put in $domain and
2668 $local_part during the evaluation of errors_copy.
2674 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2675 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2678 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2679 issue a MAIL command.
2681 PH/02 In an ACL statement such as
2683 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2685 if a client was not listed at all, or was listed with a value other than
2686 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2687 the condition was not true (as it should be), so access was not denied.
2688 The bug was that the ! inversion was incorrectly passed on to the second
2689 item. This has been fixed.
2691 PH/03 Added additional dnslists conditions == and =& which are different from
2692 = and & when the dns lookup returns more than one IP address.
2694 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2695 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2697 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2698 FSYNC, which compiles an option called disable_fsync that allows for
2699 bypassing fsync(). The documentation is heavily laced with warnings.
2701 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2703 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2704 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2705 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2706 including adding "make clean"; (3) Added -fPIC when compiling the test
2707 dynamically loaded module, to get rid of a warning.
2709 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2710 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2711 = 0s. The bug is otherwise harmless.
2713 PH/07 There was a bug in the dovecot authenticator such that the value of
2714 $auth1 could be overwritten, and so not correctly preserved, after a
2715 successful authentication. This usually meant that the value preserved by
2716 the server_setid option was incorrect.
2718 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2720 PH/09 Installed PCRE release 7.0.
2722 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2723 run for batched SMTP input. It is now run at the start of every message
2724 in the batch. While fixing this I discovered that the process information
2725 (output by running exiwhat) was not always getting set for -bs and -bS
2726 input. This is fixed, and it now also says "batched" for BSMTP.
2728 PH/11 Added control=no_pipelining.
2730 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2731 patch, slightly modified), and move the expansion of helo_data till after
2732 the connection is made in the smtp transport (so it can use these
2735 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2737 PH/14 Added log_selector = +pid.
2739 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2741 PH/16 Add ${if forany and ${if forall.
2743 PH/17 Added dsn_from option to vary the From: line in DSNs.
2745 PH/18 Flush SMTP output before performing a callout, unless control =
2746 no_callout_flush is set.
2748 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2749 was true (the default) a successful delivery failed to delete the retry
2750 item, thus causing premature timeout of the address. The bug is now
2753 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2755 PH/21 Long custom messages for fakedefer and fakereject are now split up
2756 into multiline responses in the same way that messages for "deny" and
2757 other ACL rejections are.
2759 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2760 with slight modification.
2762 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2763 draft, changing the syntax and factoring some duplicate code".
2765 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2766 for deliveries of the second and subsequent messages over the same SMTP
2769 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2770 ${reduce, with only minor "tidies".
2772 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2774 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2775 expansion side effects.
2777 PH/27 When a message times out after an over-quota error from an Exim-imposed
2778 quota, the bounce message says "mailbox is full". This message was not
2779 being given when it was a system quota that was exceeded. It now should
2782 MH/03 Made $recipients available in local_scan(). local_scan() already has
2783 better access to the recipient list through recipients_list[], but
2784 $recipients can be useful in postmaster-provided expansion strings.
2786 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2787 in the case of a MAIL command with additional options following the
2788 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2789 were accidentally chopped off.
2791 PH/29 SMTP synchronization checks are implemented when a command is read -
2792 there is a check that no more input is waiting when there shouldn't be
2793 any. However, for some commands, a delay in an ACL can mean that it is
2794 some time before the response is written. In this time, more input might
2795 arrive, invalidly. So now there are extra checks after an ACL has run for
2796 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2797 pipelining has not been advertised.
2799 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2801 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2802 returned in the SMTP response when smtp_return_error_details was false.
2803 This has been fixed.
2805 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2806 instead of the C I/O that was originally supplied, because problems were
2807 reported on Solaris.
2809 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2810 Exim which did not show up earlier: it was assuming that a call to
2811 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2812 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2813 was a macro that became an assignment, so it seemed to work. This has
2814 changed to a proper function call with a void return, hence the compile
2815 error. Exim's code has been fixed.
2817 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2820 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2822 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2824 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2825 because it turns out to be too big or there is a timeout) there is no
2826 "Completed" line in the log. When some messages of this type were
2827 selected by exigrep, they were listed as "not completed". Others were
2828 picked up by some special patterns. I have improved the selection
2829 criteria to be more general.
2831 PH/38 The host_find_failed option in the manualroute router can now be set
2832 to "ignore", to completely ignore a host whose IP address cannot be
2833 found. If all hosts are ignored, the behaviour is controlled by the new
2834 host_all_ignored option.
2836 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2837 homing or because of multiple MX records with /mx) generated more than
2838 one IP address, and the following item turned out to be the local host,
2839 all the secondary addresses of the first item were incorrectly removed
2840 from the list, along with the local host and any following hosts (which
2841 is what is supposed to happen).
2843 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2844 whoever called Exim into the -H file. In the case of the daemon it was
2845 behaving confusingly. When first started, it used values for whoever
2846 started the daemon, but after a SIGHUP it used the Exim user (because it
2847 calls itself on a restart). I have changed the code so that it now always
2850 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2851 message are rejected with the same error (e.g. no authentication or bad
2852 sender address), and a DATA command is nevertheless sent (as can happen
2853 with PIPELINING or a stupid MUA), the error message that was given to the
2854 RCPT commands is included in the rejection of the DATA command. This is
2855 intended to be helpful for MUAs that show only the final error to their
2858 PH/42 Another patch from the Sieve maintainer.
2860 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2861 Eximstats - Fixed some broken HTML links and added missing column headers
2863 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2864 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2866 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2868 PH/43 Yet another patch from the Sieve maintainer.
2870 PH/44 I found a way to check for a TCP/IP connection going away before sending
2871 the response to the final '.' that terminates a message, but only in the
2872 case where the client has not sent further data following the '.'
2873 (unfortunately, this is allowed). However, in many cases there won't be
2874 any further data because there won't be any more messages to send. A call
2875 to select() can be used: if it shows that the input is "ready", there is
2876 either input waiting, or the socket has been closed. An attempt to read
2877 the next input character can distinguish the two cases. Previously, Exim
2878 would have sent an OK response which the client would never have see.
2879 This could lead to message repetition. This fix should cure that, at
2880 least in a lot of common cases.
2882 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2883 advertised in response to EHLO.
2889 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2890 fixed by 4.65/MH/01 (is this a record?) are fixed:
2892 (i) An empty string was always treated as zero by the numeric comparison
2893 operators. This behaviour has been restored.
2895 (ii) It is documented that the numeric comparison operators always treat
2896 their arguments as decimal numbers. This was broken in that numbers
2897 starting with 0 were being interpreted as octal.
2899 While fixing these problems I realized that there was another issue that
2900 hadn't been noticed. Values of message_size_limit (both the global option
2901 and the transport option) were treated as octal if they started with 0.
2902 The documentation was vague. These values are now always treated as
2903 decimal, and I will make that clear in the documentation.
2909 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2910 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2913 MH/01 Don't check that the operands of numeric comparison operators are
2914 integers when their expansion is in "skipping" mode (fixes bug
2915 introduced by 4.64-PH/07).
2917 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2918 child addresses, Exim now panics and dies. Previously, because the count
2919 is held in a short int, deliveries were likely to be lost. As such a
2920 large number of recipients for a single message is ridiculous
2921 (performance will be very, very poor), I have chosen to impose a limit
2922 rather than extend the field.
2928 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2929 leftover -K file (the existence of which was triggered by #402).
2930 While we were at it, introduced process PID as part of the -K
2931 filename. This should rule out race conditions when creating
2934 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2935 processing considerably. Previous code took too long for large mails,
2936 triggering a timeout which in turn triggers #401.
2938 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2939 in the DK code in transports.c. sendfile() is not really portable,
2940 hence the _LINUX specificness.
2942 TF/01 In the add_headers option to the mail command in an Exim filter,
2943 there was a bug that Exim would claim a syntax error in any
2944 header after the first one which had an odd number of characters
2947 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2948 callout verification, Exim cached a "reject" for the entire domain. This
2949 is correct for most verifications, but it is not correct for a recipient
2950 verification with use_sender or use_postmaster set, because in that case
2951 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2952 case of MAIL FROM:<> rejection from other early rejections (e.g.
2953 rejection of HELO). When verifying a recipient using a non-null MAIL
2954 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2955 Whatever the result of the callout, the value of the domain cache is
2956 left unchanged (for any other kind of callout, getting as far as trying
2957 RCPT means that the domain itself is ok).
2959 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2962 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2963 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2964 session, the daemon ignores ECONNECT errors and logs others; it now
2965 ignores EPIPE as well.
2967 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2968 (quoted-printable decoding).
2970 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2971 later the small subsequent patch to fix an introduced bug.
2973 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2975 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2977 PH/08 An error is now given if message_size_limit is specified negative.
2979 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2980 to be given (somewhat) arbitrary names.
2982 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2985 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2986 miscellaneous code fixes
2988 PH/10 Added the log_reject_target ACL modifier to specify where to log
2991 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2992 hostname. This is wrong, because it relates to the incoming message (and
2993 probably the interface on which it is arriving) and not to the outgoing
2994 callout (which could be using a different interface). This has been
2995 changed to use the value of the helo_data option from the smtp transport
2996 instead - this is what is used when a message is actually being sent. If
2997 there is no remote transport (possible with a router that sets up host
2998 addresses), $smtp_active_hostname is used.
3000 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3001 tweaks were necessary in order to get it to work (see also 21 below):
3002 (a) The code assumed that strncpy() returns a negative number on buffer
3003 overflow, which isn't the case. Replaced with Exim's string_format()
3005 (b) There were several signed/unsigned issues. I just did the minimum
3006 hacking in of casts. There is scope for a larger refactoring.
3007 (c) The code used strcasecmp() which is not a standard C function.
3008 Replaced with Exim's strcmpic() function.
3009 (d) The code set only $1; it now sets $auth1 as well.
3010 (e) A simple test gave the error "authentication client didn't specify
3011 service in request". It would seem that Dovecot has changed its
3012 interface. Fortunately there's a specification; I followed it and
3013 changed what the client sends and it appears to be working now.
3015 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3018 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3019 address A is aliased to B and C, where B exists and C does not. Without
3020 -v the output is "A verified" because verification stops after a
3021 successful redirection if more than one address is generated. However,
3022 with -v the child addresses are also verified. Exim was outputting "A
3023 failed to verify" and then showing the successful verification for C,
3024 with its parentage. It now outputs "B failed to verify", showing B's
3025 parentage before showing the successful verification of C.
3027 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3028 look up a TXT record in a specific list after matching in a combined
3031 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3032 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3033 they consult the DNS. I had assumed they would set it the way they
3034 wanted; and indeed my experiments on Linux seem to show that in some
3035 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3036 To be on the safe side, however, I have now made the interface to
3037 host_find_byname() similar to host_find_bydns(), with an argument
3038 containing the DNS resolver options. The host_find_byname() function now
3039 sets these options at its start, just as host_find_bydns() does. The smtp
3040 transport options dns_qualify_single and dns_search_parents are passed to
3041 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3042 of host_find_byname() use the default settings of RES_DEFNAMES
3043 (qualify_single) but not RES_DNSRCH (search_parents).
3045 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3046 spool_read_header() do less string testing, by means of a preliminary
3047 switch on the second character of optional "-foo" lines. (This is
3048 overdue, caused by the large number of possibilities that now exist.
3049 Originally there were few.) While I was there, I also converted the
3050 str(n)cmp tests so they don't re-test the leading "-" and the first
3051 character, in the hope this might squeeze out yet more improvement.
3053 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3054 flag allowing group syntax was set by the header_syntax check but not
3055 turned off, possible causing trouble later; (2) The flag was not being
3056 set at all for the header_verify test, causing "group"-style headers to
3057 be rejected. I have now set it in this case, and also caused header_
3058 verify to ignore an empty address taken from a group. While doing this, I
3059 came across some other cases where the code for allowing group syntax
3060 while scanning a header line wasn't quite right (mostly, not resetting
3061 the flag correctly in the right place). These bugs could have caused
3062 trouble for malformed header lines. I hope it is now all correct.
3064 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3065 with the "reply" argument non-NULL. The code, however (which originally
3066 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3067 but it didn't always do it. This confused somebody who was copying the
3068 code for some other use. I have removed all the tests.
3070 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3071 feature that was used to support insecure browsers during the U.S. crypto
3072 embargo. It requires special client support, and Exim is probably the
3073 only MTA that supported it -- and would never use it because real RSA is
3074 always available. This code has been removed, because it had the bad
3075 effect of slowing Exim down by computing (never used) parameters for the
3076 RSA_EXPORT functionality.
3078 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3079 authenticator to fail if there's a tab character in the incoming data
3080 (there should never be unless someone is messing about, as it's supposed
3081 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3082 if the connection is using TLS or if the remote IP is the same as the
3083 local IP, and the "valid-client-cert option" if a client certificate has
3086 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3087 authenticators. This can be used for authorization after authentication
3088 succeeds. (In the case of plaintext, it servers for both authentication
3091 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3092 if any retry times were supplied.
3094 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3095 connection, where there is no client IP address to check. In this
3096 situation, the verify now always succeeds.
3098 PH/25 Applied John Jetmore's -Mset patch.
3100 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3102 PH/27 In a string expansion for a processed (not raw) header when multiple
3103 headers of the same name were present, leading whitespace was being
3104 removed from all of them, but trailing whitespace was being removed only
3105 from the last one. Now trailing whitespace is removed from each header
3106 before concatenation. Completely empty headers in a concatenation (as
3107 before) are ignored.
3109 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3110 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3112 PH/29 [Removed. This was a change that I later backed out, and forgot to
3113 correct the ChangeLog entry (that I had efficiently created) before
3114 committing the later change.]
3116 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3117 address errors (4xx response to RCPT) over the same connection as other
3118 messages routed to the same hosts. Such deliveries are always "forced",
3119 so retry times are not inspected. This resulted in far too many retries
3120 for the affected addresses. The effect occurred only when there were more
3121 hosts than the hosts_max_try setting in the smtp transport when it had
3122 the 4xx errors. Those hosts that it had tried were not added to the list
3123 of hosts for which the message was waiting, so if all were tried, there
3124 was no problem. Two fixes have been applied:
3126 (i) If there are any address or message errors in an SMTP delivery, none
3127 of the hosts (tried or untried) are now added to the list of hosts
3128 for which the message is waiting, so the message should not be a
3129 candidate for sending over the same connection that was used for a
3130 successful delivery of some other message. This seems entirely
3131 reasonable: after all the message is NOT "waiting for some host".
3132 This is so "obvious" that I'm not sure why it wasn't done
3133 previously. Hope I haven't missed anything, but it can't do any
3134 harm, as the worst effect is to miss an optimization.
3136 (ii) If, despite (i), such a delivery is accidentally attempted, the
3137 routing retry time is respected, so at least it doesn't keep
3138 hammering the server.
3140 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3141 in ${readsocket because some servers need this prod.
3143 PH/32 Added some extra debug output when updating a wait-xxx database.
3145 PH/33 The hint "could be header name not terminated by colon", which has been
3146 given for certain expansion errors for a long time, was not being given
3147 for the ${if def:h_colon_omitted{... case.
3149 PH/34 The spec says: "With one important exception, whenever a domain list is
3150 being scanned, $domain contains the subject domain." There was at least
3151 one case where this was not true.
3153 PH/35 The error "getsockname() failed: connection reset by peer" was being
3154 written to the panic log as well as the main log, but it isn't really
3155 panic-worthy as it just means the connection died rather early on. I have
3156 removed the panic log writing for the ECONNRESET error when getsockname()
3159 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3160 runs only) independently of the message's sender address. This meant
3161 that, if the 4xx error was in fact related to the sender, a different
3162 message to the same recipient with a different sender could confuse
3163 things. In particular, this can happen when sending to a greylisting
3164 server, but other circumstances could also provoke similar problems.
3165 I have changed the default so that the retry time for these errors is now
3166 based a combination of the sender and recipient addresses. This change
3167 can be overridden by setting address_retry_include_sender=false in the
3170 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3171 remote server are returned as part of bounce messages. This was not
3172 happening for LMTP over a pipe (the lmtp transport), but now it is the
3173 same for both kinds of LMTP.
3175 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3176 in header lines that were in fact CNAMEs. This is no longer the case.
3178 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3179 and queue runs started by the daemon processed all messages. This has
3180 been fixed so that -R and -S can now usefully be given with -q<time>.
3182 PH/40 Import PCRE release 6.7 (fixes some bugs).
3184 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3186 PH/42 Give an error if -q is specified more than once.
3188 PH/43 Renamed the variables $interface_address and $interface_port as
3189 $received_ip_address and $received_port, to make it clear that these
3190 values apply to message reception, and not to the outgoing interface when
3191 a message is delivered. (The old names remain recognized, of course.)
3193 PH/44 There was no timeout on the connect() call when using a Unix domain
3194 socket in the ${readsocket expansion. There now is.
3196 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3197 be meaningful with "accept".
3199 SC/01 Eximstats V1.43
3200 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3202 SC/02 Eximstats V1.44
3203 Use a glob alias rather than an array ref in the generated
3204 parser. This improves both readability and performance.
3206 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3207 Collect SpamAssassin and rejection statistics.
3208 Don't display local sender or destination tables unless
3209 there is data to show.
3210 Added average volumes into the top table text output.
3212 SC/04 Eximstats V1.46
3213 Collect data on the number of addresses (recipients)
3214 as well as the number of messages.
3216 SC/05 Eximstats V1.47
3217 Added 'Message too big' to the list of mail rejection
3218 reasons (thanks to Marco Gaiarin).
3220 SC/06 Eximstats V1.48
3221 Mainlog lines which have GMT offsets and are too short to
3222 have a flag are now skipped.
3224 SC/07 Eximstats V1.49 (Alain Williams)
3225 Added the -emptyok flag.
3227 SC/08 Eximstats V1.50
3228 Fixes for obtaining the IP address from reject messages.
3230 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3231 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3232 whitespace changes from 4.64-PH/27
3234 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3237 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3238 are found, allow negative numbers in numeric criteria)
3240 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3242 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3243 to match changes made in 4.64-PH/43
3245 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3247 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3248 "domain" is actually an IP address, and force a failure. This locks out
3249 those revolvers/nameservers that support "A-for-A" lookups, in
3250 contravention of the specifications.
3252 PH/48 When a host name was looked up from an IP address, and the subsequent
3253 forward lookup of the name timed out, the host name was left in
3254 $sender_host_name, contrary to the specification.
3256 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3257 restricted to single-key lookups, Exim was not diagnosing an error if
3258 * or *@ was used with a query-style lookup.
3260 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3262 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3263 long ago, but noone interested enough thought of it. Let's just say that
3264 the "1.1" means that there are some new functions that weren't there at
3265 some point in the past.
3267 PH/51 Error processing for expansion failure of helo_data from an smtp
3268 transport during callout processing was broken.
3270 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3271 tested/used via the -bh/-bhc/-bs options.
3273 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3274 bug, fixed in subsequent PCRE releases).
3276 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3277 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3279 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3285 SC/01 Use a glob alias rather than an array ref in eximstats generated
3286 parser. This improves both readability and performance.
3288 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3289 Don't display local sender or destination tables in eximstats unless
3290 there is data to show.
3291 Added average volumes into the eximstats top table text output.
3293 SC/03 Collect data on the number of addresses (recipients) as well
3294 as the number of messages in eximstats.
3296 TF/01 Correct an error in the documentation for the redirect router. Exim
3297 does (usually) call initgroups() when daemonizing.
3299 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3300 with consistent privilege compared to when running as a daemon.
3302 TF/03 Note in the spec that $authenticated_id is not set for local
3303 submissions from trusted users.
3305 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3306 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3308 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3309 by adding some example configuration directives to the default
3310 configuration file. A little bit of work is required to uncomment the
3311 directives and define how usernames and passwords are checked, but
3312 there is now a framework to start from.
3314 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3315 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3316 without this. I don't know how relevant this is to other LDAP libraries.
3318 PH/02 Add the verb name to the "unknown ACL verb" error.
3320 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3322 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3324 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3325 directory not expanded when it should be if an expanded home directory
3326 was set for the address (which is overridden by the transport).
3328 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3331 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3332 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3333 because it is too late at that time, and has no effect.
3335 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3336 security issue with \' (bugzilla #107). I could not use the
3337 PQescapeStringConn() function, because it needs a PGconn value as one of
3340 PH/08 When testing addresses using -bt, indicate those final addresses that
3341 are duplicates that would not cause an additional delivery. At least one
3342 person was confused, thinking that -bt output corresponded to deliveries.
3343 (Suppressing duplicates isn't a good idea as you lose the information
3344 about possibly different redirections that led to the duplicates.)
3346 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3347 systems where poll() doesn't work, in particular OS X.
3349 PH/10 Added more information to debugging output for retry time not reached.
3351 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3352 operations in malware.c.
3354 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3357 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3358 syslog_duplication set false, log lines that would normally be written
3359 both the the main log and to the reject log were not written to syslog at
3362 PH/14 In the default configuration, change the use of "message" in ACL warn
3363 statements to "add_header".
3365 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3366 not followed by a command (e.g. "seen endif").
3368 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3369 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3372 PH/17 Added extra conditions to the default value of delay_warning_condition
3376 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3377 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3378 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3381 The Auto-Submitted: and various List- headers are standardised, whereas I
3382 don't think Precedence: ever was.
3384 PH/18 Refactored debugging code in route_finduser() to show more information,
3385 in particular, the error code if getpwnam() issues one.
3387 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3388 This is apparently needed in addition to the PH/07 change above to avoid
3389 any possible encoding problems.
3391 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3392 but not after initializing Perl.
3394 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3395 output them only if debugging. By default they are written stderr,
3396 apparently, which is not desirable.
3398 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3401 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3404 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3406 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3407 authenticated or an ident call has been made. Suppress the default
3408 values for $authenticated_id and $authenticated_sender (but permit -oMai
3409 and -oMas) when testing with -bh.
3411 PH/24 Re-jigged the order of the tests in the default configuration so that the
3412 tests for valid domains and recipients precede the DNS black list and CSA
3413 tests, on the grounds that those ones are more expensive.
3415 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3416 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3417 This bug exists in every version of Exim that I still have, right back to
3420 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3421 However, an attempt to turn on case-sensitivity in a regex key by
3422 including (?-i) didn't work because the subject string was already
3423 lowercased, and the effects were non-intuitive. It turns out that a
3424 one-line patch can be used to allow (?-i) to work as expected.
3430 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3431 other effects) broke the use of negated acl sub-conditions.
3433 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3436 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3437 "Deny" causes Exim to reject the incoming connection with a 554 error.
3438 Unfortunately, if there is a major crisis, such as a disk failure,
3439 tcp-wrappers gives "deny", whereas what one would like would be some
3440 kind of temporary error. A kludge has been added to help with this.
3441 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3442 554 error is used if errno is still zero or contains ENOENT (which occurs
3443 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3446 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3448 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3449 errors. Otherwise a message that provokes a temporary error (when other
3450 messages do not) can cause a whole host to time out.
3452 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3453 addresses were routed directly to files or pipes from a redirect router.
3454 File deliveries just didn't batch; pipe deliveries might have suffered
3457 PH/06 A failure to get a lock for a hints database would erroneously always say
3458 "Failed to get write lock", even when it was really a read lock.
3460 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3461 of 0600. This has been changed to use the value of the lockfile_mode
3462 option (which defaults to 0600).
3464 PH/08 Applied small patch from the Sieve maintainer.
3466 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3467 folder from quota calculations, a direct delivery into this folder messed
3468 up the contents of the maildirsize file. This was because the regex was
3469 used only to exclude .Trash (or whatever) when the size of the mailbox
3470 was calculated. There was no check that a delivery was happening into an
3471 excluded directory. This bug has been fixed by ignoring all quota
3472 processing for deliveries into excluded directories.
3474 PH/10 Added the maildirfolder_create_regex option to appendfile.
3480 PH/01 The code for finding all the local interface addresses on a FreeBSD
3481 system running IPv6 was broken. This may well have applied to all BSD
3482 systems, as well as to others that have similar system calls. The broken
3483 code found IPv4 interfaces correctly, but gave incorrect values for the
3484 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3485 that it would not match correctly against @[] and not recognize the IPv6
3488 PH/02 The ipliteral router was not recognizing addresses of the form user@
3489 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3491 PH/03 Added disable_ipv6.
3493 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3494 decoded form, because it is most often used to construct To: headers
3495 lines in autoreplies, and the decoded form may well be syntactically
3496 invalid. However, $reply_address has leading white space removed, and all
3497 newlines turned into spaces so that the autoreply transport does not
3500 PH/05 If group was specified without a user on a router, and no group or user
3501 was specified on a transport, the group from the router was ignored.
3503 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3504 for visible compile-time settings that can be used to change these
3505 numbers, for those that want even more. Backwards compatibility with old
3506 spool files has been maintained. However, going back to a previous Exim
3507 release will lost any variables that are in spool files.
3509 PH/07 Two small changes when running in the test harness: increase delay when
3510 passing a TCP/IP connection to a new process, in case the original
3511 process has to generate a bounce, and remove special handling of
3512 127.0.0.2 (sic), which is no longer necessary.
3514 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3515 be the same on different OS.
3517 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3520 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3521 whether --show-vars was specified or not
3523 JJ/02 exipick: Added support for new ACL variable spool format introduced
3526 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3527 syntactically invalid From: or Reply-to: line, and a filter used this to
3528 generate an autoreply, and therefore failed to obtain an address for the
3529 autoreply, Exim could try to deliver to a non-existent relative file
3530 name, causing unrelated and misleading errors. What now happens is that
3531 it logs this as a hard delivery error, but does not attempt to create a
3534 PH/11 The exinext utility has a -C option for testing purposes, but although
3535 the given file was scanned by exinext itself; it wasn't being passed on
3536 when Exim was called.
3538 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3539 an end-of-file indication when reading a command response.
3541 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3542 compiled. In many other places in Exim, IPv6 addresses are always
3543 recognized, so I have changed this. It also means that IPv4 domain
3544 literals of the form [IPV4:n.n.n.n] are now always recognized.
3546 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3547 used if the router is not running as root, for example, when verifying at
3548 ACL time, or when using -bh. The debugging output from this situation was
3549 non-existent - all you got was a failure to exec. I have made two
3552 (a) Failures to set uid/gid, the current directory, or a process leader
3553 in a subprocess such as that created by queryprogram now generate
3554 suitable debugging output when -d is set.
3556 (b) The queryprogram router detects when it is not running as root,
3557 outputs suitable debugging information if -d is set, and then runs
3558 the subprocess without attempting to change uid/gid.
3560 PH/15 Minor change to Makefile for building test_host (undocumented testing
3563 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3564 additional section of a DNS packet that returns MX or SRV records.
3565 Instead, it always explicitly searches for A/AAAA records. This avoids
3566 major problems that occur when a DNS server includes only records of one
3567 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3568 fixed another bug: if SRV records were looked up and the corresponding
3569 address records were *not* found in the additional section, the port
3570 values from the SRV records were lost.
3572 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3573 using the correct key (the original address) when searching the retry
3574 rules in order to find which one to use for generating the retry hint.
3576 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3577 adding the default one. Similarly, if it contains a Reply-To: header, the
3578 errors_reply_to option, if set, is not used.
3580 PH/19 When calculating a retry time, Exim used to measure the "time since
3581 failure" by looking at the "first failed" field in the retry record. Now
3582 it does not use this if it is later than than the arrival time of the
3583 message. Instead it uses the arrival time. This makes for better
3584 behaviour in cases where some deliveries succeed, thus re-setting the
3585 "first failed" field. An example is a quota failure for a huge message
3586 when small messages continue to be delivered. Without this change, the
3587 "time since failure" will always be short, possible causing more frequent
3588 delivery attempts for the huge message than are intended.
3589 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3591 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3592 $1, $2, $3) because the numerical variables can be reset during some
3593 expansion items (e.g. "match"), thereby losing the authentication data.
3595 PH/21 Make -bV show the size of off_t variables so that the test suite can
3596 decide whether to run tests for quotas > 2G.
3598 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3599 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3600 filecount value is greater than 2G or if a quota value is greater than 2G
3601 on a system where the size of off_t is not greater than 4, a panic error
3604 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3605 never match. The debug and -bh output now contains an explicit error
3606 message indicating a malformed IPv4 address or mask.
3608 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3609 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3610 PH/23 above applies.
3612 PH/25 Do not write to syslog when running in the test harness. The only
3613 occasion when this arises is a failure to open the main or panic logs
3614 (for which there is an explicit test).
3616 PH/26 Added the /no_tell option to "control=freeze".
3618 PH/27 If a host name lookup failed very early in a connection, for example, if
3619 the IP address matched host_lookup and the reverse lookup yielded a name
3620 that did not have a forward lookup, an error message of the form "no IP
3621 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3622 could be logged. Now it outputs the IP address instead of "NULL".
3624 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3625 allows the sender and the authenticated sender to be set when
3626 submitting a message from within Exim. Since child_open_exim() is
3627 documented for local_scan(), the new function should be too.
3629 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3630 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3631 results in an empty string is now treated as unset.
3633 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3635 PH/31 Added sender_verify_fail logging option.
3637 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3638 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3639 tidied the source and removed it altogether.
3641 PH/33 When a queue run was abandoned because the load average was too high, a
3642 log line was always written; now it is written only if the queue_run log
3643 selector is set. In addition, the log line for abandonment now contains
3644 information about the queue run such as the pid. This is always present
3645 in "start" and "stop" lines but was omitted from the "abandon" line.
3647 PH/34 Omit spaces between a header name and the colon in the error message that
3648 is given when verify = headers_syntax fails (if there are lots of them,
3649 the message gets confusing).
3651 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3652 names, as there are now some PTR records that contain slashes. This check
3653 is only to protect against broken name servers that fall over on strange
3654 characters, so the fact that it applies to all lookups doesn't matter.
3656 PH/36 Now that the new test suite is complete, we can remove some of the
3657 special code in Exim that was needed for the old test suite. For example,
3658 sorting DNS records because real resolvers return them in an arbitrary
3659 order. The new test suite's fake resolver always returns records in the
3662 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3663 bounces) except when queue_only is set, to avoid logging races between
3664 the different processes.
3666 PH/38 Panic-die if .include specifies a non-absolute path.
3668 PH/39 A tweak to the "H" retry rule from its user.
3670 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3671 a label. They prevented compilation on older perls.
3673 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3674 a warning to be raised on newish perls.
3676 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3677 on queue. Changes to match documented behaviour of showing count of
3678 messages matching specified criteria.
3680 PH/40 Changed the default ident timeout from 30s to 5s.
3682 PH/41 Added support for the use of login_cap features, on those BSD systems
3683 that have them, for controlling the resources used by pipe deliveries.
3685 PH/42 The content-scanning code uses fopen() to create files in which to put
3686 message data. Previously it was not paying any attention to the mode of
3687 the files. Exim runs with umask(0) because the rest of the code creates
3688 files with open(), and sets the required mode explicitly. Thus, these
3689 files were ending up world-writeable. This was not a big issue, because,
3690 being within the spool directory, they were not world-accessible. I have
3691 created a function called modefopen, which takes an additional mode
3692 argument. It sets umask(777), creates the file, chmods it to the required
3693 mode, then resets the umask. All the relevant calls to fopen() in the
3694 content scanning code have been changed to use this function.
3696 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3697 to 24 hours. This avoids potential overflow problems when processing G
3698 and H retry rules. I suspect nobody ever tinkers with this value.
3700 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3702 PH/45 When the plaintext authenticator is running as a client, the server's
3703 challenges are checked to ensure they are valid base64 strings. By
3704 default, the authentication attempt is cancelled if an invalid string is
3705 received. Setting client_ignore_invalid_base64 true ignores these errors.
3706 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3707 they are received. Thus, the responses can be made to depend on the
3708 challenges. If an invalid string is ignored, an empty string is placed in
3711 PH/46 Messages that are created by the autoreply transport now contains a
3712 References: header, in accordance with RFCs 2822 and 3834.
3714 PH/47 Added authenticated_sender_force to the smtp transport.
3716 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3718 PH/49 Installed latest patch from the Sieve maintainer.
3720 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3721 was also set, the appendfile transport was unnecessarily scanning a
3722 directory of message files (e.g. for maildir delivery) to find the count
3723 of files (along with the size), even though it did not need this
3724 information. It now does the scan only if it needs to find either the
3725 size of the count of files.
3727 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3729 PH/52 Two bugs concerned with error handling when the smtp transport is
3732 (i) Exim was not creating retry information for temporary errors given
3733 for individual recipients after the DATA command when the smtp transport
3734 was used in LMTP mode. This meant that they could be retried too
3735 frequently, and not timed out correctly.
3737 (ii) Exim was setting the flag that allows error details to be returned
3738 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3739 recipients that were returned after the DATA command.
3741 PH/53 This is related to PH/52, but is more general: for any failing address,
3742 when detailed error information was permitted to be returned to the
3743 sender, but the error was temporary, then after the final timeout, only
3744 "retry timeout exceeded" was returned. Now it returns the full error as
3745 well as "retry timeout exceeded".
3747 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3748 do this, and (what is worse) MTAs that accept it.
3750 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3751 will now be deprecated.
3753 PH/56 New os.c-cygwin from the Cygwin maintainer.
3755 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3756 formats (previously only available in exim formats via -bpr, -bpru,
3757 and -bpra. Now also available in native and exiqgrep formats)
3759 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3760 with very large, slow to parse queues
3762 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3764 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3766 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3767 responses to authentication challenges, though it was showing the
3768 challenges; (ii) I've removed the CR characters from the debug output for
3771 PH/58 Allow for the insertion of a newline as well as a space when a string
3772 is turned into more than one encoded-word during RFC 2047 encoding. The
3773 Sieve code now uses this.
3775 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3776 data_4xx, lost_connection, tls_required.
3778 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3779 message was being sent as an SMTP response.
3781 PH/61 Add -l and -k options to exicyclog.
3783 PH/62 When verifying, if an address was redirected to one new address, so that
3784 verification continued, and the new address failed or deferred after
3785 having set something in $address_data, the value of $address_data was not
3786 passed back to the ACL. This was different to the case when no
3787 redirection occurred. The value is now passed back in both cases.
3789 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3790 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3791 use login_cap.h, so on its own it isn't the distinguishing feature. The
3792 new name refers directly to the setclassresources() function.
3794 PH/65 Added configuration files for NetBSD3.
3796 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3798 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3799 is preferred over IPv4.
3801 PH/68 The bounce_return_message and bounce_return_body options were not being
3802 honoured for bounces generated during the reception of non-SMTP messages.
3803 In particular, this applied to messages rejected by the ACL. This bug has
3804 been fixed. However, if bounce_return_message is true and bounce_return_
3805 body is false, the headers that are returned for a non-SMTP message
3806 include only those that have been read before the error was detected.
3807 (In the case of an ACL rejection, they have all been read.)
3809 PH/69 The HTML version of the specification is now built in a directory called
3810 spec_html instead of spec.html, because the latter looks like a path with
3811 a MIME-type, and this confuses some software.
3813 PH/70 Catch two compiler warnings in sieve.c.
3815 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3816 function verify_get_ident() calls ip_connect() to connect a socket, but
3817 if the "connect()" function timed out, ip_connect() used to close the
3818 socket. However, verify_get_ident() also closes the socket later, and in
3819 between Exim writes to the log, which may get opened at this point. When
3820 the socket was closed in ip_connect(), the log could get the same file
3821 descriptor number as the socket. This naturally causes chaos. The fix is
3822 not to close the socket in ip_connect(); the socket should be closed by
3823 the function that creates it. There was only one place in the code where
3824 this was missing, in the iplookup router, which I don't think anybody now
3825 uses, but I've fixed it anyway.
3827 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3828 well as to direct DNS lookups. Otherwise the handling of names in host
3829 lists is inconsistent and therefore confusing.
3835 PH/01 Two changes to the default runtime configuration:
3837 (1) Move the checks for relay_from_hosts and authenticated clients from
3838 after to before the (commented out) DNS black list checks.
3840 (2) Add control=submission to the relay_from_hosts and authenticated
3841 clients checks, on the grounds that messages accepted by these
3842 statements are most likely to be submissions.
3844 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3846 (1) Generate an error if the third argument for the ${prvs expansion is
3849 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3852 (3) Reset the variables that are obtained from the first argument of
3853 ${prvscheck and used in the second argument before leaving the code,
3854 because their memory is reclaimed, so using them afterwards may do
3857 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3858 one (it's much easier than Tom thought :-).
3860 (5) Because of (4), we can now allow for the use of $prvscheck_result
3861 inside the third argument.
3863 PH/03 For some reason, the default setting of PATH when running a command from
3864 a pipe transport was just "/usr/bin". I have changed it to
3867 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3868 anything to be listed in the output from -bV.
3870 PH/05 When a filter generated an autoreply, the entire To: header line was
3871 quoted in the delivery log line, like this:
3873 => >A.N.Other <ano@some.domain> <original@ddress> ...
3875 This has been changed so that it extracts the operative address. There
3876 may be more than one such address. If so, they are comma-separated, like
3879 => >ano@some.domain,ona@other.domain <original@ddress> ...
3881 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3882 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3883 not being looked up in the rDNS to get a host name, Exim was showing the
3884 IP address twice in Received: lines, even though the IP addresses were
3885 identical. For example:
3887 Received: from [1.2.3.4] (helo=[1.2.3.4])
3889 However, if the real host name was known, it was omitting the HELO data
3890 if it matched the actual IP address. This has been tidied up so that it
3891 doesn't show the same IP address twice.
3893 PH/07 When both +timestamp and +memory debugging was on, the value given by
3894 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3895 called by the debug printing, thereby overwriting the timestamp buffer.
3896 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3898 PH/08 When the original message was included in an autoreply transport, it
3899 always said "this is a copy of the message, including all the headers",
3900 even if body_only or headers_only was set. It now gives an appropriate
3903 PH/09 Applied a patch from the Sieve maintainer which:
3905 o fixes some comments
3906 o adds the (disabled) notify extension core
3907 o adds some debug output for the result of if/elsif tests
3908 o points to the current vacation draft in the documentation
3909 and documents the missing references header update
3913 o fixes a bug in processing the envelope test (when testing
3914 multiple envelope elements, the last element determined the
3917 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3918 Electronic Mail") by including:
3920 Auto-submitted: auto-generated
3922 in the messages that it generates (bounce messages and others, such as
3923 warnings). In the case of bounce messages for non-SMTP messages, there was
3924 also a typo: it was using "Auto_submitted" (underscore instead of
3925 hyphen). Since every message generated by Exim is necessarily in response
3926 to another message, thes have all been changed to:
3928 Auto-Submitted: auto-replied
3930 in accordance with these statements in the RFC:
3932 The auto-replied keyword:
3934 - SHOULD be used on messages sent in direct response to another
3935 message by an automatic process,
3937 - MUST NOT be used on manually-generated messages,
3939 - MAY be used on Delivery Status Notifications (DSNs) and Message
3940 Disposition Notifications (MDNs),
3942 - MUST NOT be used on messages generated by automatic or periodic
3943 processes, except for messages which are automatic responses to
3946 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3947 to the default Received: header definition.
3949 PH/12 Added log selector acl_warn_skipped (default on).
3951 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3952 variables because (a) they are in the wrong storage pool and (b) even if
3953 they were copied, it wouldn't work properly because of the caching.
3955 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3956 checking when decoding. Apparently there are clients that generate
3957 overlong encoded strings. Why am I not surprised?
3959 PH/15 If the first argument of "${if match_address" was not empty, but did not
3960 contain an "@" character, Exim crashed. Now it writes a panic log message
3961 and treats the condition as false.
3963 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3965 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3966 extension "envelope-auth". The code is finished and in agreement with
3967 other implementations, but there is no documentation so far and in fact,
3968 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3969 not changing the active code.
3971 Print executed "if" and "elsif" statements when debugging is used. This
3972 helps a great deal to understand what a filter does.
3974 Document more things not specified clearly in RFC3028. I had all this
3975 sorted out, when out of a sudden new issues came to my mind. Oops."
3977 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3980 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3981 canonical form (as documented). However, after a host name lookup from
3982 the IP address, check_host() was doing a simple string comparison with
3983 addresses acquired from the DNS when checking that the found name did
3984 have the original IP as one of its addresses. Since any found IPv6
3985 addresses are likely to be in abbreviated form, the comparison could
3986 fail. Luckily, there already exists a function for doing the comparison
3987 by converting both addresses to binary, so now that is used instead of
3988 the text comparison.
3990 PH/20 There was another similar case to PH/19, when a complete host name was
3991 given in a host list; looking up its IP address could give an abbreviated
3992 form, whereas the current host's name might or might not be abbreviated.
3993 The same fix has been applied.
3999 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4000 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4003 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4004 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4006 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4008 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4009 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4010 Also, for "Auto-Submitted", treat anything other than "no" as
4011 non-personal, in accordance with RFC 3834. (Previously it treated
4012 anything starting "auto-" as non-personal.)
4014 TF/01 The control=submission/name=... option had a problem with syntax
4015 errors if the name included a slash character. The /name= option
4016 now slurps the rest of the string, so it can include any characters
4017 but it must come last in the list of options (after /sender_retain
4020 PH/05 Some modifications to the interface to the fake nameserver for the new
4028 TK/01 Added the "success_on_redirect" address verification option. See
4029 NewStuff for rationale and an example.
4031 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4033 PH/02 Patch to exigrep to allow it to work on syslog lines.
4035 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4036 fread() to skip over the body file's header line, because in Cygwin the
4037 header line is locked and is inaccessible.
4039 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4040 co-exist for some time) to make it clear that it is the Exim ID that is
4041 referenced, not the Message-ID: header line.
4043 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4044 string_format() function, because snprintf() does not exist on all
4047 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4048 ${stat: expansion item.
4050 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4051 protocol synchronization error", to keep the pedants happy.
4053 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4054 well as for IRIX systems, when gcc is being used. See the host.c source
4057 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4059 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4062 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4063 transport and to the smtp transport in LMTP mode.
4065 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4067 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4068 run a filter in a subprocess. This could lead to confusion in subsequent
4069 lookups in the parent process. There should also be a search_tidyup() at
4070 the end of the subprocess.
4072 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4073 only if the host matched helo_try_verify_hosts, which caused the
4074 verification to occur when the EHLO/HELO command was issued. The ACL just
4075 tested the remembered result. Now, if a previous verification attempt has
4076 not happened, "verify = helo" does it there and then.
4078 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4080 TK/03 Fix log output including CR from clamd.
4082 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4083 exist provoked a memory error which could cause a segfault.
4085 PH/15 Installed PCRE 6.2
4087 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4089 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4090 of the problem. Specifically, suggested +O2 rather than +O1 for the
4093 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4095 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4096 record, the port was not correctly logged when the outgoing_port log
4097 selector was set (it logged the transort's default port).
4099 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4100 fallback_hosts, and "hosts" in the smtp transport.
4102 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4103 host errors such as "Connection refused".
4105 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4106 authentication with radiusclient 0.4.9:
4108 - Error returned from rc_read_config was caught wrongly
4109 - Username/password not passed on to radius server due to wrong length.
4111 The presumption is that some radiusclient API changes for 4.51/PH/17
4112 were not taken care of correctly. The code is still untested by me (my
4113 Linux distribution still has 0.3.2 of radiusclient), but it was
4114 contributed by a Radius user.
4116 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4117 expanding the "port" option of the smtp transport.
4119 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4120 while reading a MIME header. Thanks to Tom Hughes for a patch.
4122 PH/24 Include config.h inside local_scan.h so that configuration settings are
4125 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4126 that in an ACL for RCPT (for example), you can examine exactly what was
4129 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4130 commands, but it was not correctly comparing the address with the actual
4131 client host address. Thus, it would show the EHLO address in Received:
4132 header lines when this was not necessary.
4134 PH/27 Added the % operator to ${eval:}.
4136 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4137 it should be ignoring failures (because with -C, for example, it has lost
4138 privilege). It wasn't ignoring creation failures other than "already
4141 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4144 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4145 because an input line was too long, either on its own, or by virtue of
4146 too many continuations, the temporary file was not being removed, and the
4147 return code was incorrect.
4149 PH/31 Missing "BOOL" in function definition in filtertest.c.
4151 PH/32 Applied Sieve patches from the maintainer.
4153 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4155 PH/33 Added "verify = not_blind".
4157 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4158 Local/Makefile (with some defaults set). These are used in built scripts
4159 such as exicyclog, but they have never been used in the exim_install
4160 script (though there are many overriding facilities there). I have
4161 arranged that the exim_install script now takes note of these two
4164 PH/35 Installed configuration files for Dragonfly.
4166 PH/36 When a locally submitted message by a trusted user did not contain a
4167 From: header, and the sender address was obtained from -f or from an SMTP
4168 MAIL command, and the trusted user did not use -F to supply a sender
4169 name, $originator_name was incorrectly used when constructing a From:
4170 header. Furthermore, $originator_name was used for submission mode
4171 messages from external hosts without From: headers in a similar way,
4172 which is clearly wrong.
4174 PH/37 Added control=suppress_local_fixups.
4176 PH/38 When log_selector = +received_sender was set, and the addition of the
4177 sender made the log line's construction buffer exactly full, or one byte
4178 less than full, an overflow happened when the terminating "\n" was
4181 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4182 when the result of a list match is failure because a DNS lookup failed.
4184 PH/40 RM_COMMAND is now used in the building process.
4186 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4187 the "build-* directories that it finds.
4189 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4190 address in a domain literal was a prefix of an interface address.
4192 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4193 when verifying a sender address, unless rewrite_headers is false.
4195 PH/44 Wrote a long comment about why errors_to addresses are verified as
4196 recipients, not senders.
4198 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4199 the ratelimit ACL was added.
4201 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4203 PH/46 Added extra information about PostgreSQL errors to the error string.
4205 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4206 avoiding the need to install special zones in a real server. This is
4207 backwards compatible; if it can't find the fake resolver, it drops back.
4208 Thus, both old and new test suites can be run.
4210 TF/02 Added util/ratelimit.pl
4212 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4213 clock is set back in time.
4215 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4216 Candler <B.Candler@pobox.com>.
4218 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4219 OK for addresses that are the result of redirections.
4221 PH/48 A number of further additions for the benefit of the new test suite,
4222 including a fake gethostbyname() that interfaces to the fake DNS resolver
4225 TF/06 The fix for widen_domains has also been applied to qualify_single and
4226 search_parents which are the other dnslookup options that can cause
4229 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4232 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4234 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4235 service" if it encountered an MX record with an empty target hostname.
4236 The message is now "an MX or SRV record indicated no SMTP service".
4238 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4239 if the DNS of the sending site is misconfigured. This is quite a
4240 common situation. This change restores the behaviour of treating a
4241 helo verification defer as a failure.
4243 PH/51 If self=fail was set on a router, the bounce message did not include the
4244 actual error message.
4250 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4252 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4253 command itself ended in error, the underlying message about the transport
4254 filter timeout was being overwritten with the pipe command error. Now the
4255 underlying error message should be appended to the second error message.
4257 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4259 PH/02 Reduce the amount of output that "make" produces by default. Full output
4260 can still be requested.
4262 PH/03 The warning log line about a condition test deferring for a "warn" verb
4263 was being output only once per connection, rather than after each
4264 occurrence (because it was using the same function as for successful
4265 "warn" verbs). This seems wrong, so I have changed it.
4267 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4268 it should not have, which might have caused a crash in the right
4269 circumstances, but probably never did.
4271 PH/04 Installed a modified version of Tony Finch's patch to make submission
4272 mode fix the return path as well as the Sender: header line, and to
4273 add a /name= option so that you can make the user's friendly name appear
4276 TF/03 Added the control = fakedefer ACL modifier.
4278 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4279 Mark Lowes for thorough testing.
4281 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4283 TK/03 Merged latest SRS patch from Miles Wilton.
4285 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4286 with the definition in sysexits.h (which is #included earlier).
4287 Fortunately, Exim does not actually use EX_OK. The code used to try to
4288 preserve the sysexits.h value, by assuming that macro definitions were
4289 scanned for macro replacements. I have been disabused of this notion,
4290 so now the code just undefines EX_OK before #including unistd.h.
4292 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4293 in the smtp transport. When a block could not be written in a single
4294 write() function, the timeout was being re-applied to each part-write.
4295 This seems wrong - if the receiver was accepting one byte at a time it
4296 would take for ever. The timeout is now adjusted when this happens. It
4297 doesn't have to be particularly precise.
4299 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4300 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4302 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4303 without a domain if the check to <postmaster@domain> fails.
4305 SC/01 Eximstats: added -xls and the ability to specify output files
4306 (patch written by Frank Heydlauf).
4308 SC/02 Eximstats: use FileHandles for outputting results.
4310 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4312 SC/04 Eximstats: fixed display of large numbers with -nvr option
4314 SC/05 Eximstats: fixed merging of reports with empty tables.
4316 SC/06 Eximstats: added the -include_original_destination flag
4318 SC/07 Eximstats: removed tabs and trailing whitespace.
4320 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4322 TK/06 MBOX spool code: Add real "From " MBOX separator line
4323 so the .eml file is really in mbox format (even though
4324 most programs do not really care). Patch from Alex Miller.
4326 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4327 The latter is generated from $received_to and is only set if the
4328 message has one envelope recipient. SA can use these headers,
4329 obviously out-of-the-box. Patch from Alex Miller.
4331 PH/08 The ${def test on a variable was returning false if the variable's
4332 value was "0", contrary to what the specification has always said!
4333 The result should be true unless the variable is empty.
4335 PH/09 The syntax error of a character other than { following "${if
4336 def:variable_name" (after optional whitespace) was not being diagnosed.
4337 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4338 accidental colon was present, for example, could give incorrect results.
4340 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4341 result is used (not including appendfile, where other changes are about
4344 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4345 This involved changing a lot of size variables from int to off_t. It
4346 should work with maildirs and everything.
4348 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4349 spamd dying while we are connected to it.
4351 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4354 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4355 function for BDB 4.3.
4357 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4359 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4360 the sources for PCRE have been re-arranged and more files are now
4363 PH/15 The code I had for printing potentially long long variables in PH/11
4364 above was not the best (it lost precision). The length of off_t variables
4365 is now inspected at build time, and an appropriate printing format (%ld
4366 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4367 to be "long long int" or "long int". This is needed for the internal
4368 formatting function string_vformat().
4370 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4371 the configuration file to be ":syslog", then the script "guesses" where
4372 the logs files are, rather than using the compiled in default. In our
4373 case the guess is not the same as the compiled default, so the script
4374 suddenly stopped working when I started to use syslog. The patch checks
4375 to see if log_file_path is "". If so, it attempts to read it from exim
4376 with no configuration file to get the compiled in version, before it
4377 falls back to the previous guessing code."
4379 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4380 implementing BATV in an Exim configuration. See NewStuff for the gory
4383 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4384 Makefile that are specific to HP-UX.
4386 PH/18 If the "use_postmaster" option was set for a recipient callout together
4387 with the "random" option, the postmaster address was used as the MAIL
4388 FROM address for the random test, but not for the subsequent recipient
4389 test. It is now used for both.
4391 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4392 patch removes a few documentation additions to RFC 3028, because the
4393 latest draft now contains them. It adds the new en;ascii-case comparator
4394 and a new error check for 8bit text in MIME parts. Comparator and
4395 require names are now matched exactly. I enabled the subaddress
4396 extension, but it is not well tested yet (read: it works for me)."
4398 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4399 rework some of the code of TK/09 above to avoid the hardwired use of
4400 "%lld" and "long long". Replaced the call to snprintf() with a call to
4403 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4404 records point to non-existent hosts", "retry timeout exceeded", and
4405 "retry time not reached for any host after a long failure period".
4407 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4408 experimental DomainKeys support:
4410 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4411 (2) On an error such as an illegally used "control", the wrong name for
4412 the control was given.
4414 These problems did NOT occur unless DomainKeys support was compiled.
4416 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4418 PH/24 Added ${if match_ip condition.
4420 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4421 databases so that it will be absolutely obvious if a crash occurs in the
4422 DB library. This is a regular occurrence (often caused by mis-matched
4425 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4426 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4427 picked up on a user's system that detects such things. There doesn't seem
4428 to be a gcc warning option for this - only an attribute that has to be
4429 put on the function's prototype. It seems that in Fedora Core 4 they have
4430 set this on a number of new functions. No doubt there will be more in due
4433 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4434 specify a transport. However, if an address that was verified by such a
4435 router was the subject of a callout, Exim crashed because it tried to
4436 read the rcpt_include_affixes from the non-existent transport. Now it
4437 just assumes that the setting of that option is false. This bug was
4438 introduced by 4.51/PH/31.
4440 PH/28 Changed -d+all to exclude +memory, because that information is very
4441 rarely of interest, but it makes the output a lot bigger. People tend to
4442 do -d+all out of habit.
4444 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4445 code in os-type was giving problems when libc.so lives in lib64, like on
4448 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4449 aren't the modern standard, and it seems that some systems' include files
4450 don't always have them. Exim was already checking for some of the newer
4451 ones like T_AAAA, and defining it itself. I've added checks for all the
4452 record types that Exim uses.
4454 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4455 not automatically generating a new one, as it is supposed to. This
4456 prevented TLS from working. If the file did exist, but contained invalid
4457 data, a new version was generated, as expected. It was only the case of a
4458 non-existent file that was broken.
4460 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4461 with a change in libdomainkeys > 0.64.
4463 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4464 from DNS. If the selector record carries the flag, it now has
4465 precedence over the domain-wide flag.
4467 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4469 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4470 the use of a port name that isn't defined in /etc/services) occurred, the
4471 message was deferred as in a normal delivery, and thus remained on the
4472 spool, instead of being failed because of the mua_wrapper setting. This
4473 is now fixed, and I tidied up some of the mua_wrapper messages at the
4476 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4477 the messages in a hash of arrays rather than using individual hashes.
4478 This is a bit cleaner and results in dramatic memory savings, albeit
4479 at a slight CPU cost.
4481 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4482 as requested by Marc Sherman.
4484 SC/10 Eximstats: added histograms for user specified patterns as requested
4487 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4489 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4490 fopen() in the content-scanning modules that did not already have it.
4496 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4497 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4499 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4501 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4503 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4504 to test Sieve filters that use "vacation".
4506 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4507 that changes the way the GnuTLS parameters are stored in the cache file.
4508 The new format can be generated externally. For backward compatibility,
4509 if the data in the cache doesn't make sense, Exim assumes it has read an
4510 old-format file, and it generates new data and writes a new file. This
4511 means that you can't go back to an older release without removing the
4514 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4515 work if there are any delivery delays because "one_time" forces the
4516 parent to be marked "delivered", so its unseen clone is never tried
4517 again. For this reason, Exim now forbids the simultaneous setting of
4520 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4521 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4522 entry if you want to know the details. The fix, however, neglected to
4523 consider the case when local delivery batching is involved. The test for
4524 "previously delivered" was not happening when checking to see if an
4525 address could be batched with a previous (undelivered) one; under
4526 certain circumstances this could lead to multiple deliveries to the same
4529 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4530 in its include files, and this causes problems building Exim.
4532 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4533 header_syntax) but Exim was just ignoring anything given after a slash.
4534 In particular, this caused confusion with an attempt to use "verify =
4535 reverse_host_lookup/defer_ok". An error is now given when options are
4536 supplied for verify items that do not have them. (Maybe reverse_host_
4537 lookup should have a defer_ok option, but that's a different point.)
4539 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4540 defined by RFC 821) to 2048, because there were problems with some AUTH
4541 commands, and RFC 1869 says the size should be increased for extended
4542 SMTP commands that take arguments.
4544 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4547 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4548 "unknown" error; now it says that the functionality isn't in the binary.
4550 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4551 an address' error message when a string expansion fails (syntax or
4552 whatever). Otherwise the password may appear in the log. Following change
4553 PH/42 below, there is no longer a chance of it appearing in a bounce
4556 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4558 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4559 of its addresses was ever tried. (Bugzilla bug #2.)
4561 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4562 the result incorrectly in the debug output. (It correctly added a newline
4563 to what was transported.)
4565 TF/01 Added $received_time.
4567 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4568 commented out examples of how to interface to a virus scanner and to
4569 SpamAssassin. Also added commented examples of av_scanner and
4570 spamd_address settings.
4572 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4573 and controls are allowed in which ACLs. There were a couple of minor
4574 errors. Some of the entries in the conditions table (which is a table of
4575 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4576 negation of where the condition IS allowed.
4578 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4580 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4581 header file does not have a version number, so I've had to invent a new
4582 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4583 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4584 radiusclient), but it was contributed by a Radius user.
4586 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4587 files or directories, for OpenSSL.
4589 PH/19 When an Exim process that is running as root has to create an Exim log
4590 file, it does so in a subprocess that runs as exim:exim so as to get the
4591 ownership right at creation (otherwise, other Exim processes might see
4592 the file with the wrong ownership). There was no test for failure of this
4593 fork() call, which would lead to the process getting stuck as it waited
4594 for a non-existent subprocess. Forks do occasionally fail when resources
4595 run out. I reviewed all the other calls to fork(); they all seem to check
4598 PH/20 When checking for unexpected SMTP input at connect time (before writing
4599 the banner), Exim was not dealing correctly with a non-positive return
4600 from the read() function. If the client had disconnected by this time,
4601 the result was a log entry for a synchronization error with an empty
4602 string after "input=" when read() returned zero. If read() returned -1
4603 (an event I could not check), uninitialized data bytes were printed.
4604 There were reports of junk text (parts of files, etc) appearing after
4607 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4609 PH/22 Added support for macro redefinition, and (re)definition in between
4610 driver and ACL definitions.
4612 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4613 forgetting to use the resulting value; it was using the unexpanded value.
4615 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4616 hadn't been configured. The fix is from Juergen Kreileder, who
4617 understands it better than I do:
4619 "Here's what I see happening with three configured cyrus_sasl
4620 authenticators configured (plain, login, cram-md5):
4622 On startup auth_cyrus_sasl_init() gets called for each of these.
4623 This means three calls to sasl_listmech() without a specified mech_list.
4624 => SASL tests which mechs of all available mechs actually work
4625 => three warnings about OTP not working
4626 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4628 With the patch, sasl_listmech() also gets called three times. But now
4629 SASL's mech_list option is set to the server_mech specified in the the
4630 authenticator. Or in other words, the answer from sasl_listmech()
4631 gets limited to just the mech you're testing for (which is different
4633 => the return list contains just 'plain' or 'login', 'cram-md5' or
4634 nothing depending on the value of ob->server_mech.
4636 I've just tested the patch: Authentication still works fine,
4637 unavailable mechs specified in the exim configuration are still
4638 caught, and the auth.log warnings about OTP are gone."
4640 PH/25 When debugging is enabled, the contents of the command line are added
4641 to the debugging output, even when log_selector=+arguments is not
4644 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4645 answer is "GNU", and only if the return is "GNU/something" is the answer
4648 PH/27 $acl_verify_message is now set immediately after the failure of a
4649 verification in an ACL, and so is available in subsequent modifiers. In
4650 particular, the message can be preserved by coding like this:
4652 warn !verify = sender
4653 set acl_m0 = $acl_verify_message
4655 Previously, $acl_verify_message was set only while expanding "message"
4656 and "log_message" when a very denied access.
4658 PH/28 Modified OS/os.c-Linux with
4660 -#ifndef OS_LOAD_AVERAGE
4661 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4663 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4664 nomenclature these days.)
4666 PH/29 Installed patch from the Sieve maintainer that adds the options
4667 sieve_useraddress and sieve_subaddress to the redirect router.
4669 PH/30 In these circumstances:
4670 . Two addresses routed to the same list of hosts;
4671 . First host does not offer TLS;
4672 . First host accepts first address;
4673 . First host gives temporary error to second address;
4674 . Second host offers TLS and a TLS session is established;
4675 . Second host accepts second address.
4676 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4677 and peerdn, if requested) that were in fact used only for the second
4680 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4681 attention to any local part prefix or suffix that was matched by the
4682 router that accepted the address. It now behaves in the same way as it
4683 does for delivery: the affixes are removed from the local part unless
4684 rcpt_include_affixes is set on the transport.
4686 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4687 timeout during the DATA phase of an incoming message.
4689 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4690 applied a patch sanctioned by the Sieve maintainer.
4692 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4693 the uid or gid is negative. A case of a negative gid caused this to be
4694 noticed. The fix allows for either to be negative.
4696 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4697 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4700 PH/36 The change PH/12 above was broken. Fixed it.
4702 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4703 the grounds that routing the same address twice would always produce the
4704 same answer. This might have been true once, but it is certainly no
4705 longer true now. Routing a child address may depend on the previous
4706 routing that produced that child. Some complicated redirection strategies
4707 went wrong when messages had multiple recipients, and made Exim's
4708 behaviour dependent on the order in which the addresses were given.
4710 I have moved the duplicate checking until after the routing is complete.
4711 Exim scans the addresses that are assigned to local and remote
4712 transports, and removes any duplicates. This means that more work will be
4713 done, as duplicates will always all be routed, but duplicates are
4714 presumably rare, so I don't expect this is of any significance.
4716 For deliveries to pipes, files, and autoreplies, the duplicate checking
4717 still happens during the routing process, since they are not going to be
4720 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4721 It corrects a timeout issue with spamd. This is Ian's comment: "The
4722 background is that sometimes spamd either never reads data from a
4723 connection it has accepted, or it never writes response data. The exiscan
4724 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4725 blindly assumes that writes won't block so it may never time out."
4727 PH/39 Allow G after quota size as well as K and M.
4729 PH/40 The value set for $authenticated_id in an authenticator may not contain
4730 binary zeroes or newlines because the value is written to log lines and
4731 to spool files. There was no check on this. Now the value is run through
4732 the string_printing() function so that such characters are converted to
4733 printable escape sequences.
4735 PH/41 $message_linecount is a new variable that contains the total number of
4736 lines in the message. Compare $body_linecount, which is the count for the
4739 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4740 bounce and delay warning messages, except in certain special cases, which
4743 (a) An SMTP error message from a remote host;
4744 (b) A message specified in a :fail: redirection;
4745 (c) A message specified in a "fail" command in a system filter;
4746 (d) A message specified in a FAIL return from the queryprogram router;
4747 (e) A message specified by the cannot_route_message router option.
4749 In these cases only, Exim does include the error details in bounce and
4750 warning messages. There are also a few cases where bland messages such
4751 as "unrouteable address" or "local delivery error" are given.
4753 PH/43 $value is now also set for the "else" part of a ${run expansion.
4755 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4756 being worked on, but at least Exim now implements the latest version to
4759 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4760 process to complete was treated as a delivery failure, a timeout while
4761 writing the message to the pipe was logged, but erroneously treated as a
4762 successful delivery. Such timeouts include transport filter timeouts. For
4763 consistency with the overall process timeout, these timeouts are now
4764 treated as errors, giving rise to delivery failures by default. However,
4765 there is now a new Boolean option for the pipe transport called
4766 timeout_defer, which, if set TRUE, converts the failures into defers for
4767 both kinds of timeout. A transport filter timeout is now identified in
4770 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4771 systems where "make" and "gmake" are different, calling "gmake" at top
4772 level broke things. I've arranged for the value of $(MAKE) to be passed
4773 from the Makefile to this script so that it can call the same version of
4777 A note about Exim versions 4.44 and 4.50
4778 ----------------------------------------
4780 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4781 changes of various kinds. As a consequence, a big documentation update was
4782 needed. This delayed the release for rather longer than seemed good, especially
4783 in the light of a couple of (minor) security issues. Therefore, the changes
4784 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4785 release. So 4.44 and 4.50 are in effect two different branches that both start
4788 I have left the 4.50 change log unchanged; it contains all the changes since
4789 4.43. The change log for 4.44 is below; many of its items are identical to
4790 those for 4.50. This seems to be the most sensible way to preserve the
4791 historical information.
4797 1. Minor wording change to the doc/README.SIEVE file.
4799 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4800 computation of the current number of files was incorrect.
4802 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4803 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4806 4. Give more explanation in the error message when the command for a transport
4807 filter fails to execute.
4809 5. There are several places where Exim runs a non-Exim command in a
4810 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4811 was being done only for the command run by the queryprogram router. It is
4812 now done for all such subprocesses. The other cases are: ${run, transport
4813 filters, and the commands run by the lmtp and pipe transports.
4815 6. Added CONFIGURE_GROUP build-time option.
4817 7. Some older OS have a limit of 256 on the maximum number of file
4818 descriptors. Exim was using setrlimit() to set 1000 as a large value
4819 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4820 systems. I've change it so that if it can't get 1000, it tries for 256.
4822 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4823 was an oversight, and furthermore, ever since the addition of extra
4824 controls (e.g. 4.43/32), the checks on when to allow different forms of
4825 "control" were broken. There should now be diagnostics for all cases when a
4826 control that does not make sense is encountered.
4828 9. Added the /retain_sender option to "control=submission".
4830 10. $recipients is now available in the predata ACL (oversight).
4832 11. Tidy the search cache before the fork to do a delivery from a message
4833 received from the command line. Otherwise the child will trigger a lookup
4834 failure and thereby defer the delivery if it tries to use (for example) a
4835 cached ldap connection that the parent has called unbind on.
4837 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4838 of $address_data from the recipient verification was clobbered by the
4839 sender verification.
4841 13. The value of address_data from a sender verification is now available in
4842 $sender_address_data in subsequent conditions in the ACL statement.
4844 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4846 15. Added a new option "connect=<time>" to callout options, to set a different
4849 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4850 was its contents. (It was OK if the option was not defined at all.)
4852 17. A "Completed" log line is now written for messages that are removed from
4853 the spool by the -Mrm option.
4855 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4856 information about exactly what failed.
4858 19. Added -dd to debug only the daemon process.
4860 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4861 handles timeouts, both on the server side and network timeouts. Renamed the
4862 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4864 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4865 It is now set to "smtps".
4867 22. $host_address is now set to the target address during the checking of
4868 ignore_target_hosts.
4870 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4871 being passed; this would have caused $sender_host_name to have been used if
4872 matching the list had actually called for a host name (not very likely,
4873 since this list is usually IP addresses). A host name is now passed as
4876 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4877 code that specifies a non-restarting handler (typically sigaction() in
4878 modern systems) in an attempt to fix a rare and obscure crash bug.
4880 25. Narrowed the window for a race in the daemon that could cause it to ignore
4881 SIGCHLD signals. This is not a major problem, because they are used only to
4882 wake it up if nothing else does.
4884 26. A malformed maildirsize file could cause Exim to calculate negative values
4885 for the mailbox size or file count. Odd effects could occur as a result.
4886 The maildirsize information is now recalculated if the size or filecount
4889 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4890 support for a long time. Removed HAVE_SYS_VFS_H.
4892 28. Installed the latest version of exipick from John Jetmore.
4894 29. In an address list, if the pattern was not a regular expression, an empty
4895 subject address (from a bounce message) matched only if the pattern was an
4896 empty string. Non-empty patterns were not even tested. This was the wrong
4897 because it is perfectly reasonable to use an empty address as part of a
4898 database query. An empty address is now tested by patterns that are
4899 lookups. However, all the other forms of pattern expect the subject to
4900 contain a local part and a domain, and therefore, for them, an empty
4901 address still always fails if the pattern is not itself empty.
4903 30. Exim went into a mad DNS loop when attempting to do a callout where the
4904 host was specified on an smtp transport, and looking it up yielded more
4905 than one IP address.
4907 31. Re-factored the code for checking spool and log partition space into a
4908 function that finds that data and another that does the check. The former
4909 is then used to implement four new variables: $spool_space, $log_space,
4910 $spool_inodes, and $log_inodes.
4912 32. The RFC2047 encoding function was originally intended for short strings
4913 such as real names; it was not keeping to the 75-character limit for
4914 encoded words that the RFC imposes. It now respects the limit, and
4915 generates multiple encoded words if necessary. To be on the safe side, I
4916 have increased the buffer size for the ${rfc2047: expansion operator from
4919 33. It is now permitted to omit both strings after an "if" condition; if the
4920 condition is true, the result is "true". As before, when the second string
4921 is omitted, a false condition yields an empty string. This makes it less
4922 cumbersome to write custom ACL and router conditions.
4924 34. Failure to deliver a bounce message always caused it to be frozen, even if
4925 there was an errors_to setting on the router. The errors_to setting is now
4928 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4929 canonical form (fully expanded) before being placed in
4930 $sender_host_address.
4932 36. The table in the code that translates DNS record types into text (T_A to
4933 "A" for instance) was missing entries for NS and CNAME. It is just possible
4934 that this could have caused confusion if both these types were looked up
4935 for the same domain, because the text type is used as part of Exim's
4936 per-process caching. But the chance of anyone hitting this buglet seems
4939 37. The dnsdb lookup has been extended in a number of ways.
4941 (1) There is a new type, "zns", which walks up the domain tree until it
4942 finds some nameserver records. It should be used with care.
4944 (2) There is a new type, "mxh", which is like "mx" except that it returns
4945 just the host names, not the priorities.
4947 (3) It is now possible to give a list of domains (or IP addresses) to be
4948 looked up. The behaviour when one of the lookups defers can be
4949 controlled by a keyword.
4951 (4) It is now possible to specify the separator character for use when
4952 multiple records are returned.
4954 38. The dnslists ACL condition has been extended: it is now possible to supply
4955 a list of IP addresses and/or domains to be looked up in a particular DNS
4958 39. Added log_selector=+queue_time_overall.
4960 40. When running the queue in the test harness, wait just a tad after forking a
4961 delivery process, to get repeatability of debugging output.
4963 41. Include certificate and key file names in error message when GnuTLS fails
4964 to set them up, because the GnuTLS error message doesn't include the name
4965 of the failing file when there is a problem reading it.
4967 42. Allow both -bf and -bF in the same test run.
4969 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4971 44. The "Exiscan patch" is now merged into the mainline Exim source.
4973 45. Sometimes the final signoff response after QUIT could fail to get
4974 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4975 < 0 before doing a fflush(). This bug looks as though it goes back to the
4976 introduction of TLS in release 3.20, but "sometimes" must have been rare
4977 because the tests only now provoked it.
4979 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4980 (this can affect the format of dates).
4982 47. exim_tidydb, when checking for the continued existence of a message for
4983 which it has found a message-specific retry record, was not finding
4984 messages that were in split spool directories. Consequently, it was
4985 deleting retry records that should have stayed in existence.
4987 48. Steve fixed some bugs in eximstats.
4989 49. The SPA authentication driver was not abandoning authentication and moving
4990 on to the next authenticator when an expansion was forced to fail,
4991 contradicting the general specification for all authenticators. Instead it
4992 was generating a temporary error. It now behaves as specified.
4994 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4995 (the order specifies the preference for clients). The order is now AES256,
4996 AES128, 3DES, ARCFOUR128.
4998 51. Small patch to Sieve code - explicitly set From: when generating an
5001 52. Exim crashed if a remote delivery caused a very long error message to be
5002 recorded - for instance if somebody sent an entire SpamAssassin report back
5003 as a large number of 550 error lines. This bug was coincidentally fixed by
5004 increasing the size of one of Exim's internal buffers (big_buffer) that
5005 happened as part of the Exiscan merge. However, to be on the safe side, I
5006 have made the code more robust (and fixed the comments that describe what
5009 53. Now that there can be additional text after "Completed" in log lines (if
5010 the queue_time_overall log selector is set), a one-byte patch to exigrep
5011 was needed to allow it to recognize "Completed" as not the last thing in
5014 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5015 patch that reportedly fixes this has been added. I am not expert enough to
5016 create a test for it. This is what the patch creator wrote:
5018 "I found a little strange behaviour of ldap code when working with
5019 Windows 2003 AD Domain, where users was placed in more than one
5020 Organization Units. When I tried to give exim partial DN, the exit code
5021 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5022 But simultaneously result of request was absolutely normal ldap result,
5023 so I produce this patch..."
5025 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5026 REFERENCE, so I have modified the code to exclude the patch when that macro
5029 55. Some experimental protocols are using DNS PTR records for new purposes. The
5030 keys for these records are domain names, not reversed IP addresses. The
5031 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5032 leaves it alone. Component reversal etc. now happens only for IP addresses.
5035 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5037 57. Double the size of the debug message buffer (to 2048) so that more of very
5038 long debug lines gets shown.
5040 58. The exicyclog utility now does better if the number of log files to keep
5041 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5043 59. Two changes related to the smtp_active_hostname option:
5045 (1) $smtp_active_hostname is now available as a variable.
5046 (2) The default for smtp_banner uses $smtp_active_hostname instead
5047 of $primary_hostname.
5049 60. The host_aton() function is supposed to be passed a string that is known
5050 to be a valid IP address. However, in the case of IPv6 addresses, it was
5051 not checking this. This is a hostage to fortune. Exim now panics and dies
5052 if the condition is not met. A case was found where this could be provoked
5053 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5054 components; fortuitously, this particular loophole had already been fixed
5055 by change 4.50/55 above.
5057 If there are any other similar loopholes, the new check in host_aton()
5058 itself should stop them being exploited. The report I received stated that
5059 data on the command line could provoke the exploit when Exim was running as
5060 exim, but did not say which command line option was involved. All I could
5061 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5062 running as the user.
5065 61. There was a buffer overflow vulnerability in the SPA authentication code
5066 (which came originally from the Samba project). I have added a test to the
5067 spa_base64_to_bits() function which I hope fixes it.
5070 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5071 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5073 63. The daemon start-up calls getloadavg() while still root for those OS that
5074 need the first call to be done as root, but it missed one case: when
5075 deliver_queue_load_max is set with deliver_drop_privilege. This is
5076 necessary for the benefit of the queue runner, because there is no re-exec
5077 when deliver_drop_privilege is set.
5079 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5080 This has been fixed.
5082 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5083 was in use, was not putting the data itself into the right store pool;
5084 consequently, it could be overwritten for a subsequent message in the same
5085 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5088 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5090 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5091 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5092 the function were treating the return as a boolean value, which happened to
5093 work because 0=false and not-0=true, but is not correct code.
5095 68. The host_aton() function was not handling scoped IPv6 addresses (those
5096 with, for example, "%eth0" on the end) correctly.
5098 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5099 negated items (that is, ~something) in unsigned ints. Some compilers
5100 apparently mutter when there is no cast.
5102 70. If an address verification called from an ACL failed, and did not produce a
5103 user-specific message (i.e. there was only a "system" message), nothing was
5104 put in $acl_verify_message. In this situation, it now puts the system
5107 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5108 session; change 4.31/43 added the unwanted input to the log line - except
5109 that it did not do this in the start of session case. It now does.
5111 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5112 This is wrong and can cause the other end to generate a synchronization
5113 error if it is another Exim or anything else that does the synchronization
5114 check. A QUIT command is no longer sent after a timeout.
5116 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5117 during host lookups.
5119 74. The defer_ok option of callout verification was not working if it was used
5120 when verifying addresses in header lines, that is, for this case:
5122 verify = header_sender/callout=defer_ok
5124 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5125 those file descriptors could be used for SMTP connections. If anything
5126 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5127 could be sent to the SMTP client, causing chaos. The daemon now opens
5128 stdin, stdout, and stderr to /dev/null when it puts itself into the
5131 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5132 log by default. The user can override this with suitable Perl magic.
5134 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5135 the log message when discard triggers, was not working for the DATA ACL or
5136 for the non-SMTP ACL.
5138 78. Error message wording change in sieve.c.
5140 79. If smtp_accept_max_per_host was set, the number of connections could be
5141 restricted to fewer than expected, because the daemon was trying to set up
5142 a new connection before checking whether the processes handling previous
5143 connections had finished. The check for completed processes is now done
5144 earlier. On busy systems, this bug wouldn't be noticed because something
5145 else would have woken the daemon, and it would have reaped the completed
5148 80. If a message was submitted locally by a user whose login name contained one
5149 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5150 It caused a spool format error. I have fixed the spool reading code. A
5151 related problem was that the "from" clause in the Received: line became
5152 illegal because of the space(s). It is now covered by ${quote_local_part.
5154 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5157 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5159 83. Patch from Sieve maintainer to fix unterminated string problem in
5160 "vacation" handling.
5162 84. Some minor changes to the Linux configuration files to help with other
5163 OS variants using glibc.
5165 85. One more patch for Sieve to update vacation handling to latest spec.
5168 ----------------------------------------------------
5169 See the note above about the 4.44 and 4.50 releases.
5170 ----------------------------------------------------
5176 1. Change 4.43/35 introduced a bug that caused file counts to be
5177 incorrectly computed when quota_filecount was set in an appendfile
5180 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5181 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5184 3. Give more explanation in the error message when the command for a transport
5185 filter fails to execute.
5187 4. There are several places where Exim runs a non-Exim command in a
5188 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5189 was being done only for the command run by the queryprogram router. It is
5190 now done for all such subprocesses. The other cases are: ${run, transport
5191 filters, and the commands run by the lmtp and pipe transports.
5193 5. Some older OS have a limit of 256 on the maximum number of file
5194 descriptors. Exim was using setrlimit() to set 1000 as a large value
5195 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5196 systems. I've change it so that if it can't get 1000, it tries for 256.
5198 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5199 was an oversight, and furthermore, ever since the addition of extra
5200 controls (e.g. 4.43/32), the checks on when to allow different forms of
5201 "control" were broken. There should now be diagnostics for all cases when a
5202 control that does not make sense is encountered.
5204 7. $recipients is now available in the predata ACL (oversight).
5206 8. Tidy the search cache before the fork to do a delivery from a message
5207 received from the command line. Otherwise the child will trigger a lookup
5208 failure and thereby defer the delivery if it tries to use (for example) a
5209 cached ldap connection that the parent has called unbind on.
5211 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5212 of $address_data from the recipient verification was clobbered by the
5213 sender verification.
5215 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5216 was its contents. (It was OK if the option was not defined at all.)
5218 11. A "Completed" log line is now written for messages that are removed from
5219 the spool by the -Mrm option.
5221 12. $host_address is now set to the target address during the checking of
5222 ignore_target_hosts.
5224 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5225 being passed; this would have caused $sender_host_name to have been used if
5226 matching the list had actually called for a host name (not very likely,
5227 since this list is usually IP addresses). A host name is now passed as
5230 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5231 code that specifies a non-restarting handler (typically sigaction() in
5232 modern systems) in an attempt to fix a rare and obscure crash bug.
5234 15. Narrowed the window for a race in the daemon that could cause it to ignore
5235 SIGCHLD signals. This is not a major problem, because they are used only to
5236 wake it up if nothing else does.
5238 16. A malformed maildirsize file could cause Exim to calculate negative values
5239 for the mailbox size or file count. Odd effects could occur as a result.
5240 The maildirsize information is now recalculated if the size or filecount
5243 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5244 support for a long time. Removed HAVE_SYS_VFS_H.
5246 18. Updated exipick to current release from John Jetmore.
5248 19. Allow an empty sender to be matched against a lookup in an address list.
5249 Previously the only cases considered were a regular expression, or an
5252 20. Exim went into a mad DNS lookup loop when doing a callout where the
5253 host was specified on the transport, if the DNS lookup yielded more than
5256 21. The RFC2047 encoding function was originally intended for short strings
5257 such as real names; it was not keeping to the 75-character limit for
5258 encoded words that the RFC imposes. It now respects the limit, and
5259 generates multiple encoded words if necessary. To be on the safe side, I
5260 have increased the buffer size for the ${rfc2047: expansion operator from
5263 22. Failure to deliver a bounce message always caused it to be frozen, even if
5264 there was an errors_to setting on the router. The errors_to setting is now
5267 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5268 canonical form (fully expanded) before being placed in
5269 $sender_host_address.
5271 24. Updated eximstats to version 1.33
5273 25. Include certificate and key file names in error message when GnuTLS fails
5274 to set them up, because the GnuTLS error message doesn't include the name
5275 of the failing file when there is a problem reading it.
5277 26. Expand error message when OpenSSL has problems setting up cert/key files.
5280 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5281 (this can affect the format of dates).
5283 28. exim_tidydb, when checking for the continued existence of a message for
5284 which it has found a message-specific retry record, was not finding
5285 messages that were in split spool directories. Consequently, it was
5286 deleting retry records that should have stayed in existence.
5288 29. eximstats updated to version 1.35
5289 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5290 1.35 - bugfix such that pie charts by volume are generated correctly
5292 30. The SPA authentication driver was not abandoning authentication and moving
5293 on to the next authenticator when an expansion was forced to fail,
5294 contradicting the general specification for all authenticators. Instead it
5295 was generating a temporary error. It now behaves as specified.
5297 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5298 (the order specifies the preference for clients). The order is now AES256,
5299 AES128, 3DES, ARCFOUR128.
5301 31. Small patch to Sieve code - explicitly set From: when generating an
5304 32. Exim crashed if a remote delivery caused a very long error message to be
5305 recorded - for instance if somebody sent an entire SpamAssassin report back
5306 as a large number of 550 error lines. This bug was coincidentally fixed by
5307 increasing the size of one of Exim's internal buffers (big_buffer) that
5308 happened as part of the Exiscan merge. However, to be on the safe side, I
5309 have made the code more robust (and fixed the comments that describe what
5312 33. Some experimental protocols are using DNS PTR records for new purposes. The
5313 keys for these records are domain names, not reversed IP addresses. The
5314 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5315 leaves it alone. Component reversal etc. now happens only for IP addresses.
5318 34. The host_aton() function is supposed to be passed a string that is known
5319 to be a valid IP address. However, in the case of IPv6 addresses, it was
5320 not checking this. This is a hostage to fortune. Exim now panics and dies
5321 if the condition is not met. A case was found where this could be provoked
5322 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5323 components; fortuitously, this particular loophole had already been fixed
5324 by change 4.50/55 or 4.44/33 above.
5326 If there are any other similar loopholes, the new check in host_aton()
5327 itself should stop them being exploited. The report I received stated that
5328 data on the command line could provoke the exploit when Exim was running as
5329 exim, but did not say which command line option was involved. All I could
5330 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5331 running as the user.
5334 35. There was a buffer overflow vulnerability in the SPA authentication code
5335 (which came originally from the Samba project). I have added a test to the
5336 spa_base64_to_bits() function which I hope fixes it.
5339 36. The daemon start-up calls getloadavg() while still root for those OS that
5340 need the first call to be done as root, but it missed one case: when
5341 deliver_queue_load_max is set with deliver_drop_privilege. This is
5342 necessary for the benefit of the queue runner, because there is no re-exec
5343 when deliver_drop_privilege is set.
5345 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5346 was in use, was not putting the data itself into the right store pool;
5347 consequently, it could be overwritten for a subsequent message in the same
5348 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5351 38. Sometimes the final signoff response after QUIT could fail to get
5352 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5353 < 0 before doing a fflush(). This bug looks as though it goes back to the
5354 introduction of TLS in release 3.20, but "sometimes" must have been rare
5355 because the tests only now provoked it.
5361 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5362 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5363 Now it returns an integer. A number of calls were still expecting a T/F
5364 return. Fortuitously, in all cases, the tests worked in OK situations,
5365 which is the norm. However, things would have gone wrong on any write
5366 failures on the smtp file descriptor. This function is used when sending
5367 messages over SMTP and also when doing verify callouts.
5369 2. When Exim is called to do synchronous delivery of a locally submitted
5370 message (the -odf or -odi options), it no longer closes stderr before doing
5373 3. Implemented the mua_wrapper option.
5375 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5377 5. Implemented the functions header_remove(), header_testname(),
5378 header_add_at_position(), and receive_remove_recipient(), and exported them
5381 6. If an ACL "warn" statement specified the addition of headers, Exim already
5382 inserted X-ACL-Warn: at the start if there was no header name. However, it
5383 was not making this test for the second and subsequent header lines if
5384 there were newlines in the string. This meant that an invalid header could
5385 be inserted if Exim was badly configured.
5387 7. Allow an ACL "warn" statement to add header lines at the start or after all
5388 the Received: headers, as well as at the end.
5390 8. Added the rcpt_4xx retry error code.
5392 9. Added postmaster_mailfrom=xxx to callout verification option.
5394 10. Added mailfrom=xxxx to the callout verification option, for verify=
5397 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5398 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5400 12. Inserted some casts to stop certain compilers warning when using pointer
5401 differences as field lengths or precisions in printf-type calls (mostly
5402 affecting debugging statements).
5404 13. Added optional readline() support for -be (dynamically loaded).
5406 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5407 same clock tick as a message's arrival, so that its received time was the
5408 same as the "first fail" time on the retry record, and that message
5409 remained on the queue past the ultimate address timeout, every queue runner
5410 would try a delivery (because it was past the ultimate address timeout) but
5411 after another failure, the ultimate address timeout, which should have then
5412 bounced the address, did not kick in. This was a "< instead of <=" error;
5413 in most cases the first failure would have been in the next clock tick
5414 after the received time, and all would be well.
5416 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5417 being recognized when the domain list was tested by the match_domain
5418 condition in an expansion string.
5420 16. Added the ${str2b64: operator.
5422 17. Exim was always calling setrlimit() to set a large limit for the number of
5423 processes, without checking whether the existing limit was already
5424 adequate. (It did check for the limit on file descriptors.) Furthermore,
5425 errors from getrlimit() and setrlimit() were being ignored. Now they are
5426 logged to the main and panic logs, but Exim does carry on, to try to do its
5427 job under whatever limits there are.
5429 18. Imported PCRE 5.0.
5431 19. Trivial typo in log message " temporarily refused connection" (the leading
5434 20. If the log selector return_path_on_delivery was set and an address was
5435 redirected to /dev/null, the delivery process crashed because it assumed
5436 that a return path would always be set for a "successful" delivery. In this
5437 case, the whole delivery is bypassed as an optimization, and therefore no
5440 21. Internal re-arrangement: the function for sending a challenge and reading
5441 a response while authentication was assuming a zero-terminated challenge
5442 string. It's now changed to take a pointer and a length, to allow for
5443 binary data in such strings.
5445 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5447 23. Exim was not respecting finduser_retries when seeking the login of the
5448 uid under which it was called; it was always trying 10 times. (The default
5449 setting of finduser_retries is zero.) Also, it was sleeping after the final
5450 failure, which is pointless.
5452 24. Implemented tls_on_connect_ports.
5454 25. Implemented acl_smtp_predata.
5456 26. If the domain in control=submission is set empty, Exim assumes that the
5457 authenticated id is a complete email address when it generates From: or
5458 Sender: header lines.
5460 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5461 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5462 chown and chgrp in /bin and hostname in /usr/bin.
5464 28. Exim was keeping the "process log" file open after each use, just as it
5465 does for the main log. This opens the possibility of it remaining open for
5466 long periods when the USR1 signal hits a daemon. Occasional processlog
5467 errors were reported, that could have been caused by this. Anyway, it seems
5468 much more sensible not to leave this file open at all, so that is what now
5471 29. The long-running daemon process does not normally write to the log once it
5472 has entered its main loop, and it closes the log before doing so. This is
5473 so that log files can straightforwardly be renamed and moved. However,
5474 there are a couple of unusual error situations where the daemon does write
5475 log entries, and I had neglected to close the log afterwards.
5477 30. The text of an SMTP error response that was received during a remote
5478 delivery was being truncated at 512 bytes. This is too short for some of
5479 the long messages that one sometimes sees. I've increased the limit to
5482 31. It is now possible to make retry rules that apply only when a message has a
5483 specific sender, in particular, an empty sender.
5485 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5486 it possible to be selective about when SMTP synchronization is enforced.
5488 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5490 32. Implemented hosts_connection_nolog.
5492 33. Added an ACL for QUIT.
5494 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5497 35. Added mailbox_size and mailbox_filecount to appendfile.
5499 36. Added control = no_multiline_responses to ACLs.
5501 37. There was a bug in the logic of the code that waits for the clock to tick
5502 in the case where the clock went backwards by a substantial amount such
5503 that the microsecond fraction of "now" was more than the microsecond
5504 fraction of "then" (but the whole seconds number was less).
5506 38. Added support for the libradius Radius client library this is found on
5507 FreeBSD (previously only the radiusclient library was supported).
5513 1. When certain lookups returned multiple values in the form name=value, the
5514 quoting of the values was not always being done properly. Specifically:
5515 (a) If the value started with a double quote, but contained no whitespace,
5517 (b) If the value contained whitespace other than a space character (i.e.
5518 tabs or newlines or carriage returns) it was not quoted.
5519 This fix has been applied to the mysql and pgsql lookups by writing a
5520 separate quoting function and calling it from the lookup code. The fix
5521 should probably also be applied to nisplus, ibase and oracle lookups, but
5522 since I cannot test any of those, I have not disturbed their existing code.
5524 2. A hit in the callout cache for a specific address caused a log line with no
5525 reason for rejecting RCPT. Now it says "Previous (cached) callout
5526 verification failure".
5528 3. There was an off-by-one bug in the queryprogram router. An over-long
5529 return line was truncated at 256 instead of 255 characters, thereby
5530 overflowing its buffer with the terminating zero. As well as fixing this, I
5531 have increased the buffer size to 1024 (and made a note to document this).
5533 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5534 when Exim is waiting for an SMTP response from a remote server, Exim
5535 restarts its select() call on the socket, thereby resetting its timeout.
5536 This is not a problem when such interrupts are rare. Somebody set up a cron
5537 job to run exiwhat every 2 minutes, which is less than the normal select()
5538 timeout (5 or 10 minutes). This meant that the select() timeout never
5539 kicked in because it was always reset. I have fixed this by comparing the
5540 time when an interrupt arrives with the time at the start of the first call
5541 to select(). If more time than the timeout has elapsed, the interrupt is
5542 treated as a timeout.
5544 5. Some internal re-factoring in preparation for the addition of Sieve
5545 extensions (by MH). In particular, the "personal" test is moved to a
5546 separate function, and given an option for scanning Cc: and Bcc: (which is
5547 not set for Exim filters).
5549 6. When Exim created an email address using the login of the caller as the
5550 local part (e.g. when creating a From: or Sender: header line), it was not
5551 quoting the local part when it contained special characters such as @.
5553 7. Installed new OpenBSD configuration files.
5555 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5556 try to make them clearer.
5558 9. Callout options, other than the timeout value, were being ignored when
5559 verifying sender addresses in header lines. For example, when using
5561 verify = header_sender/callout=no_cache
5563 the cache was (incorrectly) being used.
5565 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5566 only the Cygwin environment.
5568 11. When return_path_on_delivery was set as a log selector, if different remote
5569 addresses in the same message used different return paths and parallel
5570 remote delivery occurred, the wrong values would sometimes be logged.
5571 (Whenever a remote delivery process finished, the return path value from
5572 the most recently started remote delivery process was logged.)
5574 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5575 lines when AUTH and/or TLS are in use. This is the "received protocol"
5576 field. Exim used to use "asmtp" for authenticated SMTP, without any
5577 indication (in the protocol name) for TLS use. Now it follows the RFC and
5578 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5579 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5580 names appear in log lines as well as in Received: header lines.
5582 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5583 extensions, and comparison tests, and to fix some bugs.
5585 14. Changes to the "personal" filter test:
5587 (1) The test was buggy in that it was just doing the equivalent of
5588 "contains" tests on header lines. For example, if a user's address was
5589 anne@some.where, the "personal" test would incorrectly be true for
5591 To: susanne@some.where
5593 This test is now done by extracting each address from the header in turn,
5594 and checking the entire address. Other tests that are part of "personal"
5595 are now done using regular expressions (for example, to check local parts
5596 of addresses in From: header lines).
5598 (2) The list of non-personal local parts in From: addresses has been
5599 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5600 taken from the Sieve specification recommendations.
5602 (3) If the message contains any header line starting with "List-" it is
5603 treated as non-personal.
5605 (4) The test for "circular" in the Subject: header line has been removed
5606 because it now seems ill-conceived.
5608 15. Minor typos in src/EDITME comments corrected.
5610 16. Installed latest exipick from John Jetmore.
5612 17. If headers_add on a router specified a text string that was too long for
5613 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5614 of string_sprintf() is now avoided.
5616 18. $message_body_size was not set (it was always zero) when running the DATA
5617 ACL and the local_scan() function.
5619 19. For the "mail" command in an Exim filter, no default was being set for
5620 the once_repeat time, causing a random time value to be used if "once" was
5621 specified. (If the value happened to be <= 0, no repeat happened.) The
5622 default is now 0s, meaning "never repeat". The "vacation" command was OK
5623 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5624 (I found it when inspecting the code).
5626 20. There is now an overall timeout for performing a callout verification. It
5627 defaults to 4 times the callout timeout, which applies to individual SMTP
5628 commands during the callout. The overall timeout applies when there is more
5629 than one host that can be tried. The timeout is checked before trying the
5630 next host. This prevents very long delays if there are a large number of
5631 hosts and all are timing out (e.g. when the network connections are timing
5632 out). The value of the overall timeout can be changed by specifying an
5633 additional sub-option for "callout", called "maxwait". For example:
5635 verify = sender/callout=5s,maxwait=20s
5637 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5638 to the end before writing, but this should make it even safer).
5640 22. Exim was forgetting that it had advertised PIPELINING for the second and
5641 subsequent messages on an SMTP connection. It was also not resetting its
5642 memory on STARTTLS and an internal HELO.
5644 23. When Exim logs an SMTP synchronization error within a session, it now
5645 records whether PIPELINING has been advertised or not.
5647 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5648 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5649 rather than long int.
5651 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5653 26. Added the never_mail option to autoreply.
5659 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5660 crash if the getsockname() call failed; this can happen if a connection is
5661 closed very soon after it is established. The problem was simply in the
5662 order in which certain operations were done, causing Exim to try to write
5663 to the SMTP stream before it had set up the file descriptor. The bug has
5664 been fixed by making things happen in the correct order.
5670 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5671 before the connection was closed, thus losing the rejection response.
5673 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5674 some early Solaris releases, but causes trouble in current releases where
5675 socklen_t is defined.
5677 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5680 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5683 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5684 that releasing the top of it at the end releases what was used for sub-
5685 expansions (unless the block got too big). However, discard this block if
5686 the first thing is a variable or header, so that we can use its block when
5687 it is dynamic (useful for very large $message_headers, for example).
5689 6. Lookups now cache *every* query, not just the most recent. A new, separate
5690 store pool is used for this. It can be recovered when all lookup caches are
5691 flushed. Lookups now release memory at the end of their result strings.
5692 This has involved some general refactoring of the lookup sources.
5694 7. Some code has been added to the store_xxx() functions to reduce the amount
5695 of flapping under certain conditions.
5697 8. log_incoming_interface used to affect only the <= reception log lines. Now
5698 it causes the local interface and port to be added to several more SMTP log
5699 lines, for example "SMTP connection from", and rejection lines.
5701 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5703 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5705 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5706 could be overwritten at the end of the current message (or the start of a
5707 new message if it was set in a HELO ACL). The value is now preserved for
5708 the duration of the SMTP connection.
5710 12. If a transport had a headers_rewrite setting, and a matching header line
5711 contained an unqualified address, that address was qualified, even if it
5712 did not match any rewriting rules. The underlying bug was that the values
5713 of the flags that permit the existence of unqualified sender and recipient
5714 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5715 non-local messages, and by -bnq for local messages) were not being
5716 preserved with the message after it was received.
5718 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5719 "next input=" as part of the text comprising the host identity instead of
5720 the correct text. The code was using the same buffer for two different
5721 strings. However, depending on which order the printing function evaluated
5722 its arguments, the bug did not always show up. Under Linux, for example, my
5723 test suite worked just fine.
5725 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5726 doesn't work with some older versions of Perl. It has been changed to "my",
5727 which in any case is probably the better facility to use.
5729 15. A really picky compiler found some instances of statements for creating
5730 error messages that either had too many or two few arguments for the format
5733 16. The size of the buffer for calls to the DNS resolver has been increased
5734 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5735 for addresses that have a lot of PTR records. This alleviates a problem; it
5736 does not fully solve it.
5738 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5739 buffer now truncates the list and logs the incident, which is the same
5740 action as happens when Exim is looking up a host name and its aliases.
5741 Previously in this situation something unpredictable would happen;
5742 sometimes it was "internal error: store_reset failed".
5744 18. If a server dropped the connection unexpectedly when an Exim client was
5745 using GnuTLS and trying to read a response, the client delivery process
5746 crashed while trying to generate an error log message.
5748 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5749 string, for example:
5751 warn message = H1: something\nH2: something
5753 the text was added as a single header line from Exim's point of view
5754 though it ended up OK in the delivered message. However, searching for the
5755 second and subsequent header lines using $h_h2: did not work. This has been
5756 fixed. Similarly, if a system filter added multiple headers in this way,
5757 the routers could not see them.
5759 20. Expanded the error message when iplsearch is called with an invalid key to
5760 suggest using net-iplsearch in a host list.
5762 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5763 ACLs are no longer actually imposed (and a message to that effect is
5766 22. If a "gecos" field in a passwd entry contained escaped characters, in
5767 particular, if it contained a \" sequence, Exim got it wrong when building
5768 a From: or a Sender: header from that name. A second bug also caused
5769 incorrect handling when an unquoted " was present following a character
5770 that needed quoting.
5772 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5773 was not being matched caselessly.
5775 24. Arranged for all hyphens in the exim.8 source to be escaped with
5778 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5779 properly. Recipient callout cache records were still being keyed to include
5780 the sender, even when use_sender was set false. This led to far more
5781 callouts that were necessary. The sender is no longer included in the key
5782 when use_sender is false.
5784 26. Added "control = submission" modifier to ACLs.
5786 27. Added the ${base62d: operator to decode base 62 numbers.
5788 28. dnsdb lookups can now access SRV records.
5790 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5791 the configuration file.
5793 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5794 (-v) mode. This makes the output for a verbose queue run more intelligible.
5796 31. Added a use_postmaster feature to recipient callouts.
5798 32. Added the $body_zerocount variable, containing the number of binary zero
5799 bytes in the message body.
5801 33. The time of last modification of the "new" subdirectory is now used as the
5802 "mailbox time last read" when there is a quota error for a maildir
5805 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5807 35. Added +ignore_unknown as a special item in host lists.
5809 36. Code for decoding IPv6 addresses in host lists is now included, even if
5810 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5811 address was recognized as an IP address, but was then not correctly decoded
5812 into binary, causing unexpected and incorrect effects when compared with
5819 1. Very minor rewording of debugging text in manualroute to say "list of
5820 hosts" instead of "hostlist".
5822 2. If verify=header_syntax was set, and a header line with an unqualified
5823 address (no domain) and a large number of spaces between the end of the
5824 name and the colon was received, the reception process suffered a buffer
5825 overflow, and (when I tested it) crashed. This was caused by some obsolete
5826 code that should have been removed. The fix is to remove it!
5828 3. When running in the test harness, delay a bit after writing a bounce
5829 message to get a bit more predictability in the log output.
5831 4. Added a call to search_tidyup() just before forking a reception process. In
5832 theory, someone could use a lookup in the expansion of smtp_accept_max_
5833 per_host which, without the tidyup, could leave open a database connection.
5835 5. Added the variables $recipient_data and $sender_data which get set from a
5836 lookup success in an ACL "recipients" or "senders" condition, or a router
5837 "senders" option, similar to $domain_data and $local_part_data.
5839 6. Moved the writing of debug_print from before to after the "senders" test
5842 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5843 problems for message scanning, either using a data ACL, or using
5844 local_scan() because the Received: header was not generated till after they
5845 were called (in order to set the time as the time of reception completion).
5846 I have revised the way this works. The header is now generated after the
5847 body is received, but before the ACL or local_scan() are called. After they
5848 are run, the timestamp in the header is updated.
5854 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5855 before starting a queue runner without re-exec. This happened only when
5856 deliver_drop_privilege was set or when the Exim user was set to root. The
5857 effect of the bug was that timeouts during subsequent deliveries caused
5858 crashes instead of being properly handled. The handler is now left at its
5859 default (and expected) setting.
5861 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5862 message, again when deliver_drop_privilege is set or Exim is run as root.
5863 The bug described in (1) was not present in this case, but the tidying up
5864 of the other signals was missing. I have made the two cases consistent.
5866 3. The ignore_target_hosts setting on a manualroute router was being ignored
5867 for hosts that were looked up using the /MX notation.
5869 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5872 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5873 operated on the sender address. After changing the $sender_address to <>
5874 for the sender address verify, Exim was re-instated it as the original
5875 (before rewriting) address, but remembering that it had rewritten it, so it
5876 wasn't rewriting it again. This bug also had the effect of breaking the
5877 sender address verification caching when the sender address was rewritten.
5879 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5880 This has been changed so that if the ip literal address matches
5881 ignore_target_hosts, the router declines.
5883 7. Added expansion conditions match_domain, match_address, and match_local_
5884 part (NOT match_host).
5886 8. The placeholder for the Received: header didn't have a length field set.
5888 9. Added code to Exim itself and to exim_lock to test for a specific race
5889 condition that could lead to file corruption when using MBX delivery. The
5890 issue is with the lockfile that is created in /tmp. If this file is removed
5891 after a process has opened it but before that process has acquired a lock,
5892 there is the potential for a second process to recreate the file and also
5893 acquire a lock. This could lead to two Exim processes writing to the file
5894 at the same time. The added code performs the same test as UW imapd; it
5895 checks after acquiring the lock that its file descriptor still refers to
5896 the same named file.
5898 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5899 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5902 11. Added the smtp_active_hostname option. If used, this will typically be made
5903 to depend on the incoming interface address. Because $interface_address is
5904 not set up until the daemon has forked a reception process, error responses
5905 that can happen earlier (such as "too many connections") no longer contain
5908 12. If an expansion in a condition on a "warn" statement fails because a lookup
5909 defers, the "warn" statement is abandoned, and the next ACL statement is
5910 processed. Previously this caused the whole ACL to be aborted.
5912 13. Added the iplsearch lookup type.
5914 14. Added ident_timeout as a log selector.
5916 15. Added tls_certificate_verified as a log selector.
5918 16. Added a global option tls_require_ciphers (compare the smtp transport
5919 option of the same name). This controls incoming TLS connections.
5921 17. I finally figured out how to make tls_require_ciphers do a similar thing
5922 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5923 before starting the TLS session.
5925 18. Tabs are now shown as \t in -bP output.
5927 19. If the log selector return_path_on_delivery was set, Exim crashed when
5928 bouncing a message because it had too many Received: header lines.
5930 20. If two routers both had headers_remove settings, and the first one included
5931 a superfluous trailing colon, the final name in the first list and the
5932 first name in the second list were incorrectly joined into one item (with a
5933 colon in the middle).
5939 1. Added -C and -D options to the exinext utility, mainly to make it easier
5940 to include in the automated testing, but these could be helpful when
5941 multiple configurations are in use.
5943 2. The exinext utility was not formatting the output nicely when there was
5944 an alternate port involved in the retry record key, nor when there was a
5945 message id as well (for retries that were specific to a specific message
5946 and a specific host). It was also confused by IPv6 addresses, because of
5947 the additional colons they contain. I have fixed the IPv4 problem, and
5948 patched it up to do a reasonable job for IPv6.
5950 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5951 delivery, the log line now contains "pipelined" if PIPELINING was used.
5953 4. An SMTP transport process used to panic and die if the bind() call to set
5954 an explicit outgoing interface failed. This has been changed; it is now
5955 treated in the same way as a connect() failure.
5957 5. A reference to $sender_host_name in the part of a conditional expansion
5958 that was being skipped was still causing a DNS lookup. This no longer
5961 6. The def: expansion condition was not recognizing references to header lines
5962 that used bh_ and bheader_.
5964 7. Added the _cache feature to named lists.
5966 8. The code for checking quota_filecount in the appendfile transport was
5967 allowing one more file than it should have been.
5969 9. For compatibility with Sendmail, the command line option
5977 and sets the incoming protocol and host name (for trusted callers). The
5978 host name and its colon can be omitted when only the protocol is to be set.
5979 Note the Exim already has two private options, -pd and -ps, that refer to
5980 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5981 "s", but I don't think that's a major issue.
5983 10. A number of refactoring changes to the code, none of which should affect
5986 (a) The number of logging options was getting close to filling up the
5987 32-bit word that was used as a bit map. I have split them into two classes:
5988 those that are passed in the argument to log_write(), and those that are
5989 only ever tested independently outside of that function. These are now in
5990 separate 32-bit words, so there is plenty of room for expansion again.
5991 There is no change in the user interface or the logging behaviour.
5993 (b) When building, for example, log lines, the code previously used a
5994 macro that called string_cat() twice, in order to add two strings. This is
5995 not really sufficiently general. Furthermore, there was one instance where
5996 it was actually wrong because one of the argument was used twice, and in
5997 one call a function was used. (As it happened, calling the function twice
5998 did not affect the overall behaviour.) The macro has been replaced by a
5999 function that can join an arbitrary number of extra strings onto a growing
6002 (c) The code for expansion conditions now uses a table and a binary chop
6003 instead of a serial search (which was left over from when there were very
6004 few conditions). Also, it now recognizes conditions like "pam" even when
6005 the relevant support is not compiled in: a suitably worded error message is
6006 given if an attempt is made to use such a condition.
6008 11. Added ${time_interval:xxxxx}.
6010 12. A bug was causing one of the ddress fields not to be passed back correctly
6011 from remote delivery subprocesses. The field in question was not being
6012 subsequently used, so this caused to problems in practice.
6014 13. Added new log selectors queue_time and deliver_time.
6016 14. Might have fixed a bug in maildirsizefile handling that threw up
6017 "unexpected character" debug warnings, and recalculated the data
6018 unnecessarily. In any case, I expanded the warning message to give more
6021 15. Added the message "Restricted characters in address" to the statements in
6022 the default ACL that block characters like @ and % in local parts.
6024 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6025 Three changes have been made:
6027 (a) There was a serious bug; a negative response to MAIL caused the whole
6028 recipient domain to be cached as invalid, thereby blocking all messages
6029 to all local parts at the same domain, from all senders. This bug has
6030 been fixed. The domain is no longer cached after a negative response to
6031 MAIL if the sender used is not empty.
6033 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6036 (c) A new callout option, "use_sender" has been added for people who want
6037 the modified behaviour.
6043 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6046 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6047 indeed breaks things for older releases.
6049 3. Added additional logging to the case where there is a problem reading data
6050 from a filter that is running in a subprocess using a pipe, in order to
6051 try to track down a specific problem.
6053 4. Testing facility fudge: when running in the test harness and attempting
6054 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6055 getting "No route to host". Convert this to a timeout.
6057 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6060 6. Some OS don't have socklen_t but use size_t instead. This affects the
6061 fifth argument of getsockopt() amongst other things. This is now
6062 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6063 can be set for individual OS. I have set it for SunOS5, OSF1, and
6064 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6065 some earlier ones do not.
6067 7. Change 4.30/15 was not doing the test caselessly.
6069 8. The standard form for an IPv6 address literal was being rejected by address
6070 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6071 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6072 this, as well as the form without the "IPv6" on the front (but only when
6073 address literals are enabled, of course).
6075 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6077 10. Exim crashed if a message with an empty sender address specified by -f
6078 encountered a router with an errors_to setting. This could be provoked only
6079 by a command such as
6083 where an empty string was supplied; "<>" did not hit this bug.
6085 11. Installed PCRE release 4.5.
6087 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6088 remained set. It is now erased.
6090 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6091 times from message ids (which are base 36 rather than the normal 62).
6093 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6094 were being counted as actual protocol errors, and logged if the log
6095 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6096 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6097 and DATA following a set of rejected RCPTs do not count as protocol errors.
6098 In other words, Exim assumes they were pipelined, though this may not
6099 actually be the case. Of course, in all cases the client gets an
6100 appropriate error code.
6102 15. If a lookup fails in an ACL condition, a message about the failure may
6103 be available; it is used if testing the ACL cannot continue, because most
6104 such messages specify what the cause of the deferral is. However, some
6105 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6106 that caused an old message to be retained and used if a later statement
6107 caused a defer, replacing the real cause of the deferral.
6109 16. If an IP address had so many PTR records that the DNS lookup buffer
6110 was not large enough to hold them, Exim could crash while trying to process
6111 the truncated data. It now detects and logs this case.
6113 17. Further to 4.21/58, another change has been made: if (and only if) the
6114 first line of a message (the first header line) ends with CRLF, a bare LF
6115 in a subsequent header line has a space inserted after it, so as not to
6116 terminate the header.
6118 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6119 unnecessarily, used atoi() instead of strtol(), and didn't check the
6120 termination when getting file sizes from file names by regex.
6122 19. Completely re-implemented the support for maildirsize files, in the light
6123 of a number of problems with the previous contributed implementation
6124 (4.30/29). In particular:
6126 . If the quota is zero, the maildirsize file is maintained, but no quota is
6129 . If the maildir directory does not exist, it is created before any attempt
6130 to write a maildirsize file.
6132 . The quota value in the file is just a cache; if the quota is changed in
6133 the transport, the new value overrides.
6135 . A regular expression is available for excluding directories from the
6138 20. The autoreply transport checks the characters in options that define the
6139 message's headers; it allows continued headers, but it was checking with
6140 isspace() after an embedded newline instead of explicitly looking for a
6143 21. If all the "regular" hosts to which an address was routed had passed their
6144 expiry times, and had not reached their retry times, the address was
6145 bounced, even if fallback hosts were defined. Now Exim should go on to try
6148 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6149 equivalent code in the SMTP transport. Some hosts send humungous responses
6150 to HELO/EHLO, more than 1024 it seems.
6152 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6153 gives compiler warnings in some environments. I've now done it "properly",
6156 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6157 (because of problems with the built-in one) was declared to return uschar *
6158 instead of char *, causing compiler failure.
6160 25. Fixed a file descriptor leak when processing alias/forward files.
6162 26. Fixed a minor format string issue in dbfn.c.
6164 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6166 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6167 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6170 29. When a sender address is verified, it is cached, to save repeating the test
6171 when there is more than one recipient in a message. However, when the
6172 verification involves a callout, it is possible for different callout
6173 options to be set for different recipients. It is too complicated to keep
6174 track of this in the cache, so now Exim always runs a verification when a
6175 callout is required, relying on the callout cache for the optimization.
6176 The overhead is duplication of the address routing, but this should not be
6179 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6180 to be verified with callout=postmaster, and the main callout worked but the
6181 postmaster check failed, the verification correctly failed. However, if a
6182 subsequent RCPT command asked for sender verification *without* the
6183 postmaster check, incorrect caching caused this verification also to fail,
6186 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6187 it was not caching the DNS options (qualify_single, search_parents) that
6188 were used when the lookup failed. A subsequent lookup with different
6189 options therefore always gave the same answer, though there were cases
6190 where it should not have. (Example: a "domains = !$mx_any" option on a
6191 dnslookup router: the "domains" option is always processed without any
6192 widening, but the router might have qualify_single set.) Now Exim uses the
6193 cached value only when the same options are set.
6195 32. Added John Jetmore's "exipick" utility to the distribution.
6197 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6198 than a timeout (e.g. a certificate is required, and is not provided), an
6199 Exim server now closes the connection immediately. Previously it waited for
6200 the client to close - but if the client is SSL, it seems that they each
6201 wait for each other, leading to a delay before one of them times out.
6203 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6204 maintained 0.8.x compatibility because I don't think many are using it, and
6205 it is clearly obsolete.
6207 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6210 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6211 client certificate was expired. A simple patch fixes this, though I don't
6212 understand the full logic of why the verify callback is called multiple
6215 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6216 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6217 which causes problems with some clients (such as the Certicom SSL Plus
6218 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6219 disables the coutermeasure allowing Eudora to connect."
6221 38. Exim was not checking that a write() to a log file succeeded. This could
6222 lead to Bad Things if a log got too big, in particular if it hit a file
6223 size limit. Exim now panics and dies if it cannot write to a log file, just
6224 as it does if it cannot open a log file.
6226 39. Modified OS/Makefile-Linux so that it now contains
6228 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6230 The two -D definitions ensure that Exim is compiled with large file
6231 support, which makes it possible to handle log files that are bigger than
6234 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6235 instance) a domain was checked against a named list that involved a lookup,
6236 causing $domain_data to be set, then another domain was checked against the
6237 same list, then the first domain was re-checked, the value of $domain_data
6238 after the final check could be wrong. In particular, if the second check
6239 failed, it could be set empty. This bug probably also applied to
6242 41. The strip_trailing_dot option was not being applied to the address given
6243 with the -f command-line option.
6245 42. The code for reading a message's header from the spool was incrementing
6246 $received_count, but never initializing it. This meant that the value was
6247 incorrect (doubled) while delivering a message in the same process in which
6248 it was received. In the most common configuration of Exim, this never
6249 happens - a fresh exec is done - but it can happen when
6250 deliver_drop_privilege is set.
6252 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6253 it now includes up to 150 characters of the unexpected data in the log
6256 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6257 and building data strings. The size of both of these buffers was 10 000
6258 bytes - far larger than anybody would *ever* want, thought I. Needless to
6259 say, somebody hit the limit. I have increased the maximum line length to
6260 20 000 and the maximum data length of concatenated lines to 100 000. I have
6261 also fixed two bugs, because there was no checking on these buffers. Tsk,
6262 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6263 buffer is too small.
6265 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6266 lsearch lookups. Now it does.
6268 46. When parsing a route_list item in a manualroute router, a fixed-length
6269 buffer was used for the list of hosts. I made this 1024 bytes long,
6270 thinking that nobody would ever have a list of hosts that long. Wrong.
6271 Somebody had a whole pile of complicated expansion conditions, and the
6272 string was silently truncated, leading to an expansion error. It turns out
6273 that it is easier to change to an unlimited length (owing to other changes
6274 that have happened since this code was originally written) than to build
6275 structure for giving a limitation error. The length of the item that
6276 expands into the list of hosts is now unlimited.
6278 47. The lsearch lookup could not handle data where the length of text line was
6279 more than 4095 characters. Such lines were truncated, leading to shortened
6280 data being returned. It should now handle lines of any length.
6282 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6283 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6286 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6287 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6288 no longer happens in comment lines. A list of replacements is now placed
6289 at the head of all of the source files, except those whose only change is
6290 to replace PERL_COMMAND in the very first #! line.
6292 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6293 messages on the queue, with a bottom-up merge sort, using code contributed
6294 by Michael Haardt. This should make operations like -bp somewhat faster on
6295 large queues. It won't affect queue runners, except when queue_run_in_order
6298 51. Installed eximstats 1.31 in the distribution.
6300 52. Added support for SRV lookups to the dnslookup router.
6302 53. If an ACL referred to $message_body or $message_body_end, the value was not
6303 reset for any messages that followed in the same SMTP session.
6305 54. The store-handling optimization for building very long strings was not
6306 differentiating between the different store pools. I don't think this
6307 actually made any difference in practice, but I've tidied it.
6309 55. While running the routers to verify a sender address, $sender_address
6310 was still set to the sender address. This is wrong, because when routing to
6311 send a bounce to the sender, it would be empty. Therefore, I have changed
6312 it so that, while verifying a sender address, $sender_address is set to <>.
6313 (There is no change to what happens when verifying a recipient address.)
6315 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6316 target A or AAAA records (if not already returned) without resetting the
6317 qualify_single or search_parents options of the DNS resolver. These are
6318 inappropriate in this case because the targets of MX and SRV records must
6319 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6320 target that, when qualified, matched something in the local domain. These
6321 two options are now turned off when doing these lookups.
6323 57. It seems that at least some releases of Reiserfs (which does not have the
6324 concept of a fixed number of inodes) returns zero and not -1 for the
6325 number of available inodes. This interacted badly with check_spool_inodes,
6326 which assumed that -1 was the "no such thing" setting. What I have done is
6327 to check that the total number of inodes is greater than zero before doing
6328 the test of how many are available.
6330 58. When a "warn" ACL statement has a log_message modifier, the message is
6331 remembered, and not repeated. This is to avoid a lot of repetition when a
6332 message has many recipients that cause the same warning to be written.
6333 However, Exim was preserving the list of already written lines for an
6334 entire SMTP session, which doesn't seem right. The memory is now reset if a
6335 new message is started.
6337 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6338 debugging output unless log_rewrite was also set.
6340 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6341 of (int)(handle) when we know that handle contains (void *)(-1).
6343 61. The Exim daemon panic-logs an error return when it closes the incoming
6344 connection. However "connection reset by peer" seems to be common, and
6345 isn't really an error worthy of noting specially, so that particular error
6348 62. When Exim is trying to find all the local interfaces, it used to panic and
6349 die if the ioctl to get the interface flags failed. However, it seems that
6350 on at least one OS (Solaris 9) it is possible to have an interface that is
6351 included in the list of interfaces, but for which you get a failure error
6352 for this call. This happens when the interface is not "plumbed" into a
6353 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6354 failure of the "get flags" call assumes that the interface is down.
6356 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6357 makes life easier for people who are doing arithmetic on fields extracted
6358 from dates, where you often get leading zeros that should not be
6359 interpreted as octal.
6361 64. Added qualify_domain to the redirect router, to override the global
6364 65. If a pathologically long header line contained very many addresses (the
6365 report of this problem mentioned 10 000) and each of them was rewritten,
6366 Exim could use up a very large amount of memory. (It kept on making new
6367 copies of the header line as it rewrote, and never released the old ones.)
6368 At the expense of a bit more processing, the header rewriting function has
6369 been changed so that it no longer eats memory in this way.
6371 66. The generation of the Received: header has been moved from the time that a
6372 message starts to be received, to the time that it finishes. The timestamp
6373 in the Received: header should now be very close to that of the <= log
6374 line. There are two side-effects of this change:
6376 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6377 logged header lines no longer include the local Received: line, because
6378 it has not yet been created. The same applies to a copy of the message
6379 that is returned to a non-SMTP sender when a message is rejected.
6381 (b) When a filter file is tested using -bf, no additional Received: header
6382 is added to the test message. After some thought, I decided that this
6385 This change does not affect the value of $received_for. It is still set
6386 after address rewriting, but before local_scan() is called.
6388 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6390 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6391 gave an unhelpful panic error message, and a defer error. I have managed to
6392 change this behaviour so that it now rejects any supplied certificate,
6393 which seems right, as the list of acceptable certificates is empty.
6395 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6396 gave an unhelpful defer error. I have not managed to make this reject any
6397 supplied certificates, but the error message it gives is "no certificate
6398 supplied", which is not helpful.
6400 70. exigrep's output now also includes lines that are not associated with any
6401 message, but which match the given pattern. Implemented by a patch from
6402 Martin Sluka, which also tidied up the Perl a bit.
6404 71. Recipient callout verification, like sender verification, was using <> in
6405 the MAIL FROM command. This isn't really the right thing, since the actual
6406 sender may affect whether the remote host accepts the recipient or not. I
6407 have changed it to use the actual sender in the callout; this means that
6408 the cache record is now keyed on a recipient/sender pair, not just the
6409 recipient address. There doesn't seem to be a real danger of callout loops,
6410 since a callout by the remote host to check the sender would use <>.
6411 [SEE ABOVE: changed after hitting problems.]
6413 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6414 temporary errors. However, in the case of such a code being given after
6415 the end of a data transmission (i.e. after ".") Exim was failing to write
6416 a retry record for the message. (Yes, there was some broken host that was
6417 actually sending 8xx at this point.)
6419 73. An unknown lookup type in a host list could cause Exim to panic-die when
6420 the list was checked. (An example that provoked this was putting <; in the
6421 middle of a list instead of at the start.) If this happened during a DATA
6422 ACL check, a -D file could be left lying around. This kind of configuration
6423 error no longer causes Exim to die; instead it causes a defer error. The
6424 incident is still logged to the main and panic logs.
6426 74. Buglet left over from Exim 3 conversion. The message "too many messages
6427 in one connection" was written to the rejectlog but not the mainlog, except
6428 when address rewriting (yes!) was being logged.
6430 75. Added write_rejectlog option.
6432 76. When a system filter was run not as root (that is, when system_filter_user
6433 was set), the values of the $n variables were not being returned to the
6434 main process; thus, they were not subsequently available in the $sn
6437 77. Added +return_path_on_delivery log selector.
6439 78. A connection timeout was being treated differently from recipients deferred
6440 when testing hosts_max_try with a message that was older than the host's
6441 retry timeout. (The host should not be counted, thus allowing all hosts to
6442 be tried at least once before bouncing.) This may have been the cause of an
6443 occasionally reported bug whereby a message would remain on the queue
6444 longer than the retry timeout, but would be bounced if a delivery was
6445 forced. I say "may" because I never totally pinned down the problem;
6446 setting up timeout/retry tests is difficult. See also the next item.
6448 79. The ultimate address timeout was not being applied to errors that involved
6449 a combination of host plus message (for example, a timeout on a MAIL
6450 command). When an address resolved to a number of possible hosts, and they
6451 were not all tried for each delivery (e.g. because of hosts_max_try), a
6452 message could remain on the queue longer than the retry timeout.
6454 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6457 81. Fixed an obscure SMTP outgoing bug which required at least the following
6458 conditions: (a) there was another message waiting for the same server;
6459 (b) the server returned 5xx to all RCPT commands in the first message so
6460 that the message was not completed; (c) the server dropped the connection
6461 or gave a negative response to the RSET that Exim sends to abort the
6462 transaction. The observed case was a dropped connection after DATA that had
6463 been sent in pipelining mode. That is, the server had advertised PIPELINING
6464 but was not implementing it correctly. The effect of the bug was incorrect
6465 behaviour, such as trying another host, and this could lead to a crash.
6471 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6472 and daemon.c were passed as pointers to ints; they should have been
6473 pointers to socklen_t variables (which are typically unsigned ints).
6475 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6478 3. Fixed a really odd bug that affected only the testing scheme; patching a
6479 certain fixed string in the binary changed the value of another string that
6480 happened to be identical to the end of the original first string.
6482 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6483 name", it returns that address as the IP address. On some operating
6484 systems (e.g. Solaris), it also passes back the IP address string as the
6485 "host name". However, on others (e.g. Linux), it passes back an empty
6486 string. Exim wasn't checking for this, and was changing the host name to an
6487 empty string, assuming it had been canonicalized.
6489 5. Although rare, it is permitted to have more than one PTR record for a given
6490 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6491 all the names associated with an address, because they do in Solaris.
6492 However, it seems that they do not in Linux for data that comes from the
6493 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6494 I found this out when I moved to a new Linux workstation and tried to run
6495 the Exim test suite.
6497 To get round this problem I have changed the code so that it now does its
6498 own call to the DNS to look up PTR records when searching for a host name.
6499 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6500 addresses that are only in /etc/hosts are still found.
6502 This behaviour is, however, controlled by an option called host_lookup_
6503 order, which defaults to "bydns:byaddr". If people want to use the other
6504 order, or indeed, just use one or the other means of lookup, they can
6505 specify it in this variable.
6507 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6508 some operating systems, this comes back from gethostbyaddr() as an empty
6509 string, and this is what Exim used to test for. However, it seems that in
6510 other systems, "." is yielded. Exim now tests for this case too.
6512 7. The values of check_spool_space and check_log_space are now held internally
6513 as a number of kilobytes instead of an absolute number of bytes. If a
6514 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6515 kilobyte. This means that much larger values can be stored.
6517 8. Exim monitor: an attempt to get the action menu when not actually pointing
6518 at a message produces an empty menu entitled "No message selected". This
6519 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6520 no entries in it ("Shell widget menu has zero width and/or height"). So I
6521 have added a single, blank menu entry in this case.
6523 9. Added ${quote_local_part.
6525 10. MIME decoding is now applied to the contents of Subject: header lines when
6528 11. Now that a reference to $sender_host_address automatically causes a reverse
6529 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6530 host lookup before query-style lookups in lists that might use this
6531 variable. This has therefore been abolished, and the "net-" prefix is no
6532 longer necessary for query-style lookups.
6534 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6535 have been a typo for LFLAGS, so it has been changed.
6537 13. The install script calls Exim with "-C /dev/null" in order to find the
6538 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6539 to be output. However, since Exim outputs its version number before the
6540 error, it didn't break the script. It just looked ugly. I fixed this by
6541 always allowing "-C /dev/null" if the caller is root.
6543 14. Ignore overlarge ACL variable number when reading spool file - insurance
6544 against a later release with more variables having written the file.
6546 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6547 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6548 this, as well as the form without the "IPv6" on the front.
6550 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6551 OS/Makefile-Darwin file.
6553 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6554 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6556 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6557 variable. However, some OS use ints for inodes. I've added cast to long int
6558 to get rid of the compiler warning.
6560 19. I had forgotten to lock out "/../" in configuration file names when
6561 ALT_CONFIG_PREFIX was set.
6563 20. Routers used for verification do not need to specify transports. However,
6564 if such a router generated a host list, and callout was configured, Exim
6565 crashed, because it could not find a port number from the (non-existent)
6566 transport. It now assumes port 25 in this circumstance.
6568 21. Added the -t option to exigrep.
6570 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6571 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6572 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6573 them is set, LOOKUP_LSEARCH is forced.
6575 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6577 24. Added sender and host information to the "rejected by local_scan()" log
6578 line; previously there was no indication of these.
6580 25. Added .include_if_exists.
6582 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6583 Linux. It turns out that not all file systems support this. Apparently some
6584 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6585 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6586 not supported on the file descriptor, is now ignored when Exim is trying to
6587 sync a directory. This applies only to Linux.
6589 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6591 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6592 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6595 29. Added support for maildirsize files from supplied patch (modified a bit).
6597 30. The use of :fail: followed by an empty string could lead Exim to respond to
6598 sender verification failures with (e.g.):
6600 550 Verification failed for <xxx>
6601 550 Sender verify failed
6603 where the first response line was missing the '-' that indicates it is not
6604 the final line of the response.
6606 31. The loop for finding the name of the user that called Exim had a hardwired
6607 limit of 10; it now uses the value of finduser_retries, which is used for
6608 all other user lookups.
6610 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6613 33. Exim was neglecting to zero errno before one call of strtol() when
6614 expanding a string and expecting an integer value. On some systems this
6615 resulted in spurious "integer overflow" errors. Also, it was casting the
6616 result into an int without checking.
6618 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6619 did not work. The code looks as if it has *never* worked, though it appears
6620 to have been documented since at least release 1.62. I have made it work.
6622 35. The "timeout_DNS" error in retry rules, also documented since at least
6623 1.62, also never worked. As it isn't clear exactly what this means, and
6624 clearly it isn't a major issue, I have abolished the feature by treating it
6625 as "timeout", and writing a warning to the main and panic logs.
6627 36. The display of retry rules for -brt wasn't always showing the error code
6630 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6631 timeout_connect_A, timeout_connect_MX.
6633 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6634 to the empty sender.
6636 39. The daemon was not analysing the content of -oX till after it had closed
6637 stderr and disconnected from the controlling terminal. This meant that any
6638 syntax errors were only noted on the panic log, and the return code from
6639 the command was 0. By re-arranging the code a little, I've made the
6640 decoding happen first, so such errors now appear on stderr, and the return
6641 code is 1. However, the actual setting up of the sockets still happens in
6642 the disconnected process, so errors there are still only recorded on the
6645 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6646 connections (as happens on some IP stacks) was logged at start up time as
6647 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6648 it from "IPv6 and IPv4", which means that two separate sockets are being
6651 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6652 says whether AF_INET or AF_INET6 was passed as an argument.
6654 42. Exiwhat output was messed up when time zones were included in log
6657 43. Exiwhat now gives more information about the daemon's listening ports,
6658 and whether -tls-on-connect was used.
6660 44. The "port" option of the smtp transport is now expanded.
6662 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6663 silently ignored. Now an error message is written to the main and panic
6666 46. There's a new ACL modifier called "logwrite" which writes to a log file
6667 as soon as it is encountered.
6669 47. Added $local_user_uid and $local_user_gid at routing time.
6671 48. Exim crashed when trying to verify a sender address that was being
6674 49. Exim was recognizing only a space character after ".include". It now also
6675 recognizes a tab character.
6677 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6678 extracting the relevant information from the specification. The man page no
6679 longer contains scrambled data for the -d option, and I've added a section
6680 at the front about calling Exim under different names.
6682 51. Added "extra_headers" argument to the "mail" command in filter files.
6684 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6687 53. Installed eximstats 1.29.
6689 54. Added transport_filter_timeout as a generic transport option.
6691 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6692 Cc: header lines. This was required by RFC 822, but it not required by RFC
6695 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6696 incoming messages that did not have them. Now it does so only if the
6697 message originates locally, that is, if there is no associated remote host
6698 address. When Resent- header lines are present, this applies to the Resent-
6699 lines rather than the non-Resent- lines.
6701 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6702 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6704 58. Messages for configuration errors now include the name of the main
6705 configuration file - useful now that there may be more than one file in a
6706 list (.included file names were always shown).
6708 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6709 for those rare installations that do not start the daemon as root or run it
6710 setuid root. I've cut out the call to initgroups() if the daemon is not
6713 60. The Exim user and group can now be bound into the binary as text strings
6714 that are looked up at the start of Exim's processing.
6716 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6718 62. Added $mailstore_basename variable.
6720 63. Installed patch to sieve.c from Michael Haardt.
6722 64. When Exim failed to open the panic log after failing to open the main log,
6723 the original message it was trying to log was written to stderr and debug
6724 output, but if they were not available (the usual case in production), it
6725 was lost. Now it is written to syslog before the two lines that record the
6726 failures to open the logs.
6728 65. Users' Exim filters run in subprocesses under the user's uid. It is
6729 possible for a "deliver" command or an alias in a "personal" command to
6730 provoke an address rewrite. If logging of address rewriting is configured,
6731 this fails because the process is not running as root or exim. There may be
6732 a better way of dealing with this, but for the moment (because 4.30 needs
6733 to be released), I have disabled address rewrite logging when running a
6734 filter in a non-root, non-exim process.
6740 1. The buildconfig auxiliary program wasn't quoting the value set for
6741 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6742 not defined. This bug was masked in 4.22 by the effect that was fixed in
6745 2. Some messages that were rejected after a message id was allocated were
6746 shown as "incomplete" by exigrep. It no longer does this for messages that
6747 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6749 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6750 have allow_domain_literals set, the ID did not get logged in the <= line.
6751 Domain literals are now always recognized in Message-ID: header lines.
6753 4. The first argument for a ${extract expansion item is the key name or field
6754 number. Leading and trailing spaces in this item were not being ignored,
6755 causing some misleading effects.
6757 5. When deliver_drop_privilege was set, single queue runner processes started
6758 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6759 same command in the process it spins off) were not dropping privilege.
6761 6. When the daemon running as "exim" started a queue runner, it always
6762 re-executed Exim in the spun-off process. This is a waste of effort when
6763 deliver_drop_privilege is set. The new process now just calls the
6764 queue-runner function directly.
6770 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6773 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6774 ignored. Though the use of -oP was forcing the writing of a pid file, it
6775 was always written to the default place.
6777 3. If the message "no IP address found for host xxxx" is generated during
6778 incoming verification, it is now followed by identification of the incoming
6779 connection (so you can more easily find what provoked it).
6781 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6783 5. Added some features to "harden" Exim a bit more against certain attacks:
6785 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6786 be put in Local/Makefile. This is like the never_users runtime option,
6787 but it cannot be overridden. The default setting is "root".
6789 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6790 prefix string with which any file named in a -C command line option
6793 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6794 is retained for -C and -D only if the caller of Exim is root. Without
6795 it, the exim user may also use -C and -D and retain privilege.
6797 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6798 command line option is disabled.
6800 6. Macro names set by the -D option must start with an upper case letter, just
6801 like macro names defined in the configuration file.
6803 7. Added "dereference=" facility to LDAP.
6805 8. Two instances of the typo "uknown" in the source files are fixed.
6807 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6808 the Configure-Makefile script screwed up while processing it.
6810 10. Incorporated PCRE 4.4.
6812 11. The SMTP synchronization check was not operating right at the start of an
6813 SMTP session. For example, it could not catch a HELO sent before the client
6814 waited for the greeting. There is now a check for outstanding input at the
6815 point when the greeting is written. Because of the duplex, asynchronous
6816 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6817 way, but not yet received, when the check is performed.
6819 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6820 on TCP/IP sockets, because this apparently causes some broken clients to
6823 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6824 unchanged) from the Cygwin maintainer.
6826 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6827 was supported instead of testing for mbx. Effectively a typo.
6829 15. The spa authenticator server code was not checking that the input it
6830 received was valid base64.
6832 16. The debug output line for the "set" modifier in ACLs was not showing the
6833 name of the variable that was being set.
6835 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6837 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6838 lookup on its own. Something else was needed to trigger the lookup. For
6839 example, a match in host_lookup or the need for a host name in a host list.
6840 Now, if $sender_host_name is referenced and the host name has not yet been
6841 looked up, a lookup is performed. If the lookup fails, the variable remains
6842 empty, and $host_lookup_failed is set to "1".
6844 19. Added "eqi" as a case-independent comparison operator.
6846 20. The saslauthd authentication condition could segfault if neither service
6847 nor realm was specified.
6849 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6850 error message that was logged was misleading, and incoming SMTP
6851 connections were dropped. The message is now more accurate, and temporary
6852 errors are given to SMTP connections.
6854 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6855 (except RSET) with a 421 error, until QUIT is received. However, it was
6856 failing to send a response to QUIT.
6858 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6859 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6860 ACL is now run after the helo_try_verify_hosts code.
6862 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6863 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6864 case-independent, so other case variants are also recognized). Apparently
6865 some systems use these upper case variants.
6867 25. If more than two messages were waiting for the same host, and a transport
6868 filter was specified for the transport, Exim sent two messages over the
6869 same TCP/IP connection, and then failed with "socket operation on non-
6870 socket" when it tried to send the third.
6872 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6874 27. The extern definition of crypt16() in expand.c was not being excluded when
6875 the OS had its own crypt16() function.
6877 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6878 as a preferred synonym for return_size_limit, both as an option and as an
6881 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6883 30. Changed the default configuration ACL to relax the local part checking rule
6884 for addresses that are not in any local domains. For these addresses,
6885 slashes and pipe symbols are allowed within local parts, but the sequence
6886 /../ is explicitly forbidden.
6888 31. SPA server authentication was not clearing the challenge buffer before
6891 32. log_message in a "warn" ACL statement was writing to the reject log as
6892 well as to the main log, which contradicts the documentation and doesn't
6893 seem right (because no rejection is happening). So I have stopped it.
6895 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6896 However, I am unable to do any testing of this.
6898 34. Fixed an infelicity in the appendfile transport. When checking directories
6899 for a mailbox, to see if any needed to be created, it was accidentally
6900 using path names with one or more superfluous leading slashes; tracing
6901 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6903 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6904 contents are added to the log line that is written for every discarded
6905 recipient. (Previously a log_message setting was ignored.)
6907 36. The ${quote: operator now quotes the string if it is empty.
6909 37. The install script runs exim in order to find its version number. If for
6910 some reason other than non-existence or emptiness, which it checks, it
6911 could not run './exim', it was installing it with an empty version number,
6912 i.e. as "exim-". This error state is now caught, and the installation is
6915 38. An argument was missing from the function that creates an error message
6916 when Exim fails to connect to the socket for saslauthd authentication.
6917 This could cause Exim to crash, or give a corrupted message.
6919 39. Added isip, isip4, and isip6 to ${if conditions.
6921 40. The ACL variables $acl_xx are now saved with the message, and can be
6922 accessed later in routers, transports, and filters.
6924 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6925 strings in the file are not string-expanded.
6927 42. If a MAIL command specified a SIZE value that was too large to fit into an
6928 int variable, the check against message_size_limit failed. Such values are
6929 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6930 day this will have to be increased, but I don't think I want to be around
6931 when emails are that large.
6938 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6939 iconv() is not standard in FreeBSD.
6941 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6942 IPv6 enabled. The observed symptom was a segmentation fault on return from
6943 the function os_common_find_running_interfaces() in src/os.c.
6945 3. In the check_special_case() function in daemon.c I had used "errno" as an
6946 argument name, which causes warnings on some systems. This was basically a
6947 typo, since it was named "eno" in the comments!
6949 4. The code that waits for the clock to tick (at a resolution of some fraction
6950 of a second) so as to ensure message-id uniqueness was always waiting for
6951 at least one whole tick, when it could have waited for less. [This is
6952 almost certainly not relevant at current processor speeds, where it is
6953 unlikely to ever wait at all. But we try to future-proof.]
6955 5. The function that sleeps for a time interval that includes fractions of a
6956 second contained a race. It did not block SIGALRM between setting the
6957 timer, and suspending (a couple of lines later). If the interval was short
6958 and the sigsuspend() was delayed until after it had expired, the suspension
6959 never ended. On busy systems this could lead to processes getting stuck for
6962 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6963 process, before it forks any delivery processes. The open lookup caching
6964 mechanism meant that the open file or database connection was passed into
6965 the delivery process. The problem was that delivery processes always tidy
6966 up cached lookup data. This could cause a problem for the next delivery
6967 process started by the queue runner, because the external queue runner
6968 process does not know about the closure. So the next delivery process
6969 still has data in the lookup cache. In the case of a file lookup, there was
6970 no problem because closing a file descriptor in a subprocess doesn't affect
6971 the parent. However, if the lookup was caching a connection to a database,
6972 the connection was closed, and the second delivery process was likely to
6973 see errors such as "PGSQL: query failed: server closed the connection
6974 unexpectedly". The problem has been fixed by closing all cached lookups
6975 in a queue runner before running a delivery process.
6977 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6978 seem to have the "const" qualifier which it has on other OS. I've
6981 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6982 *of the same type* (client or server) with the same public name that an
6983 error should be diagnosed.
6985 9. When Exim looked up a host name for an IP address, but failed to find the
6986 original IP address when looking up the host name (a safety check), it
6987 output the message "<ip address> does not match any IP for NULL", which was
6988 confusing, to say the least. The bug was that the host name should have
6989 appeared instead of "NULL".
6991 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6992 user that is built into the binary, and the -C or -D options is used, root
6993 privilege is dropped before the configuration file is read. In addition,
6994 logging is switched to stderr instead of the normal log files. If the
6995 configuration then re-defines the Exim user, the unprivileged environment
6996 is probably not what is expected, so Exim logs a panic warning message (but
6999 However, if deliver_drop_privilege is set, the unprivileged state may well
7000 be exactly what is intended, so the warning has been cut out in that case,
7001 and Exim is allowed to try to write to its normal log files.
7007 1. smtp_return_error_details was not giving details for temporary sender
7008 or receiver verification errors.
7010 2. Diagnose a configuration error if two authenticators have the same public
7013 3. Exim used not to create the message log file for a message until the first
7014 delivery attempt. This could be confusing when incoming messages were held
7015 for policy or load reasons. The message log file is now created at the time
7016 the message is received, and an initial "Received" line is written to it.
7018 4. The automatically generated man page for command line options had a minor
7019 bug that caused no ill effects; however, a more serious problem was that
7020 the procedure for building the man page automatically didn't always
7021 operate. Consequently, release 4.20 contains an out-of-date version. This
7022 shouldn't happen again.
7024 5. When building Exim with embedded Perl support, the script that builds the
7025 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7026 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7028 6. The freeze_tell option was not being used for messages that were frozen on
7029 arrival, either by an ACL or by local_scan().
7031 7. Added the smtp_incomplete_transaction log selector.
7033 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7034 was accepting AUTH without a new EHLO.
7036 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7037 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7040 10. Logging of TCP/IP connections (when configured) now happens in the main
7041 daemon process instead of the child process, so that the TCP/IP connection
7042 count is more accurate (but it can never be perfect).
7044 11. The use of "drop" in a nested ACL was not being handled correctly in the
7045 outer ACL. Now, if condition failure induced by the nested "drop" causes
7046 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7047 or "require"), the connection is dropped.
7049 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7050 that yield "discard" can now be used with an "accept" or a "discard" verb,
7051 but an error is generated for any others (because I can't see a useful way
7052 to define what should happen).
7054 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7055 are now processed in the same way as lines in the Exim configuration file.
7056 In particular, continuation lines are supported.
7058 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7060 15. Added -ti meaning -t -i.
7062 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7063 domains, and warn by logging if others are found.
7065 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7066 returned from GIFCONF when Exim is trying to find the list of interfaces on
7067 a host. The code in os.c has been modified to copy each ifreq to an aligned
7068 structure in all cases.
7070 Also, in some cases, the returned ifreq's were being copied to a 'struct
7071 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7072 means the last couple of bytes of an IPv6 address could be chopped if the
7073 ifreq contained only a normal sockaddr (14 bytes storage).
7075 18. Named domain lists were not supported in the hosts_treat_as_local option.
7076 An entry such as +xxxx was not recognized, and was treated as a literal
7079 19. Ensure that header lines added by a DATA ACL are included in the reject log
7080 if the ACL subsequently rejects the message.
7082 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7083 MD5 (which is deprecated).
7085 21. When testing a filter file using -bf, Exim was writing a message when it
7086 took the sender from a "From " line in the message, but it was not doing so
7087 when it took $return_path from a Return-Path: header line. It now does.
7089 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7090 with a valid header line field name (a series of printing characters
7091 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7093 23. Changed "disc" in the source to "disk" to conform to the documentation and
7094 the book and for uniformity.
7096 24. Ignore Sendmail's -Ooption=value command line item.
7098 25. When execve() failed while trying to run a command in a pipe transport,
7099 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7100 could be confused with a return value of 69 from the command itself. This
7101 has been changed to 127, the value the shell returns if it is asked to run
7102 a non-existent command. The wording for the related log line suggests a
7103 non-existent command as the problem.
7105 26. If received_header_text expands to an empty string, do not add a Received:
7106 header line to the message. (Well, it adds a token one on the spool, but
7107 marks it "old" so that it doesn't get used or transmitted.)
7109 27. Installed eximstats 1.28 (addition of -nt option).
7111 28. There was no check for failure on the call to getsockname() in the daemon
7112 code. This can fail if there is a shortage of resources on the system, with
7113 ENOMEM, for example. A temporary error is now given on failure.
7115 29. Contrary to the C standard, it seems that in some environments, the
7116 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7117 program. Exim now does this explicitly; it affects the formatting of
7118 timestamps using strftime().
7120 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7121 complaints. I've now initialized all the variables, to avoid this.
7123 32. Header lines added by a system filter were not being "seen" during
7124 transport-time rewrites.
7126 33. The info_callback() function passed to OpenSSL is set up with type void
7127 (*)(SSL *, int, int), as described somewhere. However, when calling the
7128 function (actually a macro) that sets it up, the type void(*)() is
7129 expected. I've put in a cast to prevent warnings from picky compilers.
7131 34. If a DNS black list lookup found a CNAME record, but there were no A
7132 records associated with the domain it pointed at, Exim crashed.
7134 35. If a DNS black list lookup returned more than one A record, Exim ignored
7135 all but the first. It now scans all returned addresses if a particular IP
7136 value is being sought. In this situation, the contents of the
7137 $dnslist_value variable are a list of all the addresses, separated by a
7140 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7141 to accept a host name and all its aliases if the forward lookup for any of
7142 them yielded the IP address of the incoming connection. Now it accepts only
7143 those names whose forward lookup yields the correct IP address. Any other
7144 names are discarded. This closes a loophole whereby a rogue DNS
7145 administrator could create reverse DNS records to break through a
7146 wildcarded host restriction in an ACL.
7148 37. If a user filter or a system filter that ran in a subprocess used any of
7149 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7150 the wrong values were passed to the pipe command ($thisaddress had the
7151 value of $0, $0 had the value of $1, etc). This bug was introduced by
7152 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7154 38. Improved the line breaking for long SMTP error messages from ACLs.
7155 Previously, if there was no break point between 40 and 75 characters, Exim
7156 left the rest of the message alone. Two changes have been made: (a) I've
7157 reduced the minimum length to 35 characters; (b) if it can't find a break
7158 point between 35 and 75 characters, it looks ahead and uses the first one
7159 that it finds. This may give the occasional overlong line, but at least the
7160 remaining text gets split now.
7162 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7163 file descriptors might be low, and that setting 1000 would always raise it.
7164 It turns out that in some environments, the limit is already over 1000 and
7165 that lowering it causes trouble. So now Exim takes care not to decrease it.
7167 40. When delivering a message, the value of $return_path is set to $sender_
7168 address at the start of routing (routers may change the value). By an
7169 oversight, this default was not being set up when an address was tested by
7170 -bt or -bv, which affected the outcome if any router or filter referred to
7173 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7174 log only when "message" or "log_message" are set. However, if one of the
7175 conditions was an address verification, or a call to a nested ACL, the
7176 messages generated by the underlying test were being passed through. This
7177 no longer happens. The underlying message is available in $acl_verify_
7178 message for both "message" and "log_message" expansions, so it can be
7179 passed through if needed.
7181 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7182 new expansion $bh_ to give the encoded byte string without charset
7183 translation. Translation happens only if iconv() is available; HAVE_ICONV
7184 indicates this at build time. HEADERS_CHARSET gives the charset to
7185 translate to; headers_charset can change it in the configuration, and
7186 "headers charset" can change it in an individual filter file.
7188 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7189 that creates RFC 2047 encoded "words" labels them as that charset instead
7190 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7191 expansion operator; (ii) when Exim creates a From: line for a local
7192 message; (iii) when a header line is rewritten to include a "phrase" part.
7194 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7195 buggy, causing it to skip the first lines of messages whose message ID
7196 ended in 'D'. This would not have bitten before Exim release 4.14, because
7197 message IDs were unlikely to end in 'D' before then. The effect was to have
7198 incorrect size information for certain domains.
7200 45. #include "config.h" was missing at the start of the crypt16.c module. This
7201 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7204 46. If there was a timeout during a "random" callout check, Exim treated it as
7205 a failure of the random address, and carried on sending RSET and the real
7206 address. If the delay was just some slowness somewhere, the response to the
7207 original RCPT would be taken as a response to RSET and so on, causing
7208 mayhem of various kinds.
7210 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7211 when I implemented it. It didn't allow for the fact that some option values
7212 may legitimately be negative (e.g. size_addition), and it didn't even do
7213 the right test for positive values.
7215 48. Domain names in DNS records are case-independent. Exim always looks them up
7216 in lower case. Some resolvers return domain names in exactly the case they
7217 appear in the zone file, that is, they may contain uppercase letters. Not
7218 all resolvers do this - some return always lower case. Exim was treating a
7219 change of case by a resolver as a change of domain, similar to a widening
7220 of a domain abbreviation. This triggered its re-routing code and so it was
7221 trying to route what was effectively the same domain again. This normally
7222 caused routing to fail (because the router wouldn't handle the domain
7223 twice). Now Exim checks for this case specially, and just changes the
7224 casing of the domain that it ultimately uses when it transmits the message
7227 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7230 50. If a filter generated a file delivery with a non-absolute name (possible if
7231 no home directory exists for the router), the forbid_file option was not
7234 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7235 the existing equality matching.
7237 52. Exim was using ints instead of ino_t variables in some places where it was
7238 dealing with inode numbers.
7240 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7241 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7242 called TMPDIR, and if it finds it is different, it changes its value.
7244 54. The smtp_printf() function is now made available to local_scan() so
7245 additional output lines can be written before returning. There is also an
7246 smtp_fflush() function to enable the detection of a dropped connection.
7247 The variables smtp_input and smtp_batched_input are exported to
7250 55. Changed the default runtime configuration: the message "Unknown user"
7251 has been removed from the ACL, and instead placed on the localuser router,
7252 using the cannot_route_message feature. This means that any verification
7253 failures that generate their own messages won't get overridden. Similarly,
7254 the "Unrouteable address" message that was in the ACL for unverifiable
7255 relay addresses has also been removed.
7257 56. Added hosts_avoid_esmtp to the smtp transport.
7259 57. The exicyclog script was not checking for the esoteric option
7260 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7261 will work only if exicyclog is run under the appropriate euid.
7263 58. Following a discussion on the list, the rules by which Exim recognises line
7264 endings on incoming messages have been changed. The -dropcr and drop_cr
7265 options are now no-ops, retained only for backwards compatibility. The
7266 following line terminators are recognized: LF CRLF CR. However, special
7267 processing applies to CR:
7269 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7270 nor a local message in the state where . is a terminator.
7272 (ii) If a bare CR is encountered in a header line, an extra space is added
7273 after the line terminator so as not to end the header. The reasoning
7274 behind this is that bare CRs in header lines are most likely either
7275 to be mistakes, or people trying to play silly games.
7277 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7278 was being incorrectly given as 18 bytes larger than it should have been.
7279 This is a VOB (very old bug).
7281 60. This may never have affected anything current, but just in case it has:
7282 When the local host is found other than at the start of a list of hosts,
7283 the local host, those with the same MX, and any that follow, are discarded.
7284 When the list in question was part of a longer list of hosts, the following
7285 hosts (not currently being processed) were also being discarded. This no
7286 longer happens. I'm not sure if this situation could ever has previously
7289 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7292 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7293 header. This is something that is recommended in a new Internet Draft, and
7294 is something that is documented as being done by Sendmail. There are two
7295 possible values. For messages generated by the autoreply transport, Exim
7298 Auto-Submitted: auto-replied
7300 whereas for all other generated messages (e.g. bounces) it adds
7302 Auto-Submitted: auto-generated
7304 63. The "personal" condition in filters now includes a test for the
7305 Auto-Submitted: header. If it contains the string "auto-" the message it
7306 not considered personal.
7308 64. Added rcpt_include_affixes as a generic transport option.
7310 65. Added queue_only_override (default true).
7312 66. Added the syslog_duplication option.
7314 67. If what should have been the first header line of a message consisted of
7315 a space followed by a colon, Exim was mis-interpreting it as a header line.
7316 It isn't of course - it is syntactically invalid and should therefore be
7317 treated as the start of the message body. The misbehaviour could have
7318 caused a number of strange effects, including loss of data in subsequent
7319 header lines, and spool format errors.
7321 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7322 client host had authenticated. This control can now be exercised by an ACL
7323 for more flexibility.
7325 69. By default, callouts do not happen when testing with -bh. There is now a
7326 variant, -bhc, which does actually run the callout code, including
7327 consulting and updating the callout cache.
7329 70. Added support for saslauthd authentication, courtesy of Alexander
7332 71. If statvfs() failed on the spool or log directories while checking their
7333 size for availability, Exim confusingly gave the error "space shortage".
7334 Furthermore, in debugging mode it crashed with a floating point exception.
7335 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7336 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7337 serious problem, Exim now writes to the main and panic logs when this
7338 happens, with details of the failure. It then refuses to accept the
7339 incoming message, giving the message "spool directory problem" or "log
7340 directory problem" with a 421 code for SMTP messages.
7342 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7343 0, 1, and 2 exist, because some OS complain for execs without them (see
7344 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7345 descriptors. However, the code omitted to check that the open succeeded,
7346 causing mysterious errors if for some reason the permissions on /dev/null
7347 got screwed. Now Exim writes a message to the main and panic logs, and
7348 bombs out if it can't open /dev/null.
7350 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7351 interact so that it is all more flexible. It is supposed to remain
7352 backwards compatible. Also added extra_local_interfaces.
7354 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7355 to bomb out with an assertion failure - to the client this appears as a
7356 connection drop. This problem occurs in the part of the code that was taken
7357 from the Samba project. Fortunately, the assertion is in a very simple
7358 function, so I have fixed this by reproducing the function inline in the
7359 one place where it is called, and arranging for authentication to fail
7360 instead of killing the process with assert().
7362 75. The SPA client code was not working when the server requested OEM rather
7363 than Unicode encoding.
7365 76. Added code to make require_files with a specific uid setting more usable in
7366 the case where statting the file as root fails - usually a non-root-mounted
7367 NFS file system. When this happens and the failure is EACCES, Exim now
7368 forks a subprocess and does the per-uid checking as the relevant uid.
7370 77. Added process_log_path.
7372 78. If log_file_path was not explicitly set, a setting of check_log_space or
7373 check_log_inodes was ignored.
7375 79. If a space check for the spool or log partitions fails, the incident is now
7376 logged. Of course, in the latter case the data may get lost...
7378 80. Added the %p formatting code to string_format() so that it can be used to
7379 print addresses in debug_print(). Adjusted all the address printing in the
7380 debugging in store.c to use %p rather than %d.
7382 81. There was a concern that a line of code in smtp_in.c could overflow a
7383 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7384 initially expressed, the concern was not well-founded, because trailing
7385 spaces are removed early. However, if the trailing spaces were followed by
7386 a NULL, they did not get removed, so the overflow was possible. Two fixes
7389 (a) I re-wrote the offending code in a cleaner fashion.
7390 (b) If an incoming SMTP command contains a NULL character, it is rejected
7393 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7394 runs initgroups(), so that if the Exim user is in any additional groups,
7395 they will be used during message reception.
7401 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob