1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
25 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
26 files in the spool were present and unlocked. A queue-runner could spot
27 them, resulting in a duplicate delivery. Fix that by doing the unlock
28 after the unlink. Investigation by Tim Stewart. Take the opportunity to
29 add more error-checking on spoolfile handling while that code is being
32 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
33 No known attacks, no CVE, this is defensive hardening.
35 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
36 a queue-runner could start a delivery while other operations were ongoing.
37 Cutthrough delivery was a common victim, resulting in duplicate delivery.
38 Found and investigated by Tim Stewart. Fix by using the open message data
39 file handle rather than opening another, and not locally closing it (which
40 releases a lock) for that case, while creating the temporary .eml format
41 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
43 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
44 $sender_verify_failure/$recipient_verify_failure to "random".
46 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
49 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
50 Previously this would segfault.
52 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
55 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
56 like zero, since the resolver should be doing this for us, But we need one
57 as a CNAME but no MX presence gets the CNAME returned; we need to check
58 that doesn't point to an MX to declare it "no MX returned" rather than
59 "error, loop". A new main option is added so the older capability of
60 following some limited number of chain links is maintained.
62 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
64 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
65 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
66 now-deprecated earlier definitions used only specified the range up to TLS
67 1.2 (in the older-version library docs).
69 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
71 JH/15 Rework TLS client-side context management. Stop using a global, and
72 explicitly pass a context around. This enables future use of TLS for
73 connections to service-daemons (eg. malware scanning) while a client smtp
74 connection is using TLS; with cutthrough connections this is quite likely.
76 JH/16 Fix ARC verification to do AS checks in reverse order.
78 JH/17 Support a "tls" option on the ${readsocket } expansion item.
80 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
81 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
82 Previously the "utf8" would be re-prepended for every additional message.
84 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
85 Previously thery were accepted, resulting in issues when attempting to
86 forward messages to a non-supporting MTA.
88 PP/02 Let -n work with printing macros too, not just options.
90 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
91 one parent address was copied, and bogus data was used at delivery-logging
92 time. Either a crash (after delivery) or bogus log data could result.
93 Discovery and analysis by Tim Stewart.
95 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
96 Previously if the string ended mid-character, we did not insert the
97 promised '?' replacement.
99 PP/04 Documentation: current string operators work on bytes, not codepoints.
101 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
102 should pack well giving a smaller memory footprint so better caching and
103 therefore performance. Group the declarations where this can't be done so
104 that the byte-sized flag variables are not interspersed among pointer
105 variables, giving a better chance of good packing by the compiler.
107 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
108 non-null, to avoid issues with sites running BATV. Previously reports were
109 sent with an empty envelope sender so looked like bounces.
111 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
112 The ignore_error flag wasn't being returned from the filter subprocess so
113 was not set for later routers. Investigation and fix by Matthias Kurz.
115 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
116 and a msg:complete for the whole, when a message is manually removed using
117 -Mrm. Developement by Matthias Kurz, hacked on by JH.
119 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
120 a "Gnu special" function, asprintf() in the DB utility binary builds; I
121 hope that is portable enough.
123 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
124 requiring a known-CA anchor certificate; make it now rely entirely on the
125 TLSA as an anchor. Checking the name on the leaf cert against the name
126 on the A-record for the host is still done for TA (but not for EE mode).
128 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
129 included in delivery lines for non-proxied connections, when compiled with
130 SUPPORT_SOCKS and running with proxy logging enabled.
132 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
133 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
134 move the existing event to fire before the normal logging of message
135 failure so that custom logging is bracketed by normal logging.
137 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
138 msg:fail:internal event. Developement by Matthias Kurz.
140 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
141 far too small for todays use of crypto signatures stored there. Go all
142 the way to the max DNS message size of 64kB, even though this might be
143 overmuch for IOT constrained device use.
145 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
146 copy a string over itself. The library routine is documented as not
147 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
149 JH/32 For main options check_spool_space and check_inode_space, where the
150 platform supports 64b integers, support more than the previous 2^31 kB
151 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
152 the previous G, M, k.
154 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
155 $authenticated_fail_id variable on authentication failure. Previously
158 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
159 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
160 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
161 GNUTLS_SEC_PARAM_MEDIUM.
163 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
164 side. Previously we would continue as if no SNI had been received.
166 JH/36 Harden the handling of string-lists. When a list consisted of a sole
167 "<" character, which should be a list-separator specification, we walked
168 off past the nul-terimation.
170 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
171 causes) even when the retry time is not yet met. Previously they were
172 not, meaning that when (say) an account was over-quota and temp-rejecting,
173 and multiple senders' messages were queued, only one sender would get
174 notified on each configured delay_warning cycle.
180 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
181 When redis_servers is set to a list of > 1 element, and the Redis servers
182 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
183 case of MOVED into a DEFER case instead, thus moving the query onto the
184 next server in the list. For a cluster of N elements, all N servers must
185 be defined in redis_servers.
187 GF/02 Catch and remove uninitialized value warning in exiqsumm
188 Check for existence of @ARGV before looking at $ARGV[0]
190 JH/01 Replace the store_release() internal interface with store_newblock(),
191 which internalises the check required to safely use the old one, plus
192 the allocate and data copy operations duplicated in both (!) of the
193 extant use locations.
195 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
196 modifier. This matches the restriction on the commandline.
198 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
199 Previously only the last row was returned.
201 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
202 we assumed that tags in the header were well-formed, and parsed the
203 element content after inspecting only the first char of the tag.
204 Assumptions at that stage could crash the receive process on malformed
207 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
208 While running the DKIM ACL we operate on the Permanent memory pool so that
209 variables created with "set" persist to the DATA ACL. Also (at any time)
210 DNS lookups that fail create cache records using the Permanent pool. But
211 expansions release any allocations made on the current pool - so a dnsdb
212 lookup expansion done in the DKIM ACL releases the memory used for the
213 DNS negative-cache, and bad things result. Solution is to switch to the
214 Main pool for expansions.
215 While we're in that code, add checks on the DNS cache during store_reset,
216 active in the testsuite.
217 Problem spotted, and debugging aided, by Wolfgang Breyha.
219 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
220 When none of the hosts presented to a transport match an already-open
221 connection, close it and proceed with the list. Previously we would
222 queue the message. Spotted by Lena with Yahoo, probably involving
225 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
226 Previously a spurious "250 OK id=" response was appended to the proper
229 JH/08 The "support for" informational output now, which built with Content
230 Scanning support, has a line for the malware scanner interfaces compiled
231 in. Interface can be individually included or not at build time.
233 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
234 by the template makefile "src/EDITME". The "STREAM" support for an older
235 ClamAV interface method is removed.
237 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
238 rows affected is given instead).
240 JH/11 The runtime Berkeley DB library version is now additionally output by
241 "exim -d -bV". Previously only the compile-time version was shown.
243 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
244 SMTP connection. Previously, when one had more receipients than the
245 first, an abortive onward connection was made. Move to full support for
246 multiple onward connections in sequence, handling cutthrough connection
247 for all multi-message initiating connections.
249 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
250 routers. Previously, a multi-recipient message would fail to match the
251 onward-connection opened for the first recipient, and cause its closure.
253 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
254 a timeout on read on a GnuTLS initiating connection, resulting in the
255 initiating connection being dropped. This mattered most when the callout
256 was marked defer_ok. Fix to keep the two timeout-detection methods
259 JH/15 Relax results from ACL control request to enable cutthrough, in
260 unsupported situations, from error to silently (except under debug)
261 ignoring. This covers use with PRDR, frozen messages, queue-only and
264 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
266 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
267 metadata, resulting in a crash in free().
269 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
270 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
271 Broken also in d185889f4, with init system revamp.
273 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
274 Previously we abruptly closed the connection after reading a malware-
275 found indication; now we go on to read the "scan ok" response line,
278 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
279 ACL. Previously, a crash would result.
281 JH/19 Speed up macro lookups during configuration file read, by skipping non-
282 macro text after a replacement (previously it was only once per line) and
283 by skipping builtin macros when searching for an uppercase lead character.
285 JH/20 DANE support moved from Experimental to mainline. The Makefile control
286 for the build is renamed.
288 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
289 was allocated for every new TLS startup, meaning one per message. Fix
290 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
292 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
293 reported the original. Fix to report (as far as possible) the ACL
294 result replacing the original.
296 JH/23 Fix memory leak during multi-message connections using STARTTLS under
297 OpenSSL. Certificate information is loaded for every new TLS startup,
298 and the resources needed to be freed.
300 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
302 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
305 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
306 DATA response info to the (existing) per-recipient response info for
307 the "C=" log element. It can have useful tracking info from the
308 destination system. Patch from Simon Arlott.
310 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
311 length value. Previously this would segfault.
313 HS/02 Support Avast multiline protoocol, this allows passing flags to
314 newer versions of the scanner.
316 JH/28 Ensure that variables possibly set during message acceptance are marked
317 dead before release of memory in the daemon loop. This stops complaints
318 about them when the debug_store option is enabled. Discovered specifically
319 for sender_rate_period, but applies to a whole set of variables.
320 Do the same for the queue-runner and queue-list loops, for variables set
321 from spool message files. Do the same for the SMTP per-message loop, for
322 certain variables indirectly set in ACL operations.
324 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
325 as a multi-recipient message from a mailinglist manager). The coding had
326 an arbitrary cutoff number of characters while checking for more input;
327 enforced by writing a NUL into the buffer. This corrupted long / fast
328 input. The problem was exposed more widely when more pipelineing of SMTP
329 responses was introduced, and one Exim system was feeding another.
330 The symptom is log complaints of SMTP syntax error (NUL chars) on the
331 receiving system, and refused recipients seen by the sending system
332 (propating to people being dropped from mailing lists).
333 Discovered and pinpointed by David Carter.
335 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
336 replaced by the ${authresults } expansion.
338 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
340 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
341 allows proper process termination in container environments.
343 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
344 Previously the "final dot" had a newline after it; ensure it is CR,LF.
346 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
347 and "err_perm", deprecated since 4.83 when the RFC-defined words
348 "temperror" and "permerror" were introduced.
350 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
351 transport-filters or DKIM-signing. The restriction was lost in the
352 consolidation of verify-callout and delivery SMTP handling.
353 Extend the restriction to also cover ARC-signing.
355 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
356 in defer=pass mode supply a 450 to the initiator. Previously the message
359 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
360 tls_require_ciphers is used as before.
362 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
363 "pass_unscanned". Only tmpfails from the scanner are written to
364 the paniclog, as they may require admin intervention (permission
365 denied, license issues). Other scanner errors (like decompression
366 bombs) do not cause a paniclog entry.
368 JH/36 Fix reinitialisation of DKIM logging variable between messages.
369 Previously it was possible to log spurious information in receive log
372 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
373 triggered odd behaviour from Outlook Express clients.
375 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
378 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
379 since the IETF WG has not yet settled on that versus the original
380 "bare" representation.
382 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
383 Previously the millisecond value corrupted the output.
384 Fix also for syslog_pid=no and log_selector +pid, for which the pid
385 corrupted the output.
391 JH/01 Rework error string handling in TLS interface so that the caller in
392 more cases is responsible for logging. This permits library-sourced
393 string to be attached to addresses during delivery, and collapses
394 pairs of long lines into single ones.
396 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
397 during configuration. Wildcards are allowed and expanded.
399 JH/02 Rework error string handling in DKIM to pass more info back to callers.
400 This permits better logging.
402 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
403 do not close it down and have the child transport start it up again on
404 the passed-on TCP connection. Instead, proxy the child (and any
405 subsequent ones) for TLS via a unix-domain socket channel. Logging is
406 affected: the continued delivery log lines do not have any DNSSEC, TLS
407 Certificate or OCSP information. TLS cipher information is still logged.
409 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
410 identical IP addresses on different listening ports. Will also affect
413 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
414 add noisy ifdef guards to special-case this sillyness.
415 Patch from Bernd Kuhls.
417 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
418 than 255 are no longer allowed.
420 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
421 Disable the session-cache too, which might reduce our load. Since we
422 currrectly use a new context for every connection, both as server and
423 client, there is no benefit for these.
424 GnuTLS appears to not support tickets server-side by default (we don't
425 call gnutls_session_ticket_enable_server()) but client side is enabled
426 by default on recent versions (3.1.3 +) unless the PFS priority string
429 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
430 <https://reproducible-builds.org/specs/source-date-epoch/>.
432 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
433 the check for any unsuccessful recipients did not notice the limit, and
434 erroneously found still-pending ones.
436 JH/08 Pipeline CHUNKING command and data together, on kernels that support
437 MSG_MORE. Only in-clear (not on TLS connections).
439 JH/09 Avoid using a temporary file during transport using dkim. Unless a
440 transport-filter is involved we can buffer the headers in memory for
441 creating the signature, and read the spool data file once for the
442 signature and again for transmission.
444 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
445 4.77 as the kernel support then wasn't solid, having issues in 64bit
446 mode. Now, it's been long enough. Add support for FreeBSD also.
448 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
449 case where the routing stage had gathered several addresses to send to
450 a host before calling the transport for the first, we previously failed
451 to close down TLS in the old transport process before passing the TCP
452 connection to the new process. The new one sent a STARTTLS command
453 which naturally failed, giving a failed delivery and bloating the retry
454 database. Investigation and fix prototype from Wolfgang Breyha.
456 JH/12 Fix check on SMTP command input synchronisation. Previously there were
457 false-negatives in the check that the sender had not preempted a response
458 or prompt from Exim (running as a server), due to that code's lack of
459 awareness of the SMTP input buffering.
461 PP/04 Add commandline_checks_require_admin option.
462 Exim drops privileges sanely, various checks such as -be aren't a
463 security problem, as long as you trust local users with access to their
464 own account. When invoked by services which pass untrusted data to
465 Exim, this might be an issue. Set this option in main configuration
466 AND make fixes to the calling application, such as using `--` to stop
469 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
470 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
471 responses to those, into a single TLS record each way (this usually means
472 a single packet). As a side issue, smtp_enforce_sync now works on TLS
475 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
476 affects you only if you're dancing at the edge of the param size limits.
477 If you are, and this message makes sense to you, then: raise the
478 configured limit or use OpenSSL 1.1. Nothing we can do for older
481 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
482 cmdline element to get the documented default one. Previously it was
485 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
486 are unsuable for non-message-specific reasons - eg. connection timeout,
487 banner-time rejection.
489 JH/16 Fix logging of delivery remote port, when specified by router, under
492 PP/06 Repair manualroute's ability to take options in any order, even if one
493 is the name of a transport.
496 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
498 JH/17 Change the list-building routines interface to use the expanding-string
499 triplet model, for better allocation and copying behaviour.
501 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
502 Previously it was constructed the first time a possibly-matching string
503 was met in the configuration file input during startup; now it is done
506 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
507 compatible one, to avoid the (poorly documented) possibility of a config
508 file in the working directory redirecting the DB files, possibly correpting
509 some existing file. CVE-2017-10140 assigned for BDB.
511 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
512 cache-hot. Previously, although the result was properly cached, the
513 initial verify call returned a defer.
515 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
516 the main verify for receipient in uncached-mode.
518 JH/22 Retire historical build files to an "unsupported" subdir. These are
519 defined as "ones for which we have no current evidence of testing".
521 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
522 if present. Previously it was ignored.
524 JH/24 Start using specified-initialisers in C structure init coding. This is
525 a C99 feature (it's 2017, so now considered safe).
527 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
528 if was a fixed-sized field and bitmask ops via macros; it is now more
531 PP/07 GitHub PR 56: Apply MariaDB build fix.
532 Patch provided by Jaroslav Škarvada.
534 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
535 during Coverity cleanups [4.87 JH/47]
536 Diagnosis and fix provided by Michael Fischer v. Mollard.
538 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
539 the right size to place the terminating semicolon on its own folded
540 line, the header hash was calculated to an incorrect value thanks to
541 the (relaxed) space the fold became.
543 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
544 and confused the parent.
546 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
547 which could crash as a result. This could lead to undeliverable messages.
549 JH/28 Logging: "next input sent too soon" now shows where input was truncated
552 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
553 matters on fast-turnover and PID-randomising systems, which were getting
554 out-of-order delivery.
556 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
557 a possibly-overlapping copy. The symptom was that "Remote host closed
558 connection in response to HELO" was logged instead of the actual 4xx
561 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
562 Previously only that bufferd was discarded, resulting in SYMTP command
565 JH/32 DKIM: when a message has multiple signatures matching an identity given
566 in dkim_verify_signers, run the dkim acl once for each. Previously only
567 one run was done. Bug 2189.
569 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
570 "panic the current process" to "deliberately defer". The panic log is
571 still written with the problem list name; the mail and reject logs now
572 get a temp-reject line for the message that was being handled, saying
573 something like "domains check lookup or other defer". The SMTP 451
574 message is still "Temporary local problem".
576 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
577 A crafted sequence of BDAT commands could result in in-use memory beeing
578 freed. CVE-2017-16943.
580 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
581 from SMTP input. Previously it was always done; now only done for DATA
582 and not BDAT commands. CVE-2017-16944.
584 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
585 to the message (such as an overlong header line). Previously this was
586 not done and we did not exit BDAT mode. Followon from the previous item
587 though a different problem.
593 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
594 than -2003 did; needs libidn2 in addition to libidn.
596 JH/02 The path option on a pipe transport is now expanded before use.
598 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
599 Patch provided by "Björn", documentation fix added too.
601 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
602 missing a wire-to-host endian conversion.
604 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
605 close after a BDAT command line could be taken as a following command,
606 giving a synch failure. Fix by only checking for synch immediately
607 before acknowledging the chunk.
609 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
610 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
611 macro. Patches provided by Josh Soref.
613 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
614 Previously we did not; the RFC seems ambiguous and VRFY is not listed
615 by IANA as a service extension. However, John Klensin suggests that we
618 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
619 the dkim code may be unix-mode line endings rather than smtp wire-format
620 CRLF, so prepend a CR to any bare LF.
622 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
623 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
625 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
626 the first were themselves being wrongly included in the feed into dkim
627 processing; with most chunk sizes in use this resulted in an incorrect
628 body hash calculated value.
630 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
631 DKIM signature block, for verification. Although advised against by
632 standards it is specifically not ruled illegal.
634 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
636 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
637 missing a body hash (the bh= tag).
639 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
640 It seems that HAProxy sends the Proxy Protocol information in clear and
641 only then does a TLS startup, so do the same.
643 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
644 TCP connections (such as for Spamd) unless the daemon successfully set
645 Fast Open mode on its listening sockets. This fixes breakage seen on
646 too-old kernels or those not configured for Fast Open, at the cost of
647 requiring both directions being enabled for TFO, and TFO never being used
648 by non-daemon-related Exim processes.
650 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
651 endings, at least on the first header line. Try to canonify any that get
652 past that check, despite the cost.
654 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
655 now limited to an arbitrary five deep, while parsing addresses with the
656 strip_excess_angle_brackets option enabled.
658 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
659 instead leave the unprompted TLS handshake in socket buffer for the
660 TLS library to consume.
662 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
664 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
666 JH/16 Drop variables when they go out of scope. Memory management drops a whole
667 region in one operation, for speed, and this leaves assigned pointers
668 dangling. Add checks run only under the testsuite which checks all
669 variables at a store-reset and panics on a dangling pointer; add code
670 explicitly nulling out all the variables discovered. Fixes one known
671 bug: a transport crash, where a dangling pointer for $sending_ip_address
672 originally assigned in a verify callout, is re-used.
674 PP/06 Drop '.' from @INC in various Perl scripts.
676 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
678 PP/08 Reduce a number of compilation warnings under clang; building with
679 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
680 should be warning-free.
682 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
684 HS/01 Fix portability problems introduced by PP/08 for platforms where
685 realloc(NULL) is not equivalent to malloc() [SunOS et al].
687 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
688 chunk. This allows us to accept broken chunked messages. We need a more
689 general solution here.
691 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
692 already-broken messages in the queue.
694 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
696 JH/19 Fix reference counting bug in routing-generated-address tracking.
702 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
703 supports it and a size is available (ie. the sending peer gave us one).
705 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
706 years of being deprecated). The replacements are the ACLs
707 acl_smtp_mime and acl_not_smtp_mime.
709 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
710 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
711 or even an in-clear connection were permitted. Now, if the host lookup
712 was dnssec and dane was requested then the host is only used if the
713 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
714 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
715 if one fails this test.
716 This means that a poorly-configured remote DNS will make it incommunicado;
717 but it protects against a DNS-interception attack on it.
719 JH/04 Bug 1810: make continued-use of an open smtp transport connection
720 non-noisy when a race steals the message being considered.
722 JH/05 If main configuration option tls_certificate is unset, generate a
723 self-signed certificate for inbound TLS connections.
725 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
726 in rewrites and routers.
728 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
729 and logged a warning sing 4.83; now they are a configuration file error.
731 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
732 (lacking @domain). Apply the same qualification processing as RCPT.
734 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
736 JH/10 Support ${sha256:} applied to a string (as well as the previous
739 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
740 a cutthrough deliver is pending, as we always want to make a connection.
741 This also avoids re-routing the message when later placing the cutthrough
742 connection after a verify cache hit.
743 Do not update it with the verify result either.
745 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
746 when routing results in more than one destination address.
748 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
749 signing (which inhibits the cutthrough capability). Previously only
750 the presence of an option was tested; now an expansion evaluating as
751 empty is permissible (obviously it should depend only on data available
752 when the cutthrough connection is made).
754 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
755 the relevant preceding SMTP command did not note the pipelining mode.
757 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
758 Previously they were not counted.
760 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
761 as one having no matching records. Previously we deferred the message
762 that needed the lookup.
764 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
765 distinguished as "(=".
767 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
768 for missing MX records. Previously it only worked for missing A records.
770 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
772 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
773 after the data-go-ahead and data-ack. Patch from Jason Betts.
775 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
776 even for a "none" policy. Patch from Tony Meyer.
778 JH/22 Fix continued use of a connection for further deliveries. If a port was
779 specified by a router, it must also match for the delivery to be
782 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
783 When one of the recipients of a message was unsuitable for the connection
784 (has no matching addresses), we lost track of needing to mark it
785 deferred. As a result mail would be lost.
787 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
789 JH/25 Decoding ACL controls is now done using a binary search; the source code
790 takes up less space and should be simpler to maintain. Merge the ACL
791 condition decode tables also, with similar effect.
793 JH/26 Fix problem with one_time used on a redirect router which returned the
794 parent address unchanged. A retry would see the parent address marked as
795 delivered, so not attempt the (identical) child. As a result mail would
798 JH/27 Fix a possible security hole, wherein a process operating with the Exim
799 UID can gain a root shell. Credit to http://www.halfdog.net/ for
800 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
803 JH/28 Enable {spool,log} filesystem space and inode checks as default.
804 Main config options check_{log,spool}_{inodes,space} are now
805 100 inodes, 10MB unless set otherwise in the configuration.
807 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
808 Previously it only applied to the main-section connection policy
811 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
813 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
814 by me. Added RFC7919 DH primes as an alternative.
816 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
817 are not in the system include path.
819 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
820 GnuTLS, when a session startup failed (eg because the client disconnected)
821 Exim did stdio operations after fclose. This was exposed by a recent
822 change which nulled out the file handle after the fclose.
824 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
825 signed directly by the cert-signing cert, rather than an intermediate
826 OCSP-signing cert. This is the model used by LetsEncrypt.
828 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
830 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
831 an incoming connection.
833 HS/02 Bug 1802: Do not half-close the connection after sending a request
836 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
837 fallback to "prime256v1".
839 JH/34 SECURITY: Use proper copy of DATA command in error message.
840 Could leak key material. Remotely exploitable. CVE-2016-9963.
846 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
847 and 3.4.4 - once the server is enabled to respond to an OCSP request
848 it does even when not requested, resulting in a stapling non-aware
849 client dropping the TLS connection.
851 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
852 support variable-length bit vectors. No functional change.
854 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
855 The I= interface field on outgoing lines is now after the H= remote
856 host field, same as incoming lines. There is a separate
857 outgoing_interface log selector which allows you to disable the
860 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
861 If not running log_selector +smtp_connection the mainlog would be held
862 open indefinitely after a "too many connections" event, including to a
863 deleted file after a log rotate. Leave the per net connection logging
864 leaving it open for efficiency as that will be quickly detected by the
865 check on the next write.
867 HS/01 Bug 1671: Fix post transport crash.
868 Processing the wait-<transport> messages could crash the delivery
869 process if the message IDs didn't exist for some reason. When
870 using 'split_spool_directory=yes' the construction of the spool
871 file name failed already, exposing the same netto behaviour.
873 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
874 mime_regex ACL conditions.
876 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
877 to DSN fail messages (bounces): remote IP, remote greeting, remote response
878 to HELO, local diagnostic string.
880 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
881 log line to debug. Even when configured with a tls authenticator many
882 client connections are expected to not authenticate in this way, so
883 an authenticate fail is not an error.
885 HS/02 Add the Exim version string to the process info. This way exiwhat
886 gives some more detail about the running daemon.
888 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
889 matter for fast-change records such as DNSBLs.
891 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
892 retry record, even if constant. There may be multiple transports with
893 different interface settings and the retry behaviour needs to be kept
896 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
898 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
900 JH/10 Bug 840: fix log_defer_output option of pipe transport
902 JH/11 Bug 830: use same host for all RCPTS of a message, even under
903 hosts_randomize. This matters a lot when combined with mua_wrapper.
905 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
906 ${quote_pgsql:<string>} operator.
908 JH/13 Bug 1708: avoid misaligned access in cached lookup.
910 JH/14 Change header file name for freeradius-client. Relevant if compiling
911 with Radius support; from the Gentoo tree and checked under Fedora.
913 JH/15 Bug 1712: Introduce $prdr_requested flag variable
915 JH/16 Bug 1714: Permit an empty string as expansion result for transport
916 option transport_filter, meaning no filtering.
918 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
920 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
921 defaults to "*" (all hosts). The variable is now available when not built
922 with TLS, default unset, mainly to enable keeping the testsuite sane.
923 If a server certificate is not supplied (via tls_certificate) an error is
924 logged, and clients will find TLS connections fail on startup. Presumably
925 they will retry in-clear.
926 Packagers of Exim are strongly encouraged to create a server certificate
927 at installation time.
929 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
930 with the $config_file variable.
932 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
933 in transport context, after the attempt, and per-recipient. The latter type
934 is per host attempted. The event data is the error message, and the errno
935 information encodes the lookup type (A vs. MX) used for the (first) host,
936 and the trailing two digits of the smtp 4xx response.
938 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
939 to write to mainlog (or rejectlog, paniclog) in the window between file
940 creation and permissions/ownership being changed. Particularly affects
941 installations where exicyclog is run as root, rather than exim user;
942 result is that the running daemon panics and dies.
944 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
946 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
947 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
948 "pri" and "weight". Note that the previous implicit priority given by the
949 list order is no longer honoured.
951 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
954 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
955 by defining SUPPORT_SOCKS.
957 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
958 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
959 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
960 variables are renamed to proxy_{local,external}_{address,port}.
962 JH/27 Move Internationalisation support from Experimental to mainline, enabled
963 for a build by defining SUPPORT_I18N
965 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
966 of the query string, and make ${quote_redis:} do that quoting.
968 JH/29 Move Events support from Experimental to mainline, enabled by default
969 and removable for a build by defining DISABLE_EVENT.
971 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
973 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
974 cached by the daemon.
976 JH/32 Move Redis support from Experimental to mainline, enabled for a build
977 by defining LOOKUP_REDIS. The libhiredis library is required.
979 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
980 keys are given for lookup.
982 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
983 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
984 only supported when built with TLS support. The PolarSSL SHA routines
985 are still used when the TLS library is too old for convenient support.
987 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
988 openssl_options), for security. OpenSSL forces this from version 1.1.0
989 server-side so match that on older versions.
991 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
992 allocation for $value could be released as the expansion processing
993 concluded, but leaving the global pointer active for it.
995 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
996 and to use the domains and local_parts ACL conditions.
998 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
999 incorrectly not doubled on cutthrough transmission, hence seen as a
1000 body-termination at the receiving system - resulting in truncated mails.
1001 Commonly the sender saw a TCP-level error, and retransmitted the message
1002 via the normal store-and-forward channel. This could result in duplicates
1003 received - but deduplicating mailstores were liable to retain only the
1004 initial truncated version.
1006 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1008 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1010 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1011 we're in there, support oversigning also; bug 1309.
1013 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1015 HS/04 Add support for keep_environment and add_environment options.
1017 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1018 either intentional arithmetic overflow during PRNG, or testing config-
1021 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1022 delivery resulted in actual delivery. Cancel cutthrough before DATA
1025 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1026 rejects a recipient: pass the reject to the originator.
1028 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1029 Many were false-positives and ignorable, but it's worth fixing the
1032 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1033 for the new environment-manipulation done at startup. Move the routines
1034 from being local to tls.c to being global via the os.c file.
1036 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1037 an extract embedded as result-arg for a map, the first arg for extract
1038 is unavailable so we cannot tell if this is a numbered or keyed
1039 extraction. Accept either.
1045 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1048 JH/02 The smtp transport option "multi_domain" is now expanded.
1050 JH/03 The smtp transport now requests PRDR by default, if the server offers
1053 JH/04 Certificate name checking on server certificates, when exim is a client,
1054 is now done by default. The transport option tls_verify_cert_hostnames
1055 can be used to disable this per-host. The build option
1056 EXPERIMENTAL_CERTNAMES is withdrawn.
1058 JH/05 The value of the tls_verify_certificates smtp transport and main options
1059 default to the word "system" to access the system default CA bundle.
1060 For GnuTLS, only version 3.0.20 or later.
1062 JH/06 Verification of the server certificate for a TLS connection is now tried
1063 (but not required) by default. The verification status is now logged by
1064 default, for both outbound TLS and client-certificate supplying inbound
1067 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1070 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1071 Status Notification (bounce) messages are now MIME format per RFC 3464.
1072 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1073 under the control of the dsn_advertise_hosts option, and routers may
1074 have a dsn_lasthop option.
1076 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1077 default, modifiable by a malware= option. The list separator for
1078 the options can now be changed in the usual way. Bug 68.
1080 JH/10 The smtp_receive_timeout main option is now expanded before use.
1082 JH/11 The incoming_interface log option now also enables logging of the
1083 local interface on delivery outgoing connections.
1085 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1086 if the interface and destination host and port all match.
1088 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1091 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1092 Patch from Andrew Lewis.
1094 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1095 now supports optional time-restrictions, weighting, and priority
1096 modifiers per server. Patch originally by <rommer@active.by>.
1098 JH/16 The spamd_address main option now supports a mixed list of local
1099 and remote servers. Remote servers can be IPv6 addresses, and
1100 specify a port-range.
1102 JH/17 Bug 68: The spamd_address main option now supports an optional
1103 timeout value per server.
1105 JH/18 Bug 1581: Router and transport options headers_add/remove can
1106 now have the list separator specified.
1108 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1111 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1114 JH/21 Support for the A6 type of dns record is withdrawn.
1116 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1117 rather than the verbs used.
1119 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1120 from 255 to 1024 chars.
1122 JH/24 Verification callouts now attempt to use TLS by default.
1124 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1125 are generic router options now. The defaults didn't change.
1127 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1128 Original patch from Alexander Shikoff, worked over by JH.
1130 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1131 files if ZCAT_COMMAND is not executable.
1133 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1135 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1137 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1138 Normally benign, it bites when the pair was led to by a CNAME;
1139 modern usage is to not canonicalize the domain to a CNAME target
1140 (and we were inconsistent anyway for A-only vs AAAA+A).
1142 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1144 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1145 when evaluating $sender_host_dnssec.
1147 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1148 $sender_helo_dnssec variable.
1150 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1152 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1154 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1156 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1157 documented as working, but never had. Support all but $spam_report.
1159 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1160 added for tls authenticator.
1162 HS/03 Add perl_taintmode main config option
1168 TL/01 When running the test suite, the README says that variables such as
1169 no_msglog_check are global and can be placed anywhere in a specific
1170 test's script, however it was observed that placement needed to be near
1171 the beginning for it to behave that way. Changed the runtest perl
1172 script to read through the entire script once to detect and set these
1173 variables, reset to the beginning of the script, and then run through
1174 the script parsing/test process like normal.
1176 TL/02 The BSD's have an arc4random API. One of the functions to induce
1177 adding randomness was arc4random_stir(), but it has been removed in
1178 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1179 function when detected.
1181 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1182 cause callback expansion.
1184 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1185 syntax errors in an expansion can be treated as a string instead of
1186 logging or causing an error, due to the internal use of bool_lax
1187 instead of bool when processing it.
1189 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1190 server certificates when making smtp deliveries.
1192 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1194 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1196 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1198 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1199 Merged patch from Sebastian Wiedenroth.
1201 JH/05 Fix results-pipe from transport process. Several recipients, combined
1202 with certificate use, exposed issues where response data items split
1203 over buffer boundaries were not parsed properly. This eventually
1204 resulted in duplicates being sent. This issue only became common enough
1205 to notice due to the introduction of connection certificate information,
1206 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1208 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1209 size buffer was used, resulting in syntax errors when an expansion
1212 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1213 version 3.3.6 or later.
1215 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1216 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1217 both become "event_action", the variables become $event_name, $event_data
1218 and $event_defer_errno. There is a new variable $verify_mode, usable in
1219 routers, transports and related events. The tls:cert event is now also
1220 raised for inbound connections, if the main configuration event_action
1223 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1224 early OCSP support, but no stapling (appears to be less than 1.0.0).
1226 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1227 server certificate names available under the smtp transport option
1228 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1231 JH/10 Time-related extraction expansions from certificates now use the main
1232 option "timezone" setting for output formatting, and are consistent
1233 between OpenSSL and GnuTLS compilations. Bug 1541.
1235 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1236 encoded parameter in the incoming message. Bug 1558.
1238 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1239 include certificate info, eximon was claiming there were spoolfile
1242 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1244 JH/14 Log delivery-related information more consistently, using the sequence
1245 "H=<name> [<ip>]" wherever possible.
1247 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1248 are problematic for Debian distribution, omit them from the release
1251 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1253 JH/16 Fix string representation of time values on 64bit time_t architectures.
1256 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1257 output list separator was used.
1262 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1263 checkers that were complaining about end of non-void function with no
1266 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1267 This was a regression introduced in 4.83 by another bugfix.
1269 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1271 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1272 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1278 TF/01 Correctly close the server side of TLS when forking for delivery.
1280 When a message was received over SMTP with TLS, Exim failed to clear up
1281 the incoming connection properly after forking off the child process to
1282 deliver the message. In some situations the subsequent outgoing
1283 delivery connection happened to have the same fd number as the incoming
1284 connection previously had. Exim would try to use TLS and fail, logging
1285 a "Bad file descriptor" error.
1287 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1288 utilities have not been installed.
1290 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1291 temporary space as the ACL may create new global variables.
1293 TL/01 LDAP support uses per connection or global context settings, depending
1294 upon the detected version of the libraries at build time.
1296 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1297 to extract and use the src ip:port in logging and expansions as if it
1298 were a direct connection from the outside internet. PPv2 support was
1299 updated based on HAProxy spec change in May 2014.
1301 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1303 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1304 Properly escape header and check for NULL return.
1306 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1309 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1311 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1312 characters in header names, implemented as a verify condition.
1313 Contributed by Michael Fischer v. Mollard.
1315 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1316 results permerror and temperror. Previous values are deprecated but
1317 still accepted. In a future release, err_perm and err_temp will be
1318 completely removed, which will be a backward incompatibility if the
1319 ACL tests for either of these two old results. Patch contributed by
1320 user bes-internal on the mailing list.
1322 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1324 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1325 selectors, in both main and reject logs.
1327 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1330 JH/07 Add malware type "sock" for talking to simple daemon.
1332 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1334 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1335 routers/transports under cutthrough routing.
1337 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1338 numbers. Touch up "bool" conditional to keep the same definition.
1340 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1342 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1344 JH/12 Expand items in router/transport headers_add or headers_remove lists
1345 individually rather than the list as a whole. Bug 1452.
1347 Required for reasonable handling of multiple headers_ options when
1348 they may be empty; requires that headers_remove items with embedded
1349 colons must have them doubled (or the list-separator changed).
1351 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1352 view the policy declared in the DMARC record. Currently, $dmarc_status
1353 is a combined value of both the record presence and the result of the
1356 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1358 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1359 dnslookup router and the smtp transport (applying to the forward
1362 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1363 of ldap servers used for a specific lookup. Patch provided by Heiko
1366 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1367 New variable $lookup_dnssec_authenticated for observability.
1369 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1370 Patch submitted by Lars Timman.
1372 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1374 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1375 Requires trusted mode and valid format message id, aborts otherwise.
1376 Patch contributed by Heiko Schlichting.
1378 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1379 certextract with support for various fields. Bug 1358.
1381 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1382 is requested by default, modifiable by smtp transport option
1385 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1386 operate on certificate variables to give certificate fingerprints
1387 Also new ${sha256:cert_variable}.
1389 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1391 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1392 Christian Aistleitner.
1394 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1396 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1397 file. Patch from Wolfgang Breyha.
1399 JH/25 Expand the coverage of the delivery $host and $host_address to
1400 client authenticators run in verify callout. Bug 1476.
1402 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1403 align with daemon_smtp_ports. Bug 72.
1405 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1406 support and error reporting did not work properly.
1408 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1409 and is readable. Patch from Andrew Colin Kissa.
1411 TL/14 Enhance documentation of ${run expansion and how it parses the
1412 commandline after expansion, particularly in the case when an
1413 unquoted variable expansion results in an empty value.
1415 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1417 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1418 by renaming away from it. Addresses GH issue 15, reported by
1421 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1423 TL/15 SECURITY: prevent double expansion in math comparison functions
1424 (can expand unsanitized data). Not remotely exploitable.
1431 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1433 PP/02 Make -n do something, by making it not do something.
1434 When combined with -bP, the name of an option is not output.
1436 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1439 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1440 $sender_host_name and config options to manage this, and basic check
1443 PP/05 DSCP support for outbound connections and control modifier for inbound.
1445 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1446 (Only plugin which currently uses this is kerberos4, which nobody should
1447 be using, but we should make it available and other future plugins might
1448 conceivably use it, even though it would break NAT; stuff *should* be
1449 using channel bindings instead).
1451 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1452 name; added for Sendmail compatibility; requires admin caller.
1453 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1454 just ignore it); requires trusted caller.
1455 Also parse but ignore: -Ac -Am -X<logfile>
1458 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1460 TL/02 Add +smtp_confirmation as a default logging option.
1462 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1463 Patch by Magnus Holmgren from 2007-02-20.
1465 TL/04 Bugzilla 1281 - Spec typo.
1466 Bugzilla 1283 - Spec typo.
1467 Bugzilla 1290 - Spec grammar fixes.
1469 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1471 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1473 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1474 mailing list by by Dmitry Isaikin.
1476 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1478 JH/02 Support "G" suffix to numbers in ${if comparisons.
1480 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1482 NM/01 Bugzilla 1197 - Spec typo
1483 Bugzilla 1196 - Spec examples corrections
1485 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1487 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1488 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1491 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1492 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1494 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1495 "acl {{name}{arg}...}", and optional args on acl condition
1498 JH/05 Permit multiple router/transport headers_add/remove lines.
1500 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1502 JH/07 Avoid using a waiting database for a single-message-only transport.
1503 Performance patch from Paul Fisher. Bugzilla 1262.
1505 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1508 JH/09 Add $headers_added variable, with content from use of ACL modifier
1509 add_header (but not yet added to the message). Bugzilla 199.
1511 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1512 Pulled from Bugzilla 817 by Wolfgang Breyha.
1514 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1516 (nb: this is the same fix as in Exim 4.80.1)
1518 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1521 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1524 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1526 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1527 advertise SMTP AUTH mechanism to us, instead of a generic
1528 protocol violation error. Also, make Exim more robust to bad
1529 data from the Dovecot auth socket.
1531 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1533 When a queue runner is handling a message, Exim first routes the
1534 recipient addresses, during which it prunes them based on the retry
1535 hints database. After that it attempts to deliver the message to
1536 any remaining recipients. It then updates the hints database using
1539 So if a recipient address works intermittently, it can get repeatedly
1540 deferred at routing time. The retry hints record remains fresh so the
1541 address never reaches the final cutoff time.
1543 This is a fairly common occurrence when a user is bumping up against
1544 their storage quota. Exim had some logic in its local delivery code
1545 to deal with this. However it did not apply to per-recipient defers
1546 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1548 This change adds a proper retry rule check during routing so that the
1549 final cutoff time is checked against the message's age. We only do
1550 this check if there is an address retry record and there is not a
1551 domain retry record; this implies that previous attempts to handle
1552 the address had the retry_use_local_parts option turned on. We use
1553 this as an approximation for the destination being like a local
1554 delivery, as in LMTP.
1556 I suspect this new check makes the old local delivery cutoff check
1557 redundant, but I have not verified this so I left the code in place.
1559 TF/02 Correct gecos expansion when From: is a prefix of the username.
1561 Test 0254 submits a message to Exim with the header
1565 When I ran the test suite under the user fanf2, Exim expanded
1566 the header to contain my full name, whereas it should have added
1567 a Resent-Sender: header. It erroneously treats any prefix of the
1568 username as equal to the username.
1570 This change corrects that bug.
1572 GF/01 DCC debug and logging tidyup
1573 Error conditions log to paniclog rather than rejectlog.
1574 Debug lines prefixed by "DCC: " to remove any ambiguity.
1576 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1578 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1579 Bug spotted by Jeremy Harris; was flawed since initial commit.
1580 Would have resulted in OCSP responses post-SNI triggering an Exim
1581 NULL dereference and crash.
1583 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1585 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1586 Bug detection, analysis and fix by Samuel Thibault.
1587 Bugzilla 1331, Debian bug #698092.
1589 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1591 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1592 Server implementation by Todd Lyons, client by JH.
1593 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1594 config variable "prdr_enable" controls whether the server
1595 advertises the facility. If the client requests PRDR a new
1596 acl_data_smtp_prdr ACL is called once for each recipient, after
1597 the body content is received and before the acl_smtp_data ACL.
1598 The client is controlled by both of: a hosts_try_prdr option
1599 on the smtp transport, and the server advertisement.
1600 Default client logging of deliveries and rejections involving
1601 PRDR are flagged with the string "PRDR".
1603 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1604 fclose(). Diagnosis by Todd Lyons.
1606 PP/17 Update configure.default to handle IPv6 localhost better.
1607 Patch by Alain Williams (plus minor tweaks).
1610 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1611 This is now consistent with GnuTLS, and is now documented: the
1612 previous undocumented portable approach to treating the option as
1613 unset was to force an expansion failure. That still works, and
1614 an empty string is now equivalent.
1616 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1617 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1618 not performing validation itself.
1620 PP/20 Added force_command boolean option to pipe transport.
1621 Patch from Nick Koston, of cPanel Inc.
1623 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1626 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1628 PP/21 Fix eximon continuous updating with timestamped log-files.
1629 Broken in a format-string cleanup in 4.80, missed when I repaired the
1630 other false fix of the same issue.
1631 Report and fix from Heiko Schlichting.
1634 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1635 Report from Prashanth Katuri.
1637 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1638 It's SecureTransport, so affects any MacOS clients which use the
1639 system-integrated TLS libraries, including email clients.
1641 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1642 using a MIME ACL for non-SMTP local injection.
1643 Report and assistance in diagnosis by Warren Baker.
1645 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1647 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1649 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1650 last id that failed so it may be referenced in subsequent ACL's.
1652 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1655 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1656 ldap library initialization, allowing self-signed CA's to be
1657 used. Also properly sets require_cert option later in code by
1658 using NULL (global ldap config) instead of ldap handle (per
1659 session). Bug diagnosis and testing by alxgomz.
1661 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1662 the src/util/ subdirectory.
1664 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1665 renamed to Transport Post Delivery Action by Jeremy Harris, as
1668 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1669 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1670 redis_servers = needs to be configured which will be used by the redis
1671 lookup. Patch from Warren Baker, of The Packet Hub.
1673 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1675 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1676 hostname or reverse DNS when processing a host list. Used suggestions
1677 from multiple comments on this bug.
1679 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1681 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1682 Missed a few lines, added it to make the runtest require no keyboard
1685 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1686 contains upper case chars. Make router use caseful_local_part.
1688 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1689 support when GnuTLS has been built with p11-kit.
1695 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1697 This, or similar/improved, will also be change PP/11 of 4.82.
1703 PP/01 Handle short writes when writing local log-files.
1704 In practice, only affects FreeBSD (8 onwards).
1705 Bugzilla 1053, with thanks to Dmitry Isaikin.
1707 NM/01 Bugzilla 949 - Documentation tweak
1709 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1712 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1714 PP/02 Implemented gsasl authenticator.
1716 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1718 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1719 `pkg-config foo` for cflags/libs.
1721 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1722 with rest of GSASL and with heimdal_gssapi.
1724 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1725 `pkg-config foo` for cflags/libs for the TLS implementation.
1727 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1728 properties get this fed in as external SSF. A number of robustness
1729 and debugging improvements to the cyrus_sasl authenticator.
1731 PP/08 cyrus_sasl server now expands the server_realm option.
1733 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1734 Patch by Jeremy Harris.
1736 PP/10 Added dbmjz lookup type.
1738 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1740 PP/12 MAIL args handles TAB as well as SP, for better interop with
1741 non-compliant senders.
1742 Analysis and variant patch by Todd Lyons.
1744 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1745 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1746 Patch from Dirk Mueller <dmueller@suse.com>
1748 PP/13 tls_peerdn now print-escaped for spool files.
1749 Observed some $tls_peerdn in wild which contained \n, which resulted
1750 in spool file corruption.
1752 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1753 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1754 or write after TLS renegotiation, which otherwise led to messages
1757 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1758 as a tracking header (ie: a signed header comes before the signature).
1759 Patch from Wolfgang Breyha.
1761 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1762 comma-sep list; embedded commas doubled.
1764 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1766 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1768 Report and patch from Dmitry Banschikov.
1770 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1771 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1772 failures. We appear to now support TLS1.1+ with Exim.
1774 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1775 lets Exim select keys and certificates based upon TLS SNI from client.
1776 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1777 before an outbound SMTP session. New log_selector, +tls_sni.
1779 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1780 NULL dereference. Report and patch from Alun Jones.
1782 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1783 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1784 Not seeing resolver debug output on NetBSD, but suspect this is a
1785 resolver implementation change.
1787 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1788 Left warnings. Added "eximon gdb" invocation mode.
1790 PP/21 Defaulting "accept_8bitmime" to true, not false.
1792 PP/22 Added -bw for inetd wait mode support.
1794 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1795 locate the relevant includes and libraries. Made this the default.
1797 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1798 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1800 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1801 This may cause build issues on older platforms.
1803 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1804 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1805 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1806 Added SNI support via GnuTLS too.
1807 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1809 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1811 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1812 Applied second patch from Janne, implementing suggestion to default
1813 multiple-strings-in-record handling to match SPF spec.
1815 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1817 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1818 read-only, out of scope).
1819 Patch from Wolfgang Breyha, report from Stuart Northfield.
1821 PP/29 Fix three issues highlighted by clang analyser static analysis.
1822 Only crash-plausible issue would require the Cambridge-specific
1823 iplookup router and a misconfiguration.
1824 Report from Marcin Mirosław.
1826 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1828 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1829 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1830 As part of this, removing so much warning spew let me fix some minor
1831 real issues in debug logging.
1833 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1834 assignment on my part. Fixed.
1836 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1837 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1838 Janne Snabb (who went above and beyond: thank you).
1840 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1841 string otherwise requires a connection and a bunch more work and it's
1842 relatively easy to get wrong. Should also expose TLS library linkage
1845 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1846 64-bit ${eval} (JH/03).
1848 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1849 GNU libc to support some of the 64-bit stuff, should not lead to
1850 conflicts. Defined before os.h is pulled in, so if a given platform
1851 needs to override this, it can.
1853 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1854 protection layer was required, which is not implemented.
1855 Bugzilla 1254, patch from Wolfgang Breyha.
1857 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1858 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1859 tls_dhparam take prime identifiers. Also unbreak combination of
1860 OpenSSL+DH_params+TLSSNI.
1862 PP/39 Disable SSLv2 by default in OpenSSL support.
1868 PP/01 Solaris build fix for Oracle's LDAP libraries.
1869 Bugzilla 1109, patch from Stephen Usher.
1871 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1873 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1876 TF/02 Fix a couple more cases where we did not log the error message
1877 when unlink() failed. See also change 4.74-TF/03.
1879 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1880 lock up or crash if it happened to be inside a call to libc when it
1881 got a SIGUSR1 from exiwhat.
1883 The SIGUSR1 handler appends the current process status to the process
1884 log which is later printed by exiwhat. It used to use the general
1885 purpose logging code to do this, but several functions it calls are
1886 not safe for signals.
1888 The new output code in the SIGUSR1 handler is specific to the process
1889 log, and simple enough that it's easy to inspect for signal safety.
1890 Removing some special cases also simplifies the general logging code.
1891 Removing the spurious timestamps from the process log simplifies
1894 TF/04 Improved ratelimit ACL condition.
1896 The /noupdate option has been deprecated in favour of /readonly which
1897 has clearer semantics. The /leaky, /strict, and /readonly update modes
1898 are mutually exclusive. The update mode is no longer included in the
1899 database key; it just determines when the database is updated. (This
1900 means that when you upgrade Exim will forget old rate measurements.)
1902 Exim now checks that the per_* options are used with an update mode that
1903 makes sense for the current ACL. For example, when Exim is processing a
1904 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1905 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1906 must specify per_mail/readonly. If you omit the update mode it defaults to
1907 /leaky where that makes sense (as before) or /readonly where required.
1909 The /noupdate option is now undocumented but still supported for
1910 backwards compatibility. It is equivalent to /readonly except that in
1911 ACLs where /readonly is required you may specify /leaky/noupdate or
1912 /strict/noupdate which are treated the same as /readonly.
1914 A useful new feature is the /count= option. This is a generalization
1915 of the per_byte option, so that you can measure the throughput of other
1916 aggregate values. For example, the per_byte option is now equivalent
1917 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1919 The per_rcpt option has been generalized using the /count= mechanism
1920 (though it's more complicated than the per_byte equivalence). When it is
1921 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1922 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1923 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1924 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1925 Note that using per_rcpt with a non-readonly update mode in more than
1926 one ACL will cause the recipients to be double-counted. (The per_mail
1927 and per_byte options don't have this problem.)
1929 The handling of very low rates has changed slightly. If the computed rate
1930 is less than the event's count (usually one) then this event is the first
1931 after a long gap. In this case the rate is set to the same as this event's
1932 count, so that the first message of a spam run is counted properly.
1934 The major new feature is a mechanism for counting the rate of unique
1935 events. The new per_addr option counts the number of different
1936 recipients that someone has sent messages to in the last time period. It
1937 behaves like per_rcpt if all the recipient addresses are different, but
1938 duplicate recipient addresses do not increase the measured rate. Like
1939 the /count= option this is a general mechanism, so the per_addr option
1940 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1941 example, measure the rate that a client uses different sender addresses
1942 with the options per_mail/unique=$sender_address. There are further
1943 details in the main documentation.
1945 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1947 TF/06 Removed a few PCRE remnants.
1949 TF/07 Automatically extract Exim's version number from tags in the git
1950 repository when doing development or release builds.
1952 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1953 Bugzilla 879. Patch from Paul Fisher.
1955 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1956 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1959 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1961 PP/05 Variable $av_failed, true if the AV scanner deferred.
1962 Bugzilla 1078. Patch from John Horne.
1964 PP/06 Stop make process more reliably on build failure.
1965 Bugzilla 1087. Patch from Heiko Schlittermann.
1967 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1968 Bugzilla 1089. Patch from Heiko Schlittermann.
1970 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1971 Bugzilla 1131. Patch from Holger Weiß.
1973 PP/09 Handle IPv6 addresses with SPF.
1974 Bugzilla 860. Patch from Wolfgang Breyha.
1976 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1978 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1981 PP/11 match_* no longer expand right-hand-side by default.
1982 New compile-time build option, EXPAND_LISTMATCH_RHS.
1983 New expansion conditions, "inlist", "inlisti".
1985 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1987 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1989 PP/14 fix log_write() format string regression from TF/03.
1990 Bugzilla 1152. Patch from Dmitry Isaikin.
1996 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1998 PP/02 Harmonised TLS library version reporting; only show if debugging.
1999 Layout now matches that introduced for other libraries in 4.74 PP/03.
2001 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2003 PP/04 New "dns_use_edns0" global option.
2005 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2008 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2009 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2011 TK/01 Updated PolarSSL code to 0.14.2.
2012 Bugzilla 1097. Patch from Andreas Metzler.
2014 PP/07 Catch divide-by-zero in ${eval:...}.
2015 Fixes bugzilla 1102.
2017 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2020 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2021 format-string attack -- SECURITY: remote arbitrary code execution.
2023 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2024 time unintentionally subject to list matching rules, letting the header
2025 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2026 arbitrary string expansion). This allowed for information disclosure.
2028 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2029 INT_MIN/-1 -- value coerced to INT_MAX.
2035 NM/01 Workaround for PCRE version dependency in version reporting
2038 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2039 This fixes portability to compilers other than gcc, notably
2040 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2042 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2043 makefiles for portability to HP-UX and POSIX correctness.
2045 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2046 Also via indented variable definition in the Makefile.
2047 (Debugging by Oliver Heesakkers).
2049 PP/02 Restore caching of spamd results with expanded spamd_address.
2050 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2052 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2053 Improves build reliability. Fix from: Frank Elsner
2055 NM/02 Fix wide character breakage in the rfc2047 coding
2056 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2058 NM/03 Allow underscore in dnslist lookups
2059 Fixes bug 1026. Patch from Graeme Fowler
2061 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2062 Code patches from Adam Ciarcinski of NetBSD.
2064 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2067 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2068 is logged, to avoid truncation. Patch from John Horne.
2070 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2071 Patch from Jakob Hirsch.
2073 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2074 SQL string expansion failure details.
2075 Patch from Andrey Oktyabrski.
2077 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2078 Patch from Simon Arlott.
2080 PP/09 New lookups functionality failed to compile on old gcc which rejects
2081 extern declarations in function scope.
2082 Patch from Oliver Fleischmann
2084 PP/10 Use sig_atomic_t for flags set from signal handlers.
2085 Check getgroups() return and improve debugging.
2086 Fixed developed for diagnosis in bug 927 (which turned out to be
2089 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2090 Patch from Mark Zealey.
2092 PP/12 Bugzilla 1056: Improved spamd server selection.
2093 Patch from Mark Zealey.
2095 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2096 Based on patch from Heiko Schlittermann.
2098 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2099 Patch from Uwe Doering, sign-off by Michael Haardt.
2101 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2102 variable declaration deep within a block. Bug and patch from
2105 PP/15 lookups-Makefile IRIX compatibility coercion.
2107 PP/16 Make DISABLE_DKIM build knob functional.
2109 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2110 Patch by Simon Arlott
2112 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2113 variable argument macros. Our copy now differs from upstream.
2119 TF/01 Failure to get a lock on a hints database can have serious
2120 consequences so log it to the panic log.
2122 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2123 controlled using the smtp_confirmation log selector.
2125 TF/03 Include the error message when we fail to unlink a spool file.
2127 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2128 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2129 for maintaining out-of-tree patches for some time.
2131 PP/01 Bugzilla 139: Documentation and portability issues.
2132 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2133 Handle per-OS dynamic-module compilation flags.
2135 PP/02 Let /dev/null have normal permissions.
2136 The 4.73 fixes were a little too stringent and complained about the
2137 permissions on /dev/null. Exempt it from some checks.
2138 Reported by Andreas M. Kirchwitz.
2140 PP/03 Report version information for many libraries, including
2141 Exim version information for dynamically loaded libraries. Created
2142 version.h, now support a version extension string for distributors
2143 who patch heavily. Dynamic module ABI change.
2145 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2146 privilege escalation vulnerability whereby the Exim run-time user
2147 can cause root to append content of the attacker's choosing to
2150 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2153 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2154 If dropping privileges for untrusted macros, we disabled normal logging
2155 on the basis that it would fail; for the Exim run-time user, this is not
2156 the case, and it resulted in successful deliveries going unlogged.
2157 Fixed. Reported by Andreas Metzler.
2163 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2164 only prepend for the Resent-* case. Fixes regression introduced in
2165 Exim 4.70 by NM/22 for Bugzilla 607.
2167 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2168 increasing numbers of administrators be bitten by this.
2170 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2172 PP/03 Bugzilla 994: added openssl_options main configuration option.
2174 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2176 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2178 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2180 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2181 without a peer certificate, leading to a segfault because of an
2182 assumption that peers always have certificates. Be a little more
2183 paranoid. Problem reported by Martin Tscholak.
2185 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2186 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2187 NB: ClamAV planning to remove STREAM in "middle of 2010".
2188 CL also introduces -bmalware, various -d+acl logging additions and
2189 more caution in buffer sizes.
2191 PP/09 Implemented reverse_ip expansion operator.
2193 PP/10 Bugzilla 937: provide a "debug" ACL control.
2195 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2197 PP/12 Bugzilla 973: Implement --version.
2199 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2201 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2203 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2205 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2206 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2207 ignore trailing whitespace.
2209 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2211 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2214 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2215 Notification from Dr Andrew Aitchison.
2217 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2218 ExtendedDetectionInfo response format.
2219 Notification from John Horne.
2221 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2224 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2225 XSL and documented dependency on system catalogs, with examples of how
2228 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2231 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2232 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2233 configuration file which is writeable by the Exim user or group.
2235 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2236 of configuration files to cover files specified with the -C option if
2237 they are going to be used with root privileges, not just the default
2240 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2241 option (effectively making it always true).
2243 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2244 files to be used while preserving root privileges.
2246 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2247 that rogue child processes cannot use them.
2249 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2250 run-time user, instead of root.
2252 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2253 Exim run-time user without dropping privileges.
2255 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2256 result string, instead of calling string_vformat() twice with the same
2259 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2260 for other users. Others should always drop root privileges if they use
2261 -C on the command line, even for a whitelisted configure file.
2263 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2265 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2271 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2272 $data_path, and $header_path variables; fixed documentation bugs and
2275 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2276 exipick to access non-standard spools, including the "frozen" queue
2279 NM/01 Bugzilla 965: Support mysql stored procedures.
2280 Patch from Alain Williams
2282 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2284 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2285 Patch from Andreas Metzler
2287 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2288 Patch from Kirill Miazine
2290 NM/05 Bugzilla 671: Added umask to procmail example.
2292 JJ/03 installed exipick 20100323.0, fixing doc bug
2294 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2295 directory. Notification and patch from Dan Rosenberg.
2297 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2299 TK/02 Improve log output when DKIM signing operation fails.
2301 MH/01 Treat the transport option dkim_domain as a colon separated
2302 list, not as a single string, and sign the message with each element,
2303 omitting multiple occurences of the same signer.
2305 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2306 Bugzilla 985, 986. Patch by Simon Arlott
2308 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2309 Patch by Simon Arlott
2311 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2312 MBX locking. Notification from Dan Rosenberg.
2318 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2320 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2322 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2324 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2326 NM/04 Bugzilla 915: Flag broken perl installation during build.
2332 TK/01 Added patch by Johannes Berg that expands the main option
2333 "spamd_address" if it starts with a dollar sign.
2335 TK/02 Write list of recipients to X-Envelope-Sender header when building
2336 the mbox-format spool file for content scanning (suggested by Jakob
2339 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2340 (http://www.dcc-servers.net/) support via dccifd. Activated by
2341 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2343 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2344 by Mark Daniel Reidel <mr@df.eu>.
2346 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2347 When building exim an external PCRE library is now needed -
2348 PCRE is a system library on the majority of modern systems.
2349 See entry on PCRE_LIBS in EDITME file.
2351 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2352 conversation. Added nologin parameter to request.
2353 Patch contributed by Kirill Miazine.
2355 TF/01 Do not log submission mode rewrites if they do not change the address.
2357 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2359 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2360 log files in place. Contributed by Roberto Lima.
2362 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2364 TF/03 Bugzilla 615: When checking the local_parts router precondition
2365 after a local_part_suffix or local_part_prefix option, Exim now
2366 does not use the address's named list lookup cache, since this
2367 contains cached lookups for the whole local part.
2369 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2370 Robert Millan. Documentation is in experimental-spec.txt.
2372 TF/04 Bugzilla 668: Fix parallel build (make -j).
2374 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2376 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2377 Patch provided by Jan Srzednicki.
2379 TF/05 Leading white space used to be stripped from $spam_report which
2380 wrecked the formatting. Now it is preserved.
2382 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2383 that they are available at delivery time.
2385 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2387 TF/08 TLS error reporting now respects the incoming_interface and
2388 incoming_port log selectors.
2390 TF/09 Produce a more useful error message if an SMTP transport's hosts
2391 setting expands to an empty string.
2393 NM/06 Bugzilla 744: EXPN did not work under TLS.
2394 Patch provided by Phil Pennock.
2396 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2397 Patch provided by Richard Godbee.
2399 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2400 acl_smtp_notquit, added index entry.
2402 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2403 Patch provided by Eugene Bujak.
2405 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2406 accept(). Patch provided by Maxim Dounin.
2408 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2409 Patch provided by Phil Pennock.
2411 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2413 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2414 Patch provided by Brad "anomie" Jorsch.
2416 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2417 Patch provided by Dean Brooks.
2419 TK/05 Add native DKIM support (does not depend on external libraries).
2421 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2422 Patch provided by Graeme Fowler.
2424 NM/16 Bugzilla 851: Documentation example syntax fix.
2426 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2428 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2431 NM/19 Bugzilla 745: TLS version reporting.
2432 Patch provided by Phil Pennock.
2434 NM/20 Bugzilla 167: bool: condition support.
2435 Patch provided by Phil Pennock.
2437 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2438 clients. Patch provided by Phil Pennock.
2440 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2441 Patch provided by Brad "anomie" Jorsch.
2443 NM/23 Bugzilla 687: Fix misparses in eximstats.
2444 Patch provided by Heiko Schlittermann.
2446 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2447 Patch provided by Heiko Schlittermann.
2449 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2450 plus update to original patch.
2452 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2454 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2455 Patch provided by David Brownlee.
2457 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2459 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2461 NM/30 Bugzilla 888: TLS documentation bugfixes.
2463 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2465 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2466 Unlike the original bugzilla I have changed all shell scripts in src tree.
2468 NM/33 Bugzilla 898: Transport filter timeout fix.
2469 Patch by Todd Rinaldo.
2471 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2472 Patch by Serge Demonchaux.
2474 NM/35 Bugzilla 39: Base64 decode bug fixes.
2475 Patch by Jakob Hirsch.
2477 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2479 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2481 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2483 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2489 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2490 ALT-N's libdkim that I have put here:
2491 http://duncanthrax.net/exim-experimental/
2493 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2494 were called 'true' and it seems that C99 defines that as a reserved
2495 keyword to be used with 'bool' variable types. That means you could
2496 not include C99-style headers which use bools without triggering
2497 build errors in sieve.c.
2499 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2500 as mailq or other aliases. Changed the --help handling significantly
2501 to do whats expected. exim_usage() emits usage/help information.
2503 SC/01 Added the -bylocaldomain option to eximstats.
2505 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2507 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2509 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2515 PH/01 Another patch from the Sieve maintainer.
2517 PH/02 When an IPv6 address is converted to a string for single-key lookup
2518 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2519 dots are used instead of colons so that keys in lsearch files need not
2520 contain colons. This was done some time before quoting was made available
2521 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2522 (notated using the quote facility) so as to distinguish them from IPv4
2523 keys. This meant that lookups for IP addresses in host lists did not work
2524 for iplsearch lookups.
2526 This has been fixed by arranging for IPv6 addresses to be expressed with
2527 colons if the lookup type is iplsearch. This is not incompatible, because
2528 previously such lookups could never work.
2530 The situation is now rather anomalous, since one *can* have colons in
2531 ordinary lsearch keys. However, making the change in all cases is
2532 incompatible and would probably break a number of configurations.
2534 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2537 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2538 conversion specification without a maximum field width, thereby enabling
2539 a rogue spamd server to cause a buffer overflow. While nobody in their
2540 right mind would setup Exim to query an untrusted spamd server, an
2541 attacker that gains access to a server running spamd could potentially
2542 exploit this vulnerability to run arbitrary code as the Exim user.
2544 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2545 $primary_hostname instead of what libspf2 thinks the hosts name is.
2547 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2548 a directory entry by the name of the lookup key. Previously, if a
2549 symlink pointed to a non-existing file or a file in a directory that
2550 Exim lacked permissions to read, a lookup for a key matching that
2551 symlink would fail. Now it is enough that a matching directory entry
2552 exists, symlink or not. (Bugzilla 503.)
2554 PH/03 The body_linecount and body_zerocount variables are now exported in the
2557 PH/04 Added the $dnslist_matched variable.
2559 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2560 This means they are set thereafter only if the connection becomes
2563 PH/06 Added the client_condition to authenticators so that some can be skipped
2564 by clients under certain conditions.
2566 PH/07 The error message for a badly-placed control=no_multiline_responses left
2567 "_responses" off the end of the name.
2569 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2571 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2572 (without spaces) instead of just copying the configuration text.
2574 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2576 PH/11 Added $max_received_linelength.
2578 PH/12 Added +ignore_defer and +include_defer to host lists.
2580 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2581 way in which PCRE > 7.0 is built.
2583 PH/14 Implemented queue_only_load_latch.
2585 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2586 MAIL command. The effect was to mangle the value on 64-bit systems.
2588 PH/16 Another patch from the Sieve maintainer.
2590 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2592 PH/18 If a system quota error occurred while trying to create the file for
2593 a maildir delivery, the message "Mailbox is full" was not appended to the
2594 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2595 only to a quota excession during the actual writing of the file.
2597 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2598 characters?) which causes problems in log lines. The DN values are now
2599 passed through string_printing() before being added to log lines.
2601 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2602 and InterBase are left for another time.)
2604 PH/21 Added message_body_newlines option.
2606 PH/22 Guard against possible overflow in moan_check_errorcopy().
2608 PH/23 POSIX allows open() to be a macro; guard against that.
2610 PH/24 If the recipient of an error message contained an @ in the local part
2611 (suitably quoted, of course), incorrect values were put in $domain and
2612 $local_part during the evaluation of errors_copy.
2618 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2619 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2622 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2623 issue a MAIL command.
2625 PH/02 In an ACL statement such as
2627 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2629 if a client was not listed at all, or was listed with a value other than
2630 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2631 the condition was not true (as it should be), so access was not denied.
2632 The bug was that the ! inversion was incorrectly passed on to the second
2633 item. This has been fixed.
2635 PH/03 Added additional dnslists conditions == and =& which are different from
2636 = and & when the dns lookup returns more than one IP address.
2638 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2639 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2641 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2642 FSYNC, which compiles an option called disable_fsync that allows for
2643 bypassing fsync(). The documentation is heavily laced with warnings.
2645 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2647 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2648 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2649 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2650 including adding "make clean"; (3) Added -fPIC when compiling the test
2651 dynamically loaded module, to get rid of a warning.
2653 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2654 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2655 = 0s. The bug is otherwise harmless.
2657 PH/07 There was a bug in the dovecot authenticator such that the value of
2658 $auth1 could be overwritten, and so not correctly preserved, after a
2659 successful authentication. This usually meant that the value preserved by
2660 the server_setid option was incorrect.
2662 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2664 PH/09 Installed PCRE release 7.0.
2666 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2667 run for batched SMTP input. It is now run at the start of every message
2668 in the batch. While fixing this I discovered that the process information
2669 (output by running exiwhat) was not always getting set for -bs and -bS
2670 input. This is fixed, and it now also says "batched" for BSMTP.
2672 PH/11 Added control=no_pipelining.
2674 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2675 patch, slightly modified), and move the expansion of helo_data till after
2676 the connection is made in the smtp transport (so it can use these
2679 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2681 PH/14 Added log_selector = +pid.
2683 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2685 PH/16 Add ${if forany and ${if forall.
2687 PH/17 Added dsn_from option to vary the From: line in DSNs.
2689 PH/18 Flush SMTP output before performing a callout, unless control =
2690 no_callout_flush is set.
2692 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2693 was true (the default) a successful delivery failed to delete the retry
2694 item, thus causing premature timeout of the address. The bug is now
2697 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2699 PH/21 Long custom messages for fakedefer and fakereject are now split up
2700 into multiline responses in the same way that messages for "deny" and
2701 other ACL rejections are.
2703 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2704 with slight modification.
2706 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2707 draft, changing the syntax and factoring some duplicate code".
2709 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2710 for deliveries of the second and subsequent messages over the same SMTP
2713 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2714 ${reduce, with only minor "tidies".
2716 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2718 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2719 expansion side effects.
2721 PH/27 When a message times out after an over-quota error from an Exim-imposed
2722 quota, the bounce message says "mailbox is full". This message was not
2723 being given when it was a system quota that was exceeded. It now should
2726 MH/03 Made $recipients available in local_scan(). local_scan() already has
2727 better access to the recipient list through recipients_list[], but
2728 $recipients can be useful in postmaster-provided expansion strings.
2730 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2731 in the case of a MAIL command with additional options following the
2732 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2733 were accidentally chopped off.
2735 PH/29 SMTP synchronization checks are implemented when a command is read -
2736 there is a check that no more input is waiting when there shouldn't be
2737 any. However, for some commands, a delay in an ACL can mean that it is
2738 some time before the response is written. In this time, more input might
2739 arrive, invalidly. So now there are extra checks after an ACL has run for
2740 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2741 pipelining has not been advertised.
2743 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2745 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2746 returned in the SMTP response when smtp_return_error_details was false.
2747 This has been fixed.
2749 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2750 instead of the C I/O that was originally supplied, because problems were
2751 reported on Solaris.
2753 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2754 Exim which did not show up earlier: it was assuming that a call to
2755 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2756 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2757 was a macro that became an assignment, so it seemed to work. This has
2758 changed to a proper function call with a void return, hence the compile
2759 error. Exim's code has been fixed.
2761 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2764 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2766 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2768 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2769 because it turns out to be too big or there is a timeout) there is no
2770 "Completed" line in the log. When some messages of this type were
2771 selected by exigrep, they were listed as "not completed". Others were
2772 picked up by some special patterns. I have improved the selection
2773 criteria to be more general.
2775 PH/38 The host_find_failed option in the manualroute router can now be set
2776 to "ignore", to completely ignore a host whose IP address cannot be
2777 found. If all hosts are ignored, the behaviour is controlled by the new
2778 host_all_ignored option.
2780 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2781 homing or because of multiple MX records with /mx) generated more than
2782 one IP address, and the following item turned out to be the local host,
2783 all the secondary addresses of the first item were incorrectly removed
2784 from the list, along with the local host and any following hosts (which
2785 is what is supposed to happen).
2787 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2788 whoever called Exim into the -H file. In the case of the daemon it was
2789 behaving confusingly. When first started, it used values for whoever
2790 started the daemon, but after a SIGHUP it used the Exim user (because it
2791 calls itself on a restart). I have changed the code so that it now always
2794 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2795 message are rejected with the same error (e.g. no authentication or bad
2796 sender address), and a DATA command is nevertheless sent (as can happen
2797 with PIPELINING or a stupid MUA), the error message that was given to the
2798 RCPT commands is included in the rejection of the DATA command. This is
2799 intended to be helpful for MUAs that show only the final error to their
2802 PH/42 Another patch from the Sieve maintainer.
2804 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2805 Eximstats - Fixed some broken HTML links and added missing column headers
2807 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2808 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2810 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2812 PH/43 Yet another patch from the Sieve maintainer.
2814 PH/44 I found a way to check for a TCP/IP connection going away before sending
2815 the response to the final '.' that terminates a message, but only in the
2816 case where the client has not sent further data following the '.'
2817 (unfortunately, this is allowed). However, in many cases there won't be
2818 any further data because there won't be any more messages to send. A call
2819 to select() can be used: if it shows that the input is "ready", there is
2820 either input waiting, or the socket has been closed. An attempt to read
2821 the next input character can distinguish the two cases. Previously, Exim
2822 would have sent an OK response which the client would never have see.
2823 This could lead to message repetition. This fix should cure that, at
2824 least in a lot of common cases.
2826 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2827 advertised in response to EHLO.
2833 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2834 fixed by 4.65/MH/01 (is this a record?) are fixed:
2836 (i) An empty string was always treated as zero by the numeric comparison
2837 operators. This behaviour has been restored.
2839 (ii) It is documented that the numeric comparison operators always treat
2840 their arguments as decimal numbers. This was broken in that numbers
2841 starting with 0 were being interpreted as octal.
2843 While fixing these problems I realized that there was another issue that
2844 hadn't been noticed. Values of message_size_limit (both the global option
2845 and the transport option) were treated as octal if they started with 0.
2846 The documentation was vague. These values are now always treated as
2847 decimal, and I will make that clear in the documentation.
2853 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2854 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2857 MH/01 Don't check that the operands of numeric comparison operators are
2858 integers when their expansion is in "skipping" mode (fixes bug
2859 introduced by 4.64-PH/07).
2861 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2862 child addresses, Exim now panics and dies. Previously, because the count
2863 is held in a short int, deliveries were likely to be lost. As such a
2864 large number of recipients for a single message is ridiculous
2865 (performance will be very, very poor), I have chosen to impose a limit
2866 rather than extend the field.
2872 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2873 leftover -K file (the existence of which was triggered by #402).
2874 While we were at it, introduced process PID as part of the -K
2875 filename. This should rule out race conditions when creating
2878 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2879 processing considerably. Previous code took too long for large mails,
2880 triggering a timeout which in turn triggers #401.
2882 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2883 in the DK code in transports.c. sendfile() is not really portable,
2884 hence the _LINUX specificness.
2886 TF/01 In the add_headers option to the mail command in an Exim filter,
2887 there was a bug that Exim would claim a syntax error in any
2888 header after the first one which had an odd number of characters
2891 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2892 callout verification, Exim cached a "reject" for the entire domain. This
2893 is correct for most verifications, but it is not correct for a recipient
2894 verification with use_sender or use_postmaster set, because in that case
2895 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2896 case of MAIL FROM:<> rejection from other early rejections (e.g.
2897 rejection of HELO). When verifying a recipient using a non-null MAIL
2898 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2899 Whatever the result of the callout, the value of the domain cache is
2900 left unchanged (for any other kind of callout, getting as far as trying
2901 RCPT means that the domain itself is ok).
2903 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2906 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2907 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2908 session, the daemon ignores ECONNECT errors and logs others; it now
2909 ignores EPIPE as well.
2911 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2912 (quoted-printable decoding).
2914 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2915 later the small subsequent patch to fix an introduced bug.
2917 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2919 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2921 PH/08 An error is now given if message_size_limit is specified negative.
2923 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2924 to be given (somewhat) arbitrary names.
2926 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2929 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2930 miscellaneous code fixes
2932 PH/10 Added the log_reject_target ACL modifier to specify where to log
2935 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2936 hostname. This is wrong, because it relates to the incoming message (and
2937 probably the interface on which it is arriving) and not to the outgoing
2938 callout (which could be using a different interface). This has been
2939 changed to use the value of the helo_data option from the smtp transport
2940 instead - this is what is used when a message is actually being sent. If
2941 there is no remote transport (possible with a router that sets up host
2942 addresses), $smtp_active_hostname is used.
2944 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2945 tweaks were necessary in order to get it to work (see also 21 below):
2946 (a) The code assumed that strncpy() returns a negative number on buffer
2947 overflow, which isn't the case. Replaced with Exim's string_format()
2949 (b) There were several signed/unsigned issues. I just did the minimum
2950 hacking in of casts. There is scope for a larger refactoring.
2951 (c) The code used strcasecmp() which is not a standard C function.
2952 Replaced with Exim's strcmpic() function.
2953 (d) The code set only $1; it now sets $auth1 as well.
2954 (e) A simple test gave the error "authentication client didn't specify
2955 service in request". It would seem that Dovecot has changed its
2956 interface. Fortunately there's a specification; I followed it and
2957 changed what the client sends and it appears to be working now.
2959 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2962 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2963 address A is aliased to B and C, where B exists and C does not. Without
2964 -v the output is "A verified" because verification stops after a
2965 successful redirection if more than one address is generated. However,
2966 with -v the child addresses are also verified. Exim was outputting "A
2967 failed to verify" and then showing the successful verification for C,
2968 with its parentage. It now outputs "B failed to verify", showing B's
2969 parentage before showing the successful verification of C.
2971 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2972 look up a TXT record in a specific list after matching in a combined
2975 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2976 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2977 they consult the DNS. I had assumed they would set it the way they
2978 wanted; and indeed my experiments on Linux seem to show that in some
2979 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2980 To be on the safe side, however, I have now made the interface to
2981 host_find_byname() similar to host_find_bydns(), with an argument
2982 containing the DNS resolver options. The host_find_byname() function now
2983 sets these options at its start, just as host_find_bydns() does. The smtp
2984 transport options dns_qualify_single and dns_search_parents are passed to
2985 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2986 of host_find_byname() use the default settings of RES_DEFNAMES
2987 (qualify_single) but not RES_DNSRCH (search_parents).
2989 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2990 spool_read_header() do less string testing, by means of a preliminary
2991 switch on the second character of optional "-foo" lines. (This is
2992 overdue, caused by the large number of possibilities that now exist.
2993 Originally there were few.) While I was there, I also converted the
2994 str(n)cmp tests so they don't re-test the leading "-" and the first
2995 character, in the hope this might squeeze out yet more improvement.
2997 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2998 flag allowing group syntax was set by the header_syntax check but not
2999 turned off, possible causing trouble later; (2) The flag was not being
3000 set at all for the header_verify test, causing "group"-style headers to
3001 be rejected. I have now set it in this case, and also caused header_
3002 verify to ignore an empty address taken from a group. While doing this, I
3003 came across some other cases where the code for allowing group syntax
3004 while scanning a header line wasn't quite right (mostly, not resetting
3005 the flag correctly in the right place). These bugs could have caused
3006 trouble for malformed header lines. I hope it is now all correct.
3008 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3009 with the "reply" argument non-NULL. The code, however (which originally
3010 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3011 but it didn't always do it. This confused somebody who was copying the
3012 code for some other use. I have removed all the tests.
3014 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3015 feature that was used to support insecure browsers during the U.S. crypto
3016 embargo. It requires special client support, and Exim is probably the
3017 only MTA that supported it -- and would never use it because real RSA is
3018 always available. This code has been removed, because it had the bad
3019 effect of slowing Exim down by computing (never used) parameters for the
3020 RSA_EXPORT functionality.
3022 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3023 authenticator to fail if there's a tab character in the incoming data
3024 (there should never be unless someone is messing about, as it's supposed
3025 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3026 if the connection is using TLS or if the remote IP is the same as the
3027 local IP, and the "valid-client-cert option" if a client certificate has
3030 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3031 authenticators. This can be used for authorization after authentication
3032 succeeds. (In the case of plaintext, it servers for both authentication
3035 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3036 if any retry times were supplied.
3038 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3039 connection, where there is no client IP address to check. In this
3040 situation, the verify now always succeeds.
3042 PH/25 Applied John Jetmore's -Mset patch.
3044 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3046 PH/27 In a string expansion for a processed (not raw) header when multiple
3047 headers of the same name were present, leading whitespace was being
3048 removed from all of them, but trailing whitespace was being removed only
3049 from the last one. Now trailing whitespace is removed from each header
3050 before concatenation. Completely empty headers in a concatenation (as
3051 before) are ignored.
3053 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3054 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3056 PH/29 [Removed. This was a change that I later backed out, and forgot to
3057 correct the ChangeLog entry (that I had efficiently created) before
3058 committing the later change.]
3060 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3061 address errors (4xx response to RCPT) over the same connection as other
3062 messages routed to the same hosts. Such deliveries are always "forced",
3063 so retry times are not inspected. This resulted in far too many retries
3064 for the affected addresses. The effect occurred only when there were more
3065 hosts than the hosts_max_try setting in the smtp transport when it had
3066 the 4xx errors. Those hosts that it had tried were not added to the list
3067 of hosts for which the message was waiting, so if all were tried, there
3068 was no problem. Two fixes have been applied:
3070 (i) If there are any address or message errors in an SMTP delivery, none
3071 of the hosts (tried or untried) are now added to the list of hosts
3072 for which the message is waiting, so the message should not be a
3073 candidate for sending over the same connection that was used for a
3074 successful delivery of some other message. This seems entirely
3075 reasonable: after all the message is NOT "waiting for some host".
3076 This is so "obvious" that I'm not sure why it wasn't done
3077 previously. Hope I haven't missed anything, but it can't do any
3078 harm, as the worst effect is to miss an optimization.
3080 (ii) If, despite (i), such a delivery is accidentally attempted, the
3081 routing retry time is respected, so at least it doesn't keep
3082 hammering the server.
3084 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3085 in ${readsocket because some servers need this prod.
3087 PH/32 Added some extra debug output when updating a wait-xxx database.
3089 PH/33 The hint "could be header name not terminated by colon", which has been
3090 given for certain expansion errors for a long time, was not being given
3091 for the ${if def:h_colon_omitted{... case.
3093 PH/34 The spec says: "With one important exception, whenever a domain list is
3094 being scanned, $domain contains the subject domain." There was at least
3095 one case where this was not true.
3097 PH/35 The error "getsockname() failed: connection reset by peer" was being
3098 written to the panic log as well as the main log, but it isn't really
3099 panic-worthy as it just means the connection died rather early on. I have
3100 removed the panic log writing for the ECONNRESET error when getsockname()
3103 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3104 runs only) independently of the message's sender address. This meant
3105 that, if the 4xx error was in fact related to the sender, a different
3106 message to the same recipient with a different sender could confuse
3107 things. In particular, this can happen when sending to a greylisting
3108 server, but other circumstances could also provoke similar problems.
3109 I have changed the default so that the retry time for these errors is now
3110 based a combination of the sender and recipient addresses. This change
3111 can be overridden by setting address_retry_include_sender=false in the
3114 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3115 remote server are returned as part of bounce messages. This was not
3116 happening for LMTP over a pipe (the lmtp transport), but now it is the
3117 same for both kinds of LMTP.
3119 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3120 in header lines that were in fact CNAMEs. This is no longer the case.
3122 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3123 and queue runs started by the daemon processed all messages. This has
3124 been fixed so that -R and -S can now usefully be given with -q<time>.
3126 PH/40 Import PCRE release 6.7 (fixes some bugs).
3128 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3130 PH/42 Give an error if -q is specified more than once.
3132 PH/43 Renamed the variables $interface_address and $interface_port as
3133 $received_ip_address and $received_port, to make it clear that these
3134 values apply to message reception, and not to the outgoing interface when
3135 a message is delivered. (The old names remain recognized, of course.)
3137 PH/44 There was no timeout on the connect() call when using a Unix domain
3138 socket in the ${readsocket expansion. There now is.
3140 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3141 be meaningful with "accept".
3143 SC/01 Eximstats V1.43
3144 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3146 SC/02 Eximstats V1.44
3147 Use a glob alias rather than an array ref in the generated
3148 parser. This improves both readability and performance.
3150 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3151 Collect SpamAssassin and rejection statistics.
3152 Don't display local sender or destination tables unless
3153 there is data to show.
3154 Added average volumes into the top table text output.
3156 SC/04 Eximstats V1.46
3157 Collect data on the number of addresses (recipients)
3158 as well as the number of messages.
3160 SC/05 Eximstats V1.47
3161 Added 'Message too big' to the list of mail rejection
3162 reasons (thanks to Marco Gaiarin).
3164 SC/06 Eximstats V1.48
3165 Mainlog lines which have GMT offsets and are too short to
3166 have a flag are now skipped.
3168 SC/07 Eximstats V1.49 (Alain Williams)
3169 Added the -emptyok flag.
3171 SC/08 Eximstats V1.50
3172 Fixes for obtaining the IP address from reject messages.
3174 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3175 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3176 whitespace changes from 4.64-PH/27
3178 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3181 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3182 are found, allow negative numbers in numeric criteria)
3184 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3186 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3187 to match changes made in 4.64-PH/43
3189 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3191 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3192 "domain" is actually an IP address, and force a failure. This locks out
3193 those revolvers/nameservers that support "A-for-A" lookups, in
3194 contravention of the specifications.
3196 PH/48 When a host name was looked up from an IP address, and the subsequent
3197 forward lookup of the name timed out, the host name was left in
3198 $sender_host_name, contrary to the specification.
3200 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3201 restricted to single-key lookups, Exim was not diagnosing an error if
3202 * or *@ was used with a query-style lookup.
3204 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3206 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3207 long ago, but noone interested enough thought of it. Let's just say that
3208 the "1.1" means that there are some new functions that weren't there at
3209 some point in the past.
3211 PH/51 Error processing for expansion failure of helo_data from an smtp
3212 transport during callout processing was broken.
3214 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3215 tested/used via the -bh/-bhc/-bs options.
3217 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3218 bug, fixed in subsequent PCRE releases).
3220 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3221 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3223 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3229 SC/01 Use a glob alias rather than an array ref in eximstats generated
3230 parser. This improves both readability and performance.
3232 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3233 Don't display local sender or destination tables in eximstats unless
3234 there is data to show.
3235 Added average volumes into the eximstats top table text output.
3237 SC/03 Collect data on the number of addresses (recipients) as well
3238 as the number of messages in eximstats.
3240 TF/01 Correct an error in the documentation for the redirect router. Exim
3241 does (usually) call initgroups() when daemonizing.
3243 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3244 with consistent privilege compared to when running as a daemon.
3246 TF/03 Note in the spec that $authenticated_id is not set for local
3247 submissions from trusted users.
3249 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3250 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3252 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3253 by adding some example configuration directives to the default
3254 configuration file. A little bit of work is required to uncomment the
3255 directives and define how usernames and passwords are checked, but
3256 there is now a framework to start from.
3258 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3259 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3260 without this. I don't know how relevant this is to other LDAP libraries.
3262 PH/02 Add the verb name to the "unknown ACL verb" error.
3264 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3266 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3268 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3269 directory not expanded when it should be if an expanded home directory
3270 was set for the address (which is overridden by the transport).
3272 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3275 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3276 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3277 because it is too late at that time, and has no effect.
3279 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3280 security issue with \' (bugzilla #107). I could not use the
3281 PQescapeStringConn() function, because it needs a PGconn value as one of
3284 PH/08 When testing addresses using -bt, indicate those final addresses that
3285 are duplicates that would not cause an additional delivery. At least one
3286 person was confused, thinking that -bt output corresponded to deliveries.
3287 (Suppressing duplicates isn't a good idea as you lose the information
3288 about possibly different redirections that led to the duplicates.)
3290 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3291 systems where poll() doesn't work, in particular OS X.
3293 PH/10 Added more information to debugging output for retry time not reached.
3295 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3296 operations in malware.c.
3298 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3301 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3302 syslog_duplication set false, log lines that would normally be written
3303 both the the main log and to the reject log were not written to syslog at
3306 PH/14 In the default configuration, change the use of "message" in ACL warn
3307 statements to "add_header".
3309 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3310 not followed by a command (e.g. "seen endif").
3312 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3313 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3316 PH/17 Added extra conditions to the default value of delay_warning_condition
3320 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3321 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3322 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3325 The Auto-Submitted: and various List- headers are standardised, whereas I
3326 don't think Precedence: ever was.
3328 PH/18 Refactored debugging code in route_finduser() to show more information,
3329 in particular, the error code if getpwnam() issues one.
3331 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3332 This is apparently needed in addition to the PH/07 change above to avoid
3333 any possible encoding problems.
3335 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3336 but not after initializing Perl.
3338 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3339 output them only if debugging. By default they are written stderr,
3340 apparently, which is not desirable.
3342 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3345 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3348 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3350 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3351 authenticated or an ident call has been made. Suppress the default
3352 values for $authenticated_id and $authenticated_sender (but permit -oMai
3353 and -oMas) when testing with -bh.
3355 PH/24 Re-jigged the order of the tests in the default configuration so that the
3356 tests for valid domains and recipients precede the DNS black list and CSA
3357 tests, on the grounds that those ones are more expensive.
3359 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3360 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3361 This bug exists in every version of Exim that I still have, right back to
3364 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3365 However, an attempt to turn on case-sensitivity in a regex key by
3366 including (?-i) didn't work because the subject string was already
3367 lowercased, and the effects were non-intuitive. It turns out that a
3368 one-line patch can be used to allow (?-i) to work as expected.
3374 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3375 other effects) broke the use of negated acl sub-conditions.
3377 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3380 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3381 "Deny" causes Exim to reject the incoming connection with a 554 error.
3382 Unfortunately, if there is a major crisis, such as a disk failure,
3383 tcp-wrappers gives "deny", whereas what one would like would be some
3384 kind of temporary error. A kludge has been added to help with this.
3385 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3386 554 error is used if errno is still zero or contains ENOENT (which occurs
3387 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3390 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3392 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3393 errors. Otherwise a message that provokes a temporary error (when other
3394 messages do not) can cause a whole host to time out.
3396 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3397 addresses were routed directly to files or pipes from a redirect router.
3398 File deliveries just didn't batch; pipe deliveries might have suffered
3401 PH/06 A failure to get a lock for a hints database would erroneously always say
3402 "Failed to get write lock", even when it was really a read lock.
3404 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3405 of 0600. This has been changed to use the value of the lockfile_mode
3406 option (which defaults to 0600).
3408 PH/08 Applied small patch from the Sieve maintainer.
3410 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3411 folder from quota calculations, a direct delivery into this folder messed
3412 up the contents of the maildirsize file. This was because the regex was
3413 used only to exclude .Trash (or whatever) when the size of the mailbox
3414 was calculated. There was no check that a delivery was happening into an
3415 excluded directory. This bug has been fixed by ignoring all quota
3416 processing for deliveries into excluded directories.
3418 PH/10 Added the maildirfolder_create_regex option to appendfile.
3424 PH/01 The code for finding all the local interface addresses on a FreeBSD
3425 system running IPv6 was broken. This may well have applied to all BSD
3426 systems, as well as to others that have similar system calls. The broken
3427 code found IPv4 interfaces correctly, but gave incorrect values for the
3428 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3429 that it would not match correctly against @[] and not recognize the IPv6
3432 PH/02 The ipliteral router was not recognizing addresses of the form user@
3433 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3435 PH/03 Added disable_ipv6.
3437 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3438 decoded form, because it is most often used to construct To: headers
3439 lines in autoreplies, and the decoded form may well be syntactically
3440 invalid. However, $reply_address has leading white space removed, and all
3441 newlines turned into spaces so that the autoreply transport does not
3444 PH/05 If group was specified without a user on a router, and no group or user
3445 was specified on a transport, the group from the router was ignored.
3447 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3448 for visible compile-time settings that can be used to change these
3449 numbers, for those that want even more. Backwards compatibility with old
3450 spool files has been maintained. However, going back to a previous Exim
3451 release will lost any variables that are in spool files.
3453 PH/07 Two small changes when running in the test harness: increase delay when
3454 passing a TCP/IP connection to a new process, in case the original
3455 process has to generate a bounce, and remove special handling of
3456 127.0.0.2 (sic), which is no longer necessary.
3458 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3459 be the same on different OS.
3461 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3464 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3465 whether --show-vars was specified or not
3467 JJ/02 exipick: Added support for new ACL variable spool format introduced
3470 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3471 syntactically invalid From: or Reply-to: line, and a filter used this to
3472 generate an autoreply, and therefore failed to obtain an address for the
3473 autoreply, Exim could try to deliver to a non-existent relative file
3474 name, causing unrelated and misleading errors. What now happens is that
3475 it logs this as a hard delivery error, but does not attempt to create a
3478 PH/11 The exinext utility has a -C option for testing purposes, but although
3479 the given file was scanned by exinext itself; it wasn't being passed on
3480 when Exim was called.
3482 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3483 an end-of-file indication when reading a command response.
3485 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3486 compiled. In many other places in Exim, IPv6 addresses are always
3487 recognized, so I have changed this. It also means that IPv4 domain
3488 literals of the form [IPV4:n.n.n.n] are now always recognized.
3490 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3491 used if the router is not running as root, for example, when verifying at
3492 ACL time, or when using -bh. The debugging output from this situation was
3493 non-existent - all you got was a failure to exec. I have made two
3496 (a) Failures to set uid/gid, the current directory, or a process leader
3497 in a subprocess such as that created by queryprogram now generate
3498 suitable debugging output when -d is set.
3500 (b) The queryprogram router detects when it is not running as root,
3501 outputs suitable debugging information if -d is set, and then runs
3502 the subprocess without attempting to change uid/gid.
3504 PH/15 Minor change to Makefile for building test_host (undocumented testing
3507 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3508 additional section of a DNS packet that returns MX or SRV records.
3509 Instead, it always explicitly searches for A/AAAA records. This avoids
3510 major problems that occur when a DNS server includes only records of one
3511 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3512 fixed another bug: if SRV records were looked up and the corresponding
3513 address records were *not* found in the additional section, the port
3514 values from the SRV records were lost.
3516 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3517 using the correct key (the original address) when searching the retry
3518 rules in order to find which one to use for generating the retry hint.
3520 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3521 adding the default one. Similarly, if it contains a Reply-To: header, the
3522 errors_reply_to option, if set, is not used.
3524 PH/19 When calculating a retry time, Exim used to measure the "time since
3525 failure" by looking at the "first failed" field in the retry record. Now
3526 it does not use this if it is later than than the arrival time of the
3527 message. Instead it uses the arrival time. This makes for better
3528 behaviour in cases where some deliveries succeed, thus re-setting the
3529 "first failed" field. An example is a quota failure for a huge message
3530 when small messages continue to be delivered. Without this change, the
3531 "time since failure" will always be short, possible causing more frequent
3532 delivery attempts for the huge message than are intended.
3533 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3535 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3536 $1, $2, $3) because the numerical variables can be reset during some
3537 expansion items (e.g. "match"), thereby losing the authentication data.
3539 PH/21 Make -bV show the size of off_t variables so that the test suite can
3540 decide whether to run tests for quotas > 2G.
3542 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3543 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3544 filecount value is greater than 2G or if a quota value is greater than 2G
3545 on a system where the size of off_t is not greater than 4, a panic error
3548 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3549 never match. The debug and -bh output now contains an explicit error
3550 message indicating a malformed IPv4 address or mask.
3552 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3553 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3554 PH/23 above applies.
3556 PH/25 Do not write to syslog when running in the test harness. The only
3557 occasion when this arises is a failure to open the main or panic logs
3558 (for which there is an explicit test).
3560 PH/26 Added the /no_tell option to "control=freeze".
3562 PH/27 If a host name lookup failed very early in a connection, for example, if
3563 the IP address matched host_lookup and the reverse lookup yielded a name
3564 that did not have a forward lookup, an error message of the form "no IP
3565 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3566 could be logged. Now it outputs the IP address instead of "NULL".
3568 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3569 allows the sender and the authenticated sender to be set when
3570 submitting a message from within Exim. Since child_open_exim() is
3571 documented for local_scan(), the new function should be too.
3573 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3574 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3575 results in an empty string is now treated as unset.
3577 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3579 PH/31 Added sender_verify_fail logging option.
3581 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3582 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3583 tidied the source and removed it altogether.
3585 PH/33 When a queue run was abandoned because the load average was too high, a
3586 log line was always written; now it is written only if the queue_run log
3587 selector is set. In addition, the log line for abandonment now contains
3588 information about the queue run such as the pid. This is always present
3589 in "start" and "stop" lines but was omitted from the "abandon" line.
3591 PH/34 Omit spaces between a header name and the colon in the error message that
3592 is given when verify = headers_syntax fails (if there are lots of them,
3593 the message gets confusing).
3595 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3596 names, as there are now some PTR records that contain slashes. This check
3597 is only to protect against broken name servers that fall over on strange
3598 characters, so the fact that it applies to all lookups doesn't matter.
3600 PH/36 Now that the new test suite is complete, we can remove some of the
3601 special code in Exim that was needed for the old test suite. For example,
3602 sorting DNS records because real resolvers return them in an arbitrary
3603 order. The new test suite's fake resolver always returns records in the
3606 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3607 bounces) except when queue_only is set, to avoid logging races between
3608 the different processes.
3610 PH/38 Panic-die if .include specifies a non-absolute path.
3612 PH/39 A tweak to the "H" retry rule from its user.
3614 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3615 a label. They prevented compilation on older perls.
3617 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3618 a warning to be raised on newish perls.
3620 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3621 on queue. Changes to match documented behaviour of showing count of
3622 messages matching specified criteria.
3624 PH/40 Changed the default ident timeout from 30s to 5s.
3626 PH/41 Added support for the use of login_cap features, on those BSD systems
3627 that have them, for controlling the resources used by pipe deliveries.
3629 PH/42 The content-scanning code uses fopen() to create files in which to put
3630 message data. Previously it was not paying any attention to the mode of
3631 the files. Exim runs with umask(0) because the rest of the code creates
3632 files with open(), and sets the required mode explicitly. Thus, these
3633 files were ending up world-writeable. This was not a big issue, because,
3634 being within the spool directory, they were not world-accessible. I have
3635 created a function called modefopen, which takes an additional mode
3636 argument. It sets umask(777), creates the file, chmods it to the required
3637 mode, then resets the umask. All the relevant calls to fopen() in the
3638 content scanning code have been changed to use this function.
3640 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3641 to 24 hours. This avoids potential overflow problems when processing G
3642 and H retry rules. I suspect nobody ever tinkers with this value.
3644 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3646 PH/45 When the plaintext authenticator is running as a client, the server's
3647 challenges are checked to ensure they are valid base64 strings. By
3648 default, the authentication attempt is cancelled if an invalid string is
3649 received. Setting client_ignore_invalid_base64 true ignores these errors.
3650 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3651 they are received. Thus, the responses can be made to depend on the
3652 challenges. If an invalid string is ignored, an empty string is placed in
3655 PH/46 Messages that are created by the autoreply transport now contains a
3656 References: header, in accordance with RFCs 2822 and 3834.
3658 PH/47 Added authenticated_sender_force to the smtp transport.
3660 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3662 PH/49 Installed latest patch from the Sieve maintainer.
3664 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3665 was also set, the appendfile transport was unnecessarily scanning a
3666 directory of message files (e.g. for maildir delivery) to find the count
3667 of files (along with the size), even though it did not need this
3668 information. It now does the scan only if it needs to find either the
3669 size of the count of files.
3671 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3673 PH/52 Two bugs concerned with error handling when the smtp transport is
3676 (i) Exim was not creating retry information for temporary errors given
3677 for individual recipients after the DATA command when the smtp transport
3678 was used in LMTP mode. This meant that they could be retried too
3679 frequently, and not timed out correctly.
3681 (ii) Exim was setting the flag that allows error details to be returned
3682 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3683 recipients that were returned after the DATA command.
3685 PH/53 This is related to PH/52, but is more general: for any failing address,
3686 when detailed error information was permitted to be returned to the
3687 sender, but the error was temporary, then after the final timeout, only
3688 "retry timeout exceeded" was returned. Now it returns the full error as
3689 well as "retry timeout exceeded".
3691 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3692 do this, and (what is worse) MTAs that accept it.
3694 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3695 will now be deprecated.
3697 PH/56 New os.c-cygwin from the Cygwin maintainer.
3699 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3700 formats (previously only available in exim formats via -bpr, -bpru,
3701 and -bpra. Now also available in native and exiqgrep formats)
3703 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3704 with very large, slow to parse queues
3706 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3708 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3710 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3711 responses to authentication challenges, though it was showing the
3712 challenges; (ii) I've removed the CR characters from the debug output for
3715 PH/58 Allow for the insertion of a newline as well as a space when a string
3716 is turned into more than one encoded-word during RFC 2047 encoding. The
3717 Sieve code now uses this.
3719 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3720 data_4xx, lost_connection, tls_required.
3722 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3723 message was being sent as an SMTP response.
3725 PH/61 Add -l and -k options to exicyclog.
3727 PH/62 When verifying, if an address was redirected to one new address, so that
3728 verification continued, and the new address failed or deferred after
3729 having set something in $address_data, the value of $address_data was not
3730 passed back to the ACL. This was different to the case when no
3731 redirection occurred. The value is now passed back in both cases.
3733 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3734 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3735 use login_cap.h, so on its own it isn't the distinguishing feature. The
3736 new name refers directly to the setclassresources() function.
3738 PH/65 Added configuration files for NetBSD3.
3740 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3742 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3743 is preferred over IPv4.
3745 PH/68 The bounce_return_message and bounce_return_body options were not being
3746 honoured for bounces generated during the reception of non-SMTP messages.
3747 In particular, this applied to messages rejected by the ACL. This bug has
3748 been fixed. However, if bounce_return_message is true and bounce_return_
3749 body is false, the headers that are returned for a non-SMTP message
3750 include only those that have been read before the error was detected.
3751 (In the case of an ACL rejection, they have all been read.)
3753 PH/69 The HTML version of the specification is now built in a directory called
3754 spec_html instead of spec.html, because the latter looks like a path with
3755 a MIME-type, and this confuses some software.
3757 PH/70 Catch two compiler warnings in sieve.c.
3759 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3760 function verify_get_ident() calls ip_connect() to connect a socket, but
3761 if the "connect()" function timed out, ip_connect() used to close the
3762 socket. However, verify_get_ident() also closes the socket later, and in
3763 between Exim writes to the log, which may get opened at this point. When
3764 the socket was closed in ip_connect(), the log could get the same file
3765 descriptor number as the socket. This naturally causes chaos. The fix is
3766 not to close the socket in ip_connect(); the socket should be closed by
3767 the function that creates it. There was only one place in the code where
3768 this was missing, in the iplookup router, which I don't think anybody now
3769 uses, but I've fixed it anyway.
3771 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3772 well as to direct DNS lookups. Otherwise the handling of names in host
3773 lists is inconsistent and therefore confusing.
3779 PH/01 Two changes to the default runtime configuration:
3781 (1) Move the checks for relay_from_hosts and authenticated clients from
3782 after to before the (commented out) DNS black list checks.
3784 (2) Add control=submission to the relay_from_hosts and authenticated
3785 clients checks, on the grounds that messages accepted by these
3786 statements are most likely to be submissions.
3788 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3790 (1) Generate an error if the third argument for the ${prvs expansion is
3793 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3796 (3) Reset the variables that are obtained from the first argument of
3797 ${prvscheck and used in the second argument before leaving the code,
3798 because their memory is reclaimed, so using them afterwards may do
3801 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3802 one (it's much easier than Tom thought :-).
3804 (5) Because of (4), we can now allow for the use of $prvscheck_result
3805 inside the third argument.
3807 PH/03 For some reason, the default setting of PATH when running a command from
3808 a pipe transport was just "/usr/bin". I have changed it to
3811 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3812 anything to be listed in the output from -bV.
3814 PH/05 When a filter generated an autoreply, the entire To: header line was
3815 quoted in the delivery log line, like this:
3817 => >A.N.Other <ano@some.domain> <original@ddress> ...
3819 This has been changed so that it extracts the operative address. There
3820 may be more than one such address. If so, they are comma-separated, like
3823 => >ano@some.domain,ona@other.domain <original@ddress> ...
3825 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3826 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3827 not being looked up in the rDNS to get a host name, Exim was showing the
3828 IP address twice in Received: lines, even though the IP addresses were
3829 identical. For example:
3831 Received: from [1.2.3.4] (helo=[1.2.3.4])
3833 However, if the real host name was known, it was omitting the HELO data
3834 if it matched the actual IP address. This has been tidied up so that it
3835 doesn't show the same IP address twice.
3837 PH/07 When both +timestamp and +memory debugging was on, the value given by
3838 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3839 called by the debug printing, thereby overwriting the timestamp buffer.
3840 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3842 PH/08 When the original message was included in an autoreply transport, it
3843 always said "this is a copy of the message, including all the headers",
3844 even if body_only or headers_only was set. It now gives an appropriate
3847 PH/09 Applied a patch from the Sieve maintainer which:
3849 o fixes some comments
3850 o adds the (disabled) notify extension core
3851 o adds some debug output for the result of if/elsif tests
3852 o points to the current vacation draft in the documentation
3853 and documents the missing references header update
3857 o fixes a bug in processing the envelope test (when testing
3858 multiple envelope elements, the last element determined the
3861 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3862 Electronic Mail") by including:
3864 Auto-submitted: auto-generated
3866 in the messages that it generates (bounce messages and others, such as
3867 warnings). In the case of bounce messages for non-SMTP messages, there was
3868 also a typo: it was using "Auto_submitted" (underscore instead of
3869 hyphen). Since every message generated by Exim is necessarily in response
3870 to another message, thes have all been changed to:
3872 Auto-Submitted: auto-replied
3874 in accordance with these statements in the RFC:
3876 The auto-replied keyword:
3878 - SHOULD be used on messages sent in direct response to another
3879 message by an automatic process,
3881 - MUST NOT be used on manually-generated messages,
3883 - MAY be used on Delivery Status Notifications (DSNs) and Message
3884 Disposition Notifications (MDNs),
3886 - MUST NOT be used on messages generated by automatic or periodic
3887 processes, except for messages which are automatic responses to
3890 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3891 to the default Received: header definition.
3893 PH/12 Added log selector acl_warn_skipped (default on).
3895 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3896 variables because (a) they are in the wrong storage pool and (b) even if
3897 they were copied, it wouldn't work properly because of the caching.
3899 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3900 checking when decoding. Apparently there are clients that generate
3901 overlong encoded strings. Why am I not surprised?
3903 PH/15 If the first argument of "${if match_address" was not empty, but did not
3904 contain an "@" character, Exim crashed. Now it writes a panic log message
3905 and treats the condition as false.
3907 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3909 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3910 extension "envelope-auth". The code is finished and in agreement with
3911 other implementations, but there is no documentation so far and in fact,
3912 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3913 not changing the active code.
3915 Print executed "if" and "elsif" statements when debugging is used. This
3916 helps a great deal to understand what a filter does.
3918 Document more things not specified clearly in RFC3028. I had all this
3919 sorted out, when out of a sudden new issues came to my mind. Oops."
3921 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3924 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3925 canonical form (as documented). However, after a host name lookup from
3926 the IP address, check_host() was doing a simple string comparison with
3927 addresses acquired from the DNS when checking that the found name did
3928 have the original IP as one of its addresses. Since any found IPv6
3929 addresses are likely to be in abbreviated form, the comparison could
3930 fail. Luckily, there already exists a function for doing the comparison
3931 by converting both addresses to binary, so now that is used instead of
3932 the text comparison.
3934 PH/20 There was another similar case to PH/19, when a complete host name was
3935 given in a host list; looking up its IP address could give an abbreviated
3936 form, whereas the current host's name might or might not be abbreviated.
3937 The same fix has been applied.
3943 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3944 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3947 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3948 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3950 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3952 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3953 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3954 Also, for "Auto-Submitted", treat anything other than "no" as
3955 non-personal, in accordance with RFC 3834. (Previously it treated
3956 anything starting "auto-" as non-personal.)
3958 TF/01 The control=submission/name=... option had a problem with syntax
3959 errors if the name included a slash character. The /name= option
3960 now slurps the rest of the string, so it can include any characters
3961 but it must come last in the list of options (after /sender_retain
3964 PH/05 Some modifications to the interface to the fake nameserver for the new
3972 TK/01 Added the "success_on_redirect" address verification option. See
3973 NewStuff for rationale and an example.
3975 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3977 PH/02 Patch to exigrep to allow it to work on syslog lines.
3979 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3980 fread() to skip over the body file's header line, because in Cygwin the
3981 header line is locked and is inaccessible.
3983 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3984 co-exist for some time) to make it clear that it is the Exim ID that is
3985 referenced, not the Message-ID: header line.
3987 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3988 string_format() function, because snprintf() does not exist on all
3991 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3992 ${stat: expansion item.
3994 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3995 protocol synchronization error", to keep the pedants happy.
3997 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3998 well as for IRIX systems, when gcc is being used. See the host.c source
4001 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4003 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4006 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4007 transport and to the smtp transport in LMTP mode.
4009 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4011 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4012 run a filter in a subprocess. This could lead to confusion in subsequent
4013 lookups in the parent process. There should also be a search_tidyup() at
4014 the end of the subprocess.
4016 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4017 only if the host matched helo_try_verify_hosts, which caused the
4018 verification to occur when the EHLO/HELO command was issued. The ACL just
4019 tested the remembered result. Now, if a previous verification attempt has
4020 not happened, "verify = helo" does it there and then.
4022 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4024 TK/03 Fix log output including CR from clamd.
4026 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4027 exist provoked a memory error which could cause a segfault.
4029 PH/15 Installed PCRE 6.2
4031 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4033 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4034 of the problem. Specifically, suggested +O2 rather than +O1 for the
4037 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4039 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4040 record, the port was not correctly logged when the outgoing_port log
4041 selector was set (it logged the transort's default port).
4043 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4044 fallback_hosts, and "hosts" in the smtp transport.
4046 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4047 host errors such as "Connection refused".
4049 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4050 authentication with radiusclient 0.4.9:
4052 - Error returned from rc_read_config was caught wrongly
4053 - Username/password not passed on to radius server due to wrong length.
4055 The presumption is that some radiusclient API changes for 4.51/PH/17
4056 were not taken care of correctly. The code is still untested by me (my
4057 Linux distribution still has 0.3.2 of radiusclient), but it was
4058 contributed by a Radius user.
4060 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4061 expanding the "port" option of the smtp transport.
4063 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4064 while reading a MIME header. Thanks to Tom Hughes for a patch.
4066 PH/24 Include config.h inside local_scan.h so that configuration settings are
4069 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4070 that in an ACL for RCPT (for example), you can examine exactly what was
4073 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4074 commands, but it was not correctly comparing the address with the actual
4075 client host address. Thus, it would show the EHLO address in Received:
4076 header lines when this was not necessary.
4078 PH/27 Added the % operator to ${eval:}.
4080 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4081 it should be ignoring failures (because with -C, for example, it has lost
4082 privilege). It wasn't ignoring creation failures other than "already
4085 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4088 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4089 because an input line was too long, either on its own, or by virtue of
4090 too many continuations, the temporary file was not being removed, and the
4091 return code was incorrect.
4093 PH/31 Missing "BOOL" in function definition in filtertest.c.
4095 PH/32 Applied Sieve patches from the maintainer.
4097 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4099 PH/33 Added "verify = not_blind".
4101 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4102 Local/Makefile (with some defaults set). These are used in built scripts
4103 such as exicyclog, but they have never been used in the exim_install
4104 script (though there are many overriding facilities there). I have
4105 arranged that the exim_install script now takes note of these two
4108 PH/35 Installed configuration files for Dragonfly.
4110 PH/36 When a locally submitted message by a trusted user did not contain a
4111 From: header, and the sender address was obtained from -f or from an SMTP
4112 MAIL command, and the trusted user did not use -F to supply a sender
4113 name, $originator_name was incorrectly used when constructing a From:
4114 header. Furthermore, $originator_name was used for submission mode
4115 messages from external hosts without From: headers in a similar way,
4116 which is clearly wrong.
4118 PH/37 Added control=suppress_local_fixups.
4120 PH/38 When log_selector = +received_sender was set, and the addition of the
4121 sender made the log line's construction buffer exactly full, or one byte
4122 less than full, an overflow happened when the terminating "\n" was
4125 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4126 when the result of a list match is failure because a DNS lookup failed.
4128 PH/40 RM_COMMAND is now used in the building process.
4130 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4131 the "build-* directories that it finds.
4133 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4134 address in a domain literal was a prefix of an interface address.
4136 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4137 when verifying a sender address, unless rewrite_headers is false.
4139 PH/44 Wrote a long comment about why errors_to addresses are verified as
4140 recipients, not senders.
4142 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4143 the ratelimit ACL was added.
4145 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4147 PH/46 Added extra information about PostgreSQL errors to the error string.
4149 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4150 avoiding the need to install special zones in a real server. This is
4151 backwards compatible; if it can't find the fake resolver, it drops back.
4152 Thus, both old and new test suites can be run.
4154 TF/02 Added util/ratelimit.pl
4156 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4157 clock is set back in time.
4159 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4160 Candler <B.Candler@pobox.com>.
4162 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4163 OK for addresses that are the result of redirections.
4165 PH/48 A number of further additions for the benefit of the new test suite,
4166 including a fake gethostbyname() that interfaces to the fake DNS resolver
4169 TF/06 The fix for widen_domains has also been applied to qualify_single and
4170 search_parents which are the other dnslookup options that can cause
4173 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4176 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4178 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4179 service" if it encountered an MX record with an empty target hostname.
4180 The message is now "an MX or SRV record indicated no SMTP service".
4182 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4183 if the DNS of the sending site is misconfigured. This is quite a
4184 common situation. This change restores the behaviour of treating a
4185 helo verification defer as a failure.
4187 PH/51 If self=fail was set on a router, the bounce message did not include the
4188 actual error message.
4194 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4196 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4197 command itself ended in error, the underlying message about the transport
4198 filter timeout was being overwritten with the pipe command error. Now the
4199 underlying error message should be appended to the second error message.
4201 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4203 PH/02 Reduce the amount of output that "make" produces by default. Full output
4204 can still be requested.
4206 PH/03 The warning log line about a condition test deferring for a "warn" verb
4207 was being output only once per connection, rather than after each
4208 occurrence (because it was using the same function as for successful
4209 "warn" verbs). This seems wrong, so I have changed it.
4211 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4212 it should not have, which might have caused a crash in the right
4213 circumstances, but probably never did.
4215 PH/04 Installed a modified version of Tony Finch's patch to make submission
4216 mode fix the return path as well as the Sender: header line, and to
4217 add a /name= option so that you can make the user's friendly name appear
4220 TF/03 Added the control = fakedefer ACL modifier.
4222 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4223 Mark Lowes for thorough testing.
4225 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4227 TK/03 Merged latest SRS patch from Miles Wilton.
4229 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4230 with the definition in sysexits.h (which is #included earlier).
4231 Fortunately, Exim does not actually use EX_OK. The code used to try to
4232 preserve the sysexits.h value, by assuming that macro definitions were
4233 scanned for macro replacements. I have been disabused of this notion,
4234 so now the code just undefines EX_OK before #including unistd.h.
4236 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4237 in the smtp transport. When a block could not be written in a single
4238 write() function, the timeout was being re-applied to each part-write.
4239 This seems wrong - if the receiver was accepting one byte at a time it
4240 would take for ever. The timeout is now adjusted when this happens. It
4241 doesn't have to be particularly precise.
4243 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4244 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4246 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4247 without a domain if the check to <postmaster@domain> fails.
4249 SC/01 Eximstats: added -xls and the ability to specify output files
4250 (patch written by Frank Heydlauf).
4252 SC/02 Eximstats: use FileHandles for outputting results.
4254 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4256 SC/04 Eximstats: fixed display of large numbers with -nvr option
4258 SC/05 Eximstats: fixed merging of reports with empty tables.
4260 SC/06 Eximstats: added the -include_original_destination flag
4262 SC/07 Eximstats: removed tabs and trailing whitespace.
4264 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4266 TK/06 MBOX spool code: Add real "From " MBOX separator line
4267 so the .eml file is really in mbox format (even though
4268 most programs do not really care). Patch from Alex Miller.
4270 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4271 The latter is generated from $received_to and is only set if the
4272 message has one envelope recipient. SA can use these headers,
4273 obviously out-of-the-box. Patch from Alex Miller.
4275 PH/08 The ${def test on a variable was returning false if the variable's
4276 value was "0", contrary to what the specification has always said!
4277 The result should be true unless the variable is empty.
4279 PH/09 The syntax error of a character other than { following "${if
4280 def:variable_name" (after optional whitespace) was not being diagnosed.
4281 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4282 accidental colon was present, for example, could give incorrect results.
4284 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4285 result is used (not including appendfile, where other changes are about
4288 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4289 This involved changing a lot of size variables from int to off_t. It
4290 should work with maildirs and everything.
4292 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4293 spamd dying while we are connected to it.
4295 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4298 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4299 function for BDB 4.3.
4301 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4303 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4304 the sources for PCRE have been re-arranged and more files are now
4307 PH/15 The code I had for printing potentially long long variables in PH/11
4308 above was not the best (it lost precision). The length of off_t variables
4309 is now inspected at build time, and an appropriate printing format (%ld
4310 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4311 to be "long long int" or "long int". This is needed for the internal
4312 formatting function string_vformat().
4314 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4315 the configuration file to be ":syslog", then the script "guesses" where
4316 the logs files are, rather than using the compiled in default. In our
4317 case the guess is not the same as the compiled default, so the script
4318 suddenly stopped working when I started to use syslog. The patch checks
4319 to see if log_file_path is "". If so, it attempts to read it from exim
4320 with no configuration file to get the compiled in version, before it
4321 falls back to the previous guessing code."
4323 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4324 implementing BATV in an Exim configuration. See NewStuff for the gory
4327 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4328 Makefile that are specific to HP-UX.
4330 PH/18 If the "use_postmaster" option was set for a recipient callout together
4331 with the "random" option, the postmaster address was used as the MAIL
4332 FROM address for the random test, but not for the subsequent recipient
4333 test. It is now used for both.
4335 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4336 patch removes a few documentation additions to RFC 3028, because the
4337 latest draft now contains them. It adds the new en;ascii-case comparator
4338 and a new error check for 8bit text in MIME parts. Comparator and
4339 require names are now matched exactly. I enabled the subaddress
4340 extension, but it is not well tested yet (read: it works for me)."
4342 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4343 rework some of the code of TK/09 above to avoid the hardwired use of
4344 "%lld" and "long long". Replaced the call to snprintf() with a call to
4347 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4348 records point to non-existent hosts", "retry timeout exceeded", and
4349 "retry time not reached for any host after a long failure period".
4351 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4352 experimental DomainKeys support:
4354 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4355 (2) On an error such as an illegally used "control", the wrong name for
4356 the control was given.
4358 These problems did NOT occur unless DomainKeys support was compiled.
4360 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4362 PH/24 Added ${if match_ip condition.
4364 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4365 databases so that it will be absolutely obvious if a crash occurs in the
4366 DB library. This is a regular occurrence (often caused by mis-matched
4369 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4370 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4371 picked up on a user's system that detects such things. There doesn't seem
4372 to be a gcc warning option for this - only an attribute that has to be
4373 put on the function's prototype. It seems that in Fedora Core 4 they have
4374 set this on a number of new functions. No doubt there will be more in due
4377 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4378 specify a transport. However, if an address that was verified by such a
4379 router was the subject of a callout, Exim crashed because it tried to
4380 read the rcpt_include_affixes from the non-existent transport. Now it
4381 just assumes that the setting of that option is false. This bug was
4382 introduced by 4.51/PH/31.
4384 PH/28 Changed -d+all to exclude +memory, because that information is very
4385 rarely of interest, but it makes the output a lot bigger. People tend to
4386 do -d+all out of habit.
4388 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4389 code in os-type was giving problems when libc.so lives in lib64, like on
4392 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4393 aren't the modern standard, and it seems that some systems' include files
4394 don't always have them. Exim was already checking for some of the newer
4395 ones like T_AAAA, and defining it itself. I've added checks for all the
4396 record types that Exim uses.
4398 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4399 not automatically generating a new one, as it is supposed to. This
4400 prevented TLS from working. If the file did exist, but contained invalid
4401 data, a new version was generated, as expected. It was only the case of a
4402 non-existent file that was broken.
4404 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4405 with a change in libdomainkeys > 0.64.
4407 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4408 from DNS. If the selector record carries the flag, it now has
4409 precedence over the domain-wide flag.
4411 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4413 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4414 the use of a port name that isn't defined in /etc/services) occurred, the
4415 message was deferred as in a normal delivery, and thus remained on the
4416 spool, instead of being failed because of the mua_wrapper setting. This
4417 is now fixed, and I tidied up some of the mua_wrapper messages at the
4420 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4421 the messages in a hash of arrays rather than using individual hashes.
4422 This is a bit cleaner and results in dramatic memory savings, albeit
4423 at a slight CPU cost.
4425 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4426 as requested by Marc Sherman.
4428 SC/10 Eximstats: added histograms for user specified patterns as requested
4431 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4433 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4434 fopen() in the content-scanning modules that did not already have it.
4440 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4441 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4443 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4445 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4447 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4448 to test Sieve filters that use "vacation".
4450 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4451 that changes the way the GnuTLS parameters are stored in the cache file.
4452 The new format can be generated externally. For backward compatibility,
4453 if the data in the cache doesn't make sense, Exim assumes it has read an
4454 old-format file, and it generates new data and writes a new file. This
4455 means that you can't go back to an older release without removing the
4458 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4459 work if there are any delivery delays because "one_time" forces the
4460 parent to be marked "delivered", so its unseen clone is never tried
4461 again. For this reason, Exim now forbids the simultaneous setting of
4464 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4465 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4466 entry if you want to know the details. The fix, however, neglected to
4467 consider the case when local delivery batching is involved. The test for
4468 "previously delivered" was not happening when checking to see if an
4469 address could be batched with a previous (undelivered) one; under
4470 certain circumstances this could lead to multiple deliveries to the same
4473 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4474 in its include files, and this causes problems building Exim.
4476 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4477 header_syntax) but Exim was just ignoring anything given after a slash.
4478 In particular, this caused confusion with an attempt to use "verify =
4479 reverse_host_lookup/defer_ok". An error is now given when options are
4480 supplied for verify items that do not have them. (Maybe reverse_host_
4481 lookup should have a defer_ok option, but that's a different point.)
4483 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4484 defined by RFC 821) to 2048, because there were problems with some AUTH
4485 commands, and RFC 1869 says the size should be increased for extended
4486 SMTP commands that take arguments.
4488 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4491 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4492 "unknown" error; now it says that the functionality isn't in the binary.
4494 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4495 an address' error message when a string expansion fails (syntax or
4496 whatever). Otherwise the password may appear in the log. Following change
4497 PH/42 below, there is no longer a chance of it appearing in a bounce
4500 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4502 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4503 of its addresses was ever tried. (Bugzilla bug #2.)
4505 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4506 the result incorrectly in the debug output. (It correctly added a newline
4507 to what was transported.)
4509 TF/01 Added $received_time.
4511 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4512 commented out examples of how to interface to a virus scanner and to
4513 SpamAssassin. Also added commented examples of av_scanner and
4514 spamd_address settings.
4516 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4517 and controls are allowed in which ACLs. There were a couple of minor
4518 errors. Some of the entries in the conditions table (which is a table of
4519 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4520 negation of where the condition IS allowed.
4522 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4524 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4525 header file does not have a version number, so I've had to invent a new
4526 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4527 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4528 radiusclient), but it was contributed by a Radius user.
4530 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4531 files or directories, for OpenSSL.
4533 PH/19 When an Exim process that is running as root has to create an Exim log
4534 file, it does so in a subprocess that runs as exim:exim so as to get the
4535 ownership right at creation (otherwise, other Exim processes might see
4536 the file with the wrong ownership). There was no test for failure of this
4537 fork() call, which would lead to the process getting stuck as it waited
4538 for a non-existent subprocess. Forks do occasionally fail when resources
4539 run out. I reviewed all the other calls to fork(); they all seem to check
4542 PH/20 When checking for unexpected SMTP input at connect time (before writing
4543 the banner), Exim was not dealing correctly with a non-positive return
4544 from the read() function. If the client had disconnected by this time,
4545 the result was a log entry for a synchronization error with an empty
4546 string after "input=" when read() returned zero. If read() returned -1
4547 (an event I could not check), uninitialized data bytes were printed.
4548 There were reports of junk text (parts of files, etc) appearing after
4551 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4553 PH/22 Added support for macro redefinition, and (re)definition in between
4554 driver and ACL definitions.
4556 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4557 forgetting to use the resulting value; it was using the unexpanded value.
4559 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4560 hadn't been configured. The fix is from Juergen Kreileder, who
4561 understands it better than I do:
4563 "Here's what I see happening with three configured cyrus_sasl
4564 authenticators configured (plain, login, cram-md5):
4566 On startup auth_cyrus_sasl_init() gets called for each of these.
4567 This means three calls to sasl_listmech() without a specified mech_list.
4568 => SASL tests which mechs of all available mechs actually work
4569 => three warnings about OTP not working
4570 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4572 With the patch, sasl_listmech() also gets called three times. But now
4573 SASL's mech_list option is set to the server_mech specified in the the
4574 authenticator. Or in other words, the answer from sasl_listmech()
4575 gets limited to just the mech you're testing for (which is different
4577 => the return list contains just 'plain' or 'login', 'cram-md5' or
4578 nothing depending on the value of ob->server_mech.
4580 I've just tested the patch: Authentication still works fine,
4581 unavailable mechs specified in the exim configuration are still
4582 caught, and the auth.log warnings about OTP are gone."
4584 PH/25 When debugging is enabled, the contents of the command line are added
4585 to the debugging output, even when log_selector=+arguments is not
4588 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4589 answer is "GNU", and only if the return is "GNU/something" is the answer
4592 PH/27 $acl_verify_message is now set immediately after the failure of a
4593 verification in an ACL, and so is available in subsequent modifiers. In
4594 particular, the message can be preserved by coding like this:
4596 warn !verify = sender
4597 set acl_m0 = $acl_verify_message
4599 Previously, $acl_verify_message was set only while expanding "message"
4600 and "log_message" when a very denied access.
4602 PH/28 Modified OS/os.c-Linux with
4604 -#ifndef OS_LOAD_AVERAGE
4605 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4607 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4608 nomenclature these days.)
4610 PH/29 Installed patch from the Sieve maintainer that adds the options
4611 sieve_useraddress and sieve_subaddress to the redirect router.
4613 PH/30 In these circumstances:
4614 . Two addresses routed to the same list of hosts;
4615 . First host does not offer TLS;
4616 . First host accepts first address;
4617 . First host gives temporary error to second address;
4618 . Second host offers TLS and a TLS session is established;
4619 . Second host accepts second address.
4620 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4621 and peerdn, if requested) that were in fact used only for the second
4624 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4625 attention to any local part prefix or suffix that was matched by the
4626 router that accepted the address. It now behaves in the same way as it
4627 does for delivery: the affixes are removed from the local part unless
4628 rcpt_include_affixes is set on the transport.
4630 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4631 timeout during the DATA phase of an incoming message.
4633 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4634 applied a patch sanctioned by the Sieve maintainer.
4636 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4637 the uid or gid is negative. A case of a negative gid caused this to be
4638 noticed. The fix allows for either to be negative.
4640 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4641 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4644 PH/36 The change PH/12 above was broken. Fixed it.
4646 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4647 the grounds that routing the same address twice would always produce the
4648 same answer. This might have been true once, but it is certainly no
4649 longer true now. Routing a child address may depend on the previous
4650 routing that produced that child. Some complicated redirection strategies
4651 went wrong when messages had multiple recipients, and made Exim's
4652 behaviour dependent on the order in which the addresses were given.
4654 I have moved the duplicate checking until after the routing is complete.
4655 Exim scans the addresses that are assigned to local and remote
4656 transports, and removes any duplicates. This means that more work will be
4657 done, as duplicates will always all be routed, but duplicates are
4658 presumably rare, so I don't expect this is of any significance.
4660 For deliveries to pipes, files, and autoreplies, the duplicate checking
4661 still happens during the routing process, since they are not going to be
4664 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4665 It corrects a timeout issue with spamd. This is Ian's comment: "The
4666 background is that sometimes spamd either never reads data from a
4667 connection it has accepted, or it never writes response data. The exiscan
4668 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4669 blindly assumes that writes won't block so it may never time out."
4671 PH/39 Allow G after quota size as well as K and M.
4673 PH/40 The value set for $authenticated_id in an authenticator may not contain
4674 binary zeroes or newlines because the value is written to log lines and
4675 to spool files. There was no check on this. Now the value is run through
4676 the string_printing() function so that such characters are converted to
4677 printable escape sequences.
4679 PH/41 $message_linecount is a new variable that contains the total number of
4680 lines in the message. Compare $body_linecount, which is the count for the
4683 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4684 bounce and delay warning messages, except in certain special cases, which
4687 (a) An SMTP error message from a remote host;
4688 (b) A message specified in a :fail: redirection;
4689 (c) A message specified in a "fail" command in a system filter;
4690 (d) A message specified in a FAIL return from the queryprogram router;
4691 (e) A message specified by the cannot_route_message router option.
4693 In these cases only, Exim does include the error details in bounce and
4694 warning messages. There are also a few cases where bland messages such
4695 as "unrouteable address" or "local delivery error" are given.
4697 PH/43 $value is now also set for the "else" part of a ${run expansion.
4699 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4700 being worked on, but at least Exim now implements the latest version to
4703 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4704 process to complete was treated as a delivery failure, a timeout while
4705 writing the message to the pipe was logged, but erroneously treated as a
4706 successful delivery. Such timeouts include transport filter timeouts. For
4707 consistency with the overall process timeout, these timeouts are now
4708 treated as errors, giving rise to delivery failures by default. However,
4709 there is now a new Boolean option for the pipe transport called
4710 timeout_defer, which, if set TRUE, converts the failures into defers for
4711 both kinds of timeout. A transport filter timeout is now identified in
4714 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4715 systems where "make" and "gmake" are different, calling "gmake" at top
4716 level broke things. I've arranged for the value of $(MAKE) to be passed
4717 from the Makefile to this script so that it can call the same version of
4721 A note about Exim versions 4.44 and 4.50
4722 ----------------------------------------
4724 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4725 changes of various kinds. As a consequence, a big documentation update was
4726 needed. This delayed the release for rather longer than seemed good, especially
4727 in the light of a couple of (minor) security issues. Therefore, the changes
4728 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4729 release. So 4.44 and 4.50 are in effect two different branches that both start
4732 I have left the 4.50 change log unchanged; it contains all the changes since
4733 4.43. The change log for 4.44 is below; many of its items are identical to
4734 those for 4.50. This seems to be the most sensible way to preserve the
4735 historical information.
4741 1. Minor wording change to the doc/README.SIEVE file.
4743 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4744 computation of the current number of files was incorrect.
4746 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4747 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4750 4. Give more explanation in the error message when the command for a transport
4751 filter fails to execute.
4753 5. There are several places where Exim runs a non-Exim command in a
4754 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4755 was being done only for the command run by the queryprogram router. It is
4756 now done for all such subprocesses. The other cases are: ${run, transport
4757 filters, and the commands run by the lmtp and pipe transports.
4759 6. Added CONFIGURE_GROUP build-time option.
4761 7. Some older OS have a limit of 256 on the maximum number of file
4762 descriptors. Exim was using setrlimit() to set 1000 as a large value
4763 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4764 systems. I've change it so that if it can't get 1000, it tries for 256.
4766 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4767 was an oversight, and furthermore, ever since the addition of extra
4768 controls (e.g. 4.43/32), the checks on when to allow different forms of
4769 "control" were broken. There should now be diagnostics for all cases when a
4770 control that does not make sense is encountered.
4772 9. Added the /retain_sender option to "control=submission".
4774 10. $recipients is now available in the predata ACL (oversight).
4776 11. Tidy the search cache before the fork to do a delivery from a message
4777 received from the command line. Otherwise the child will trigger a lookup
4778 failure and thereby defer the delivery if it tries to use (for example) a
4779 cached ldap connection that the parent has called unbind on.
4781 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4782 of $address_data from the recipient verification was clobbered by the
4783 sender verification.
4785 13. The value of address_data from a sender verification is now available in
4786 $sender_address_data in subsequent conditions in the ACL statement.
4788 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4790 15. Added a new option "connect=<time>" to callout options, to set a different
4793 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4794 was its contents. (It was OK if the option was not defined at all.)
4796 17. A "Completed" log line is now written for messages that are removed from
4797 the spool by the -Mrm option.
4799 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4800 information about exactly what failed.
4802 19. Added -dd to debug only the daemon process.
4804 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4805 handles timeouts, both on the server side and network timeouts. Renamed the
4806 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4808 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4809 It is now set to "smtps".
4811 22. $host_address is now set to the target address during the checking of
4812 ignore_target_hosts.
4814 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4815 being passed; this would have caused $sender_host_name to have been used if
4816 matching the list had actually called for a host name (not very likely,
4817 since this list is usually IP addresses). A host name is now passed as
4820 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4821 code that specifies a non-restarting handler (typically sigaction() in
4822 modern systems) in an attempt to fix a rare and obscure crash bug.
4824 25. Narrowed the window for a race in the daemon that could cause it to ignore
4825 SIGCHLD signals. This is not a major problem, because they are used only to
4826 wake it up if nothing else does.
4828 26. A malformed maildirsize file could cause Exim to calculate negative values
4829 for the mailbox size or file count. Odd effects could occur as a result.
4830 The maildirsize information is now recalculated if the size or filecount
4833 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4834 support for a long time. Removed HAVE_SYS_VFS_H.
4836 28. Installed the latest version of exipick from John Jetmore.
4838 29. In an address list, if the pattern was not a regular expression, an empty
4839 subject address (from a bounce message) matched only if the pattern was an
4840 empty string. Non-empty patterns were not even tested. This was the wrong
4841 because it is perfectly reasonable to use an empty address as part of a
4842 database query. An empty address is now tested by patterns that are
4843 lookups. However, all the other forms of pattern expect the subject to
4844 contain a local part and a domain, and therefore, for them, an empty
4845 address still always fails if the pattern is not itself empty.
4847 30. Exim went into a mad DNS loop when attempting to do a callout where the
4848 host was specified on an smtp transport, and looking it up yielded more
4849 than one IP address.
4851 31. Re-factored the code for checking spool and log partition space into a
4852 function that finds that data and another that does the check. The former
4853 is then used to implement four new variables: $spool_space, $log_space,
4854 $spool_inodes, and $log_inodes.
4856 32. The RFC2047 encoding function was originally intended for short strings
4857 such as real names; it was not keeping to the 75-character limit for
4858 encoded words that the RFC imposes. It now respects the limit, and
4859 generates multiple encoded words if necessary. To be on the safe side, I
4860 have increased the buffer size for the ${rfc2047: expansion operator from
4863 33. It is now permitted to omit both strings after an "if" condition; if the
4864 condition is true, the result is "true". As before, when the second string
4865 is omitted, a false condition yields an empty string. This makes it less
4866 cumbersome to write custom ACL and router conditions.
4868 34. Failure to deliver a bounce message always caused it to be frozen, even if
4869 there was an errors_to setting on the router. The errors_to setting is now
4872 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4873 canonical form (fully expanded) before being placed in
4874 $sender_host_address.
4876 36. The table in the code that translates DNS record types into text (T_A to
4877 "A" for instance) was missing entries for NS and CNAME. It is just possible
4878 that this could have caused confusion if both these types were looked up
4879 for the same domain, because the text type is used as part of Exim's
4880 per-process caching. But the chance of anyone hitting this buglet seems
4883 37. The dnsdb lookup has been extended in a number of ways.
4885 (1) There is a new type, "zns", which walks up the domain tree until it
4886 finds some nameserver records. It should be used with care.
4888 (2) There is a new type, "mxh", which is like "mx" except that it returns
4889 just the host names, not the priorities.
4891 (3) It is now possible to give a list of domains (or IP addresses) to be
4892 looked up. The behaviour when one of the lookups defers can be
4893 controlled by a keyword.
4895 (4) It is now possible to specify the separator character for use when
4896 multiple records are returned.
4898 38. The dnslists ACL condition has been extended: it is now possible to supply
4899 a list of IP addresses and/or domains to be looked up in a particular DNS
4902 39. Added log_selector=+queue_time_overall.
4904 40. When running the queue in the test harness, wait just a tad after forking a
4905 delivery process, to get repeatability of debugging output.
4907 41. Include certificate and key file names in error message when GnuTLS fails
4908 to set them up, because the GnuTLS error message doesn't include the name
4909 of the failing file when there is a problem reading it.
4911 42. Allow both -bf and -bF in the same test run.
4913 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4915 44. The "Exiscan patch" is now merged into the mainline Exim source.
4917 45. Sometimes the final signoff response after QUIT could fail to get
4918 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4919 < 0 before doing a fflush(). This bug looks as though it goes back to the
4920 introduction of TLS in release 3.20, but "sometimes" must have been rare
4921 because the tests only now provoked it.
4923 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4924 (this can affect the format of dates).
4926 47. exim_tidydb, when checking for the continued existence of a message for
4927 which it has found a message-specific retry record, was not finding
4928 messages that were in split spool directories. Consequently, it was
4929 deleting retry records that should have stayed in existence.
4931 48. Steve fixed some bugs in eximstats.
4933 49. The SPA authentication driver was not abandoning authentication and moving
4934 on to the next authenticator when an expansion was forced to fail,
4935 contradicting the general specification for all authenticators. Instead it
4936 was generating a temporary error. It now behaves as specified.
4938 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4939 (the order specifies the preference for clients). The order is now AES256,
4940 AES128, 3DES, ARCFOUR128.
4942 51. Small patch to Sieve code - explicitly set From: when generating an
4945 52. Exim crashed if a remote delivery caused a very long error message to be
4946 recorded - for instance if somebody sent an entire SpamAssassin report back
4947 as a large number of 550 error lines. This bug was coincidentally fixed by
4948 increasing the size of one of Exim's internal buffers (big_buffer) that
4949 happened as part of the Exiscan merge. However, to be on the safe side, I
4950 have made the code more robust (and fixed the comments that describe what
4953 53. Now that there can be additional text after "Completed" in log lines (if
4954 the queue_time_overall log selector is set), a one-byte patch to exigrep
4955 was needed to allow it to recognize "Completed" as not the last thing in
4958 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4959 patch that reportedly fixes this has been added. I am not expert enough to
4960 create a test for it. This is what the patch creator wrote:
4962 "I found a little strange behaviour of ldap code when working with
4963 Windows 2003 AD Domain, where users was placed in more than one
4964 Organization Units. When I tried to give exim partial DN, the exit code
4965 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4966 But simultaneously result of request was absolutely normal ldap result,
4967 so I produce this patch..."
4969 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4970 REFERENCE, so I have modified the code to exclude the patch when that macro
4973 55. Some experimental protocols are using DNS PTR records for new purposes. The
4974 keys for these records are domain names, not reversed IP addresses. The
4975 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4976 leaves it alone. Component reversal etc. now happens only for IP addresses.
4979 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4981 57. Double the size of the debug message buffer (to 2048) so that more of very
4982 long debug lines gets shown.
4984 58. The exicyclog utility now does better if the number of log files to keep
4985 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4987 59. Two changes related to the smtp_active_hostname option:
4989 (1) $smtp_active_hostname is now available as a variable.
4990 (2) The default for smtp_banner uses $smtp_active_hostname instead
4991 of $primary_hostname.
4993 60. The host_aton() function is supposed to be passed a string that is known
4994 to be a valid IP address. However, in the case of IPv6 addresses, it was
4995 not checking this. This is a hostage to fortune. Exim now panics and dies
4996 if the condition is not met. A case was found where this could be provoked
4997 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4998 components; fortuitously, this particular loophole had already been fixed
4999 by change 4.50/55 above.
5001 If there are any other similar loopholes, the new check in host_aton()
5002 itself should stop them being exploited. The report I received stated that
5003 data on the command line could provoke the exploit when Exim was running as
5004 exim, but did not say which command line option was involved. All I could
5005 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5006 running as the user.
5009 61. There was a buffer overflow vulnerability in the SPA authentication code
5010 (which came originally from the Samba project). I have added a test to the
5011 spa_base64_to_bits() function which I hope fixes it.
5014 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5015 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5017 63. The daemon start-up calls getloadavg() while still root for those OS that
5018 need the first call to be done as root, but it missed one case: when
5019 deliver_queue_load_max is set with deliver_drop_privilege. This is
5020 necessary for the benefit of the queue runner, because there is no re-exec
5021 when deliver_drop_privilege is set.
5023 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5024 This has been fixed.
5026 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5027 was in use, was not putting the data itself into the right store pool;
5028 consequently, it could be overwritten for a subsequent message in the same
5029 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5032 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5034 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5035 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5036 the function were treating the return as a boolean value, which happened to
5037 work because 0=false and not-0=true, but is not correct code.
5039 68. The host_aton() function was not handling scoped IPv6 addresses (those
5040 with, for example, "%eth0" on the end) correctly.
5042 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5043 negated items (that is, ~something) in unsigned ints. Some compilers
5044 apparently mutter when there is no cast.
5046 70. If an address verification called from an ACL failed, and did not produce a
5047 user-specific message (i.e. there was only a "system" message), nothing was
5048 put in $acl_verify_message. In this situation, it now puts the system
5051 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5052 session; change 4.31/43 added the unwanted input to the log line - except
5053 that it did not do this in the start of session case. It now does.
5055 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5056 This is wrong and can cause the other end to generate a synchronization
5057 error if it is another Exim or anything else that does the synchronization
5058 check. A QUIT command is no longer sent after a timeout.
5060 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5061 during host lookups.
5063 74. The defer_ok option of callout verification was not working if it was used
5064 when verifying addresses in header lines, that is, for this case:
5066 verify = header_sender/callout=defer_ok
5068 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5069 those file descriptors could be used for SMTP connections. If anything
5070 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5071 could be sent to the SMTP client, causing chaos. The daemon now opens
5072 stdin, stdout, and stderr to /dev/null when it puts itself into the
5075 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5076 log by default. The user can override this with suitable Perl magic.
5078 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5079 the log message when discard triggers, was not working for the DATA ACL or
5080 for the non-SMTP ACL.
5082 78. Error message wording change in sieve.c.
5084 79. If smtp_accept_max_per_host was set, the number of connections could be
5085 restricted to fewer than expected, because the daemon was trying to set up
5086 a new connection before checking whether the processes handling previous
5087 connections had finished. The check for completed processes is now done
5088 earlier. On busy systems, this bug wouldn't be noticed because something
5089 else would have woken the daemon, and it would have reaped the completed
5092 80. If a message was submitted locally by a user whose login name contained one
5093 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5094 It caused a spool format error. I have fixed the spool reading code. A
5095 related problem was that the "from" clause in the Received: line became
5096 illegal because of the space(s). It is now covered by ${quote_local_part.
5098 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5101 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5103 83. Patch from Sieve maintainer to fix unterminated string problem in
5104 "vacation" handling.
5106 84. Some minor changes to the Linux configuration files to help with other
5107 OS variants using glibc.
5109 85. One more patch for Sieve to update vacation handling to latest spec.
5112 ----------------------------------------------------
5113 See the note above about the 4.44 and 4.50 releases.
5114 ----------------------------------------------------
5120 1. Change 4.43/35 introduced a bug that caused file counts to be
5121 incorrectly computed when quota_filecount was set in an appendfile
5124 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5125 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5128 3. Give more explanation in the error message when the command for a transport
5129 filter fails to execute.
5131 4. There are several places where Exim runs a non-Exim command in a
5132 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5133 was being done only for the command run by the queryprogram router. It is
5134 now done for all such subprocesses. The other cases are: ${run, transport
5135 filters, and the commands run by the lmtp and pipe transports.
5137 5. Some older OS have a limit of 256 on the maximum number of file
5138 descriptors. Exim was using setrlimit() to set 1000 as a large value
5139 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5140 systems. I've change it so that if it can't get 1000, it tries for 256.
5142 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5143 was an oversight, and furthermore, ever since the addition of extra
5144 controls (e.g. 4.43/32), the checks on when to allow different forms of
5145 "control" were broken. There should now be diagnostics for all cases when a
5146 control that does not make sense is encountered.
5148 7. $recipients is now available in the predata ACL (oversight).
5150 8. Tidy the search cache before the fork to do a delivery from a message
5151 received from the command line. Otherwise the child will trigger a lookup
5152 failure and thereby defer the delivery if it tries to use (for example) a
5153 cached ldap connection that the parent has called unbind on.
5155 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5156 of $address_data from the recipient verification was clobbered by the
5157 sender verification.
5159 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5160 was its contents. (It was OK if the option was not defined at all.)
5162 11. A "Completed" log line is now written for messages that are removed from
5163 the spool by the -Mrm option.
5165 12. $host_address is now set to the target address during the checking of
5166 ignore_target_hosts.
5168 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5169 being passed; this would have caused $sender_host_name to have been used if
5170 matching the list had actually called for a host name (not very likely,
5171 since this list is usually IP addresses). A host name is now passed as
5174 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5175 code that specifies a non-restarting handler (typically sigaction() in
5176 modern systems) in an attempt to fix a rare and obscure crash bug.
5178 15. Narrowed the window for a race in the daemon that could cause it to ignore
5179 SIGCHLD signals. This is not a major problem, because they are used only to
5180 wake it up if nothing else does.
5182 16. A malformed maildirsize file could cause Exim to calculate negative values
5183 for the mailbox size or file count. Odd effects could occur as a result.
5184 The maildirsize information is now recalculated if the size or filecount
5187 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5188 support for a long time. Removed HAVE_SYS_VFS_H.
5190 18. Updated exipick to current release from John Jetmore.
5192 19. Allow an empty sender to be matched against a lookup in an address list.
5193 Previously the only cases considered were a regular expression, or an
5196 20. Exim went into a mad DNS lookup loop when doing a callout where the
5197 host was specified on the transport, if the DNS lookup yielded more than
5200 21. The RFC2047 encoding function was originally intended for short strings
5201 such as real names; it was not keeping to the 75-character limit for
5202 encoded words that the RFC imposes. It now respects the limit, and
5203 generates multiple encoded words if necessary. To be on the safe side, I
5204 have increased the buffer size for the ${rfc2047: expansion operator from
5207 22. Failure to deliver a bounce message always caused it to be frozen, even if
5208 there was an errors_to setting on the router. The errors_to setting is now
5211 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5212 canonical form (fully expanded) before being placed in
5213 $sender_host_address.
5215 24. Updated eximstats to version 1.33
5217 25. Include certificate and key file names in error message when GnuTLS fails
5218 to set them up, because the GnuTLS error message doesn't include the name
5219 of the failing file when there is a problem reading it.
5221 26. Expand error message when OpenSSL has problems setting up cert/key files.
5224 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5225 (this can affect the format of dates).
5227 28. exim_tidydb, when checking for the continued existence of a message for
5228 which it has found a message-specific retry record, was not finding
5229 messages that were in split spool directories. Consequently, it was
5230 deleting retry records that should have stayed in existence.
5232 29. eximstats updated to version 1.35
5233 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5234 1.35 - bugfix such that pie charts by volume are generated correctly
5236 30. The SPA authentication driver was not abandoning authentication and moving
5237 on to the next authenticator when an expansion was forced to fail,
5238 contradicting the general specification for all authenticators. Instead it
5239 was generating a temporary error. It now behaves as specified.
5241 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5242 (the order specifies the preference for clients). The order is now AES256,
5243 AES128, 3DES, ARCFOUR128.
5245 31. Small patch to Sieve code - explicitly set From: when generating an
5248 32. Exim crashed if a remote delivery caused a very long error message to be
5249 recorded - for instance if somebody sent an entire SpamAssassin report back
5250 as a large number of 550 error lines. This bug was coincidentally fixed by
5251 increasing the size of one of Exim's internal buffers (big_buffer) that
5252 happened as part of the Exiscan merge. However, to be on the safe side, I
5253 have made the code more robust (and fixed the comments that describe what
5256 33. Some experimental protocols are using DNS PTR records for new purposes. The
5257 keys for these records are domain names, not reversed IP addresses. The
5258 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5259 leaves it alone. Component reversal etc. now happens only for IP addresses.
5262 34. The host_aton() function is supposed to be passed a string that is known
5263 to be a valid IP address. However, in the case of IPv6 addresses, it was
5264 not checking this. This is a hostage to fortune. Exim now panics and dies
5265 if the condition is not met. A case was found where this could be provoked
5266 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5267 components; fortuitously, this particular loophole had already been fixed
5268 by change 4.50/55 or 4.44/33 above.
5270 If there are any other similar loopholes, the new check in host_aton()
5271 itself should stop them being exploited. The report I received stated that
5272 data on the command line could provoke the exploit when Exim was running as
5273 exim, but did not say which command line option was involved. All I could
5274 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5275 running as the user.
5278 35. There was a buffer overflow vulnerability in the SPA authentication code
5279 (which came originally from the Samba project). I have added a test to the
5280 spa_base64_to_bits() function which I hope fixes it.
5283 36. The daemon start-up calls getloadavg() while still root for those OS that
5284 need the first call to be done as root, but it missed one case: when
5285 deliver_queue_load_max is set with deliver_drop_privilege. This is
5286 necessary for the benefit of the queue runner, because there is no re-exec
5287 when deliver_drop_privilege is set.
5289 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5290 was in use, was not putting the data itself into the right store pool;
5291 consequently, it could be overwritten for a subsequent message in the same
5292 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5295 38. Sometimes the final signoff response after QUIT could fail to get
5296 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5297 < 0 before doing a fflush(). This bug looks as though it goes back to the
5298 introduction of TLS in release 3.20, but "sometimes" must have been rare
5299 because the tests only now provoked it.
5305 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5306 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5307 Now it returns an integer. A number of calls were still expecting a T/F
5308 return. Fortuitously, in all cases, the tests worked in OK situations,
5309 which is the norm. However, things would have gone wrong on any write
5310 failures on the smtp file descriptor. This function is used when sending
5311 messages over SMTP and also when doing verify callouts.
5313 2. When Exim is called to do synchronous delivery of a locally submitted
5314 message (the -odf or -odi options), it no longer closes stderr before doing
5317 3. Implemented the mua_wrapper option.
5319 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5321 5. Implemented the functions header_remove(), header_testname(),
5322 header_add_at_position(), and receive_remove_recipient(), and exported them
5325 6. If an ACL "warn" statement specified the addition of headers, Exim already
5326 inserted X-ACL-Warn: at the start if there was no header name. However, it
5327 was not making this test for the second and subsequent header lines if
5328 there were newlines in the string. This meant that an invalid header could
5329 be inserted if Exim was badly configured.
5331 7. Allow an ACL "warn" statement to add header lines at the start or after all
5332 the Received: headers, as well as at the end.
5334 8. Added the rcpt_4xx retry error code.
5336 9. Added postmaster_mailfrom=xxx to callout verification option.
5338 10. Added mailfrom=xxxx to the callout verification option, for verify=
5341 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5342 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5344 12. Inserted some casts to stop certain compilers warning when using pointer
5345 differences as field lengths or precisions in printf-type calls (mostly
5346 affecting debugging statements).
5348 13. Added optional readline() support for -be (dynamically loaded).
5350 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5351 same clock tick as a message's arrival, so that its received time was the
5352 same as the "first fail" time on the retry record, and that message
5353 remained on the queue past the ultimate address timeout, every queue runner
5354 would try a delivery (because it was past the ultimate address timeout) but
5355 after another failure, the ultimate address timeout, which should have then
5356 bounced the address, did not kick in. This was a "< instead of <=" error;
5357 in most cases the first failure would have been in the next clock tick
5358 after the received time, and all would be well.
5360 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5361 being recognized when the domain list was tested by the match_domain
5362 condition in an expansion string.
5364 16. Added the ${str2b64: operator.
5366 17. Exim was always calling setrlimit() to set a large limit for the number of
5367 processes, without checking whether the existing limit was already
5368 adequate. (It did check for the limit on file descriptors.) Furthermore,
5369 errors from getrlimit() and setrlimit() were being ignored. Now they are
5370 logged to the main and panic logs, but Exim does carry on, to try to do its
5371 job under whatever limits there are.
5373 18. Imported PCRE 5.0.
5375 19. Trivial typo in log message " temporarily refused connection" (the leading
5378 20. If the log selector return_path_on_delivery was set and an address was
5379 redirected to /dev/null, the delivery process crashed because it assumed
5380 that a return path would always be set for a "successful" delivery. In this
5381 case, the whole delivery is bypassed as an optimization, and therefore no
5384 21. Internal re-arrangement: the function for sending a challenge and reading
5385 a response while authentication was assuming a zero-terminated challenge
5386 string. It's now changed to take a pointer and a length, to allow for
5387 binary data in such strings.
5389 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5391 23. Exim was not respecting finduser_retries when seeking the login of the
5392 uid under which it was called; it was always trying 10 times. (The default
5393 setting of finduser_retries is zero.) Also, it was sleeping after the final
5394 failure, which is pointless.
5396 24. Implemented tls_on_connect_ports.
5398 25. Implemented acl_smtp_predata.
5400 26. If the domain in control=submission is set empty, Exim assumes that the
5401 authenticated id is a complete email address when it generates From: or
5402 Sender: header lines.
5404 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5405 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5406 chown and chgrp in /bin and hostname in /usr/bin.
5408 28. Exim was keeping the "process log" file open after each use, just as it
5409 does for the main log. This opens the possibility of it remaining open for
5410 long periods when the USR1 signal hits a daemon. Occasional processlog
5411 errors were reported, that could have been caused by this. Anyway, it seems
5412 much more sensible not to leave this file open at all, so that is what now
5415 29. The long-running daemon process does not normally write to the log once it
5416 has entered its main loop, and it closes the log before doing so. This is
5417 so that log files can straightforwardly be renamed and moved. However,
5418 there are a couple of unusual error situations where the daemon does write
5419 log entries, and I had neglected to close the log afterwards.
5421 30. The text of an SMTP error response that was received during a remote
5422 delivery was being truncated at 512 bytes. This is too short for some of
5423 the long messages that one sometimes sees. I've increased the limit to
5426 31. It is now possible to make retry rules that apply only when a message has a
5427 specific sender, in particular, an empty sender.
5429 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5430 it possible to be selective about when SMTP synchronization is enforced.
5432 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5434 32. Implemented hosts_connection_nolog.
5436 33. Added an ACL for QUIT.
5438 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5441 35. Added mailbox_size and mailbox_filecount to appendfile.
5443 36. Added control = no_multiline_responses to ACLs.
5445 37. There was a bug in the logic of the code that waits for the clock to tick
5446 in the case where the clock went backwards by a substantial amount such
5447 that the microsecond fraction of "now" was more than the microsecond
5448 fraction of "then" (but the whole seconds number was less).
5450 38. Added support for the libradius Radius client library this is found on
5451 FreeBSD (previously only the radiusclient library was supported).
5457 1. When certain lookups returned multiple values in the form name=value, the
5458 quoting of the values was not always being done properly. Specifically:
5459 (a) If the value started with a double quote, but contained no whitespace,
5461 (b) If the value contained whitespace other than a space character (i.e.
5462 tabs or newlines or carriage returns) it was not quoted.
5463 This fix has been applied to the mysql and pgsql lookups by writing a
5464 separate quoting function and calling it from the lookup code. The fix
5465 should probably also be applied to nisplus, ibase and oracle lookups, but
5466 since I cannot test any of those, I have not disturbed their existing code.
5468 2. A hit in the callout cache for a specific address caused a log line with no
5469 reason for rejecting RCPT. Now it says "Previous (cached) callout
5470 verification failure".
5472 3. There was an off-by-one bug in the queryprogram router. An over-long
5473 return line was truncated at 256 instead of 255 characters, thereby
5474 overflowing its buffer with the terminating zero. As well as fixing this, I
5475 have increased the buffer size to 1024 (and made a note to document this).
5477 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5478 when Exim is waiting for an SMTP response from a remote server, Exim
5479 restarts its select() call on the socket, thereby resetting its timeout.
5480 This is not a problem when such interrupts are rare. Somebody set up a cron
5481 job to run exiwhat every 2 minutes, which is less than the normal select()
5482 timeout (5 or 10 minutes). This meant that the select() timeout never
5483 kicked in because it was always reset. I have fixed this by comparing the
5484 time when an interrupt arrives with the time at the start of the first call
5485 to select(). If more time than the timeout has elapsed, the interrupt is
5486 treated as a timeout.
5488 5. Some internal re-factoring in preparation for the addition of Sieve
5489 extensions (by MH). In particular, the "personal" test is moved to a
5490 separate function, and given an option for scanning Cc: and Bcc: (which is
5491 not set for Exim filters).
5493 6. When Exim created an email address using the login of the caller as the
5494 local part (e.g. when creating a From: or Sender: header line), it was not
5495 quoting the local part when it contained special characters such as @.
5497 7. Installed new OpenBSD configuration files.
5499 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5500 try to make them clearer.
5502 9. Callout options, other than the timeout value, were being ignored when
5503 verifying sender addresses in header lines. For example, when using
5505 verify = header_sender/callout=no_cache
5507 the cache was (incorrectly) being used.
5509 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5510 only the Cygwin environment.
5512 11. When return_path_on_delivery was set as a log selector, if different remote
5513 addresses in the same message used different return paths and parallel
5514 remote delivery occurred, the wrong values would sometimes be logged.
5515 (Whenever a remote delivery process finished, the return path value from
5516 the most recently started remote delivery process was logged.)
5518 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5519 lines when AUTH and/or TLS are in use. This is the "received protocol"
5520 field. Exim used to use "asmtp" for authenticated SMTP, without any
5521 indication (in the protocol name) for TLS use. Now it follows the RFC and
5522 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5523 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5524 names appear in log lines as well as in Received: header lines.
5526 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5527 extensions, and comparison tests, and to fix some bugs.
5529 14. Changes to the "personal" filter test:
5531 (1) The test was buggy in that it was just doing the equivalent of
5532 "contains" tests on header lines. For example, if a user's address was
5533 anne@some.where, the "personal" test would incorrectly be true for
5535 To: susanne@some.where
5537 This test is now done by extracting each address from the header in turn,
5538 and checking the entire address. Other tests that are part of "personal"
5539 are now done using regular expressions (for example, to check local parts
5540 of addresses in From: header lines).
5542 (2) The list of non-personal local parts in From: addresses has been
5543 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5544 taken from the Sieve specification recommendations.
5546 (3) If the message contains any header line starting with "List-" it is
5547 treated as non-personal.
5549 (4) The test for "circular" in the Subject: header line has been removed
5550 because it now seems ill-conceived.
5552 15. Minor typos in src/EDITME comments corrected.
5554 16. Installed latest exipick from John Jetmore.
5556 17. If headers_add on a router specified a text string that was too long for
5557 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5558 of string_sprintf() is now avoided.
5560 18. $message_body_size was not set (it was always zero) when running the DATA
5561 ACL and the local_scan() function.
5563 19. For the "mail" command in an Exim filter, no default was being set for
5564 the once_repeat time, causing a random time value to be used if "once" was
5565 specified. (If the value happened to be <= 0, no repeat happened.) The
5566 default is now 0s, meaning "never repeat". The "vacation" command was OK
5567 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5568 (I found it when inspecting the code).
5570 20. There is now an overall timeout for performing a callout verification. It
5571 defaults to 4 times the callout timeout, which applies to individual SMTP
5572 commands during the callout. The overall timeout applies when there is more
5573 than one host that can be tried. The timeout is checked before trying the
5574 next host. This prevents very long delays if there are a large number of
5575 hosts and all are timing out (e.g. when the network connections are timing
5576 out). The value of the overall timeout can be changed by specifying an
5577 additional sub-option for "callout", called "maxwait". For example:
5579 verify = sender/callout=5s,maxwait=20s
5581 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5582 to the end before writing, but this should make it even safer).
5584 22. Exim was forgetting that it had advertised PIPELINING for the second and
5585 subsequent messages on an SMTP connection. It was also not resetting its
5586 memory on STARTTLS and an internal HELO.
5588 23. When Exim logs an SMTP synchronization error within a session, it now
5589 records whether PIPELINING has been advertised or not.
5591 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5592 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5593 rather than long int.
5595 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5597 26. Added the never_mail option to autoreply.
5603 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5604 crash if the getsockname() call failed; this can happen if a connection is
5605 closed very soon after it is established. The problem was simply in the
5606 order in which certain operations were done, causing Exim to try to write
5607 to the SMTP stream before it had set up the file descriptor. The bug has
5608 been fixed by making things happen in the correct order.
5614 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5615 before the connection was closed, thus losing the rejection response.
5617 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5618 some early Solaris releases, but causes trouble in current releases where
5619 socklen_t is defined.
5621 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5624 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5627 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5628 that releasing the top of it at the end releases what was used for sub-
5629 expansions (unless the block got too big). However, discard this block if
5630 the first thing is a variable or header, so that we can use its block when
5631 it is dynamic (useful for very large $message_headers, for example).
5633 6. Lookups now cache *every* query, not just the most recent. A new, separate
5634 store pool is used for this. It can be recovered when all lookup caches are
5635 flushed. Lookups now release memory at the end of their result strings.
5636 This has involved some general refactoring of the lookup sources.
5638 7. Some code has been added to the store_xxx() functions to reduce the amount
5639 of flapping under certain conditions.
5641 8. log_incoming_interface used to affect only the <= reception log lines. Now
5642 it causes the local interface and port to be added to several more SMTP log
5643 lines, for example "SMTP connection from", and rejection lines.
5645 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5647 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5649 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5650 could be overwritten at the end of the current message (or the start of a
5651 new message if it was set in a HELO ACL). The value is now preserved for
5652 the duration of the SMTP connection.
5654 12. If a transport had a headers_rewrite setting, and a matching header line
5655 contained an unqualified address, that address was qualified, even if it
5656 did not match any rewriting rules. The underlying bug was that the values
5657 of the flags that permit the existence of unqualified sender and recipient
5658 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5659 non-local messages, and by -bnq for local messages) were not being
5660 preserved with the message after it was received.
5662 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5663 "next input=" as part of the text comprising the host identity instead of
5664 the correct text. The code was using the same buffer for two different
5665 strings. However, depending on which order the printing function evaluated
5666 its arguments, the bug did not always show up. Under Linux, for example, my
5667 test suite worked just fine.
5669 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5670 doesn't work with some older versions of Perl. It has been changed to "my",
5671 which in any case is probably the better facility to use.
5673 15. A really picky compiler found some instances of statements for creating
5674 error messages that either had too many or two few arguments for the format
5677 16. The size of the buffer for calls to the DNS resolver has been increased
5678 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5679 for addresses that have a lot of PTR records. This alleviates a problem; it
5680 does not fully solve it.
5682 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5683 buffer now truncates the list and logs the incident, which is the same
5684 action as happens when Exim is looking up a host name and its aliases.
5685 Previously in this situation something unpredictable would happen;
5686 sometimes it was "internal error: store_reset failed".
5688 18. If a server dropped the connection unexpectedly when an Exim client was
5689 using GnuTLS and trying to read a response, the client delivery process
5690 crashed while trying to generate an error log message.
5692 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5693 string, for example:
5695 warn message = H1: something\nH2: something
5697 the text was added as a single header line from Exim's point of view
5698 though it ended up OK in the delivered message. However, searching for the
5699 second and subsequent header lines using $h_h2: did not work. This has been
5700 fixed. Similarly, if a system filter added multiple headers in this way,
5701 the routers could not see them.
5703 20. Expanded the error message when iplsearch is called with an invalid key to
5704 suggest using net-iplsearch in a host list.
5706 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5707 ACLs are no longer actually imposed (and a message to that effect is
5710 22. If a "gecos" field in a passwd entry contained escaped characters, in
5711 particular, if it contained a \" sequence, Exim got it wrong when building
5712 a From: or a Sender: header from that name. A second bug also caused
5713 incorrect handling when an unquoted " was present following a character
5714 that needed quoting.
5716 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5717 was not being matched caselessly.
5719 24. Arranged for all hyphens in the exim.8 source to be escaped with
5722 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5723 properly. Recipient callout cache records were still being keyed to include
5724 the sender, even when use_sender was set false. This led to far more
5725 callouts that were necessary. The sender is no longer included in the key
5726 when use_sender is false.
5728 26. Added "control = submission" modifier to ACLs.
5730 27. Added the ${base62d: operator to decode base 62 numbers.
5732 28. dnsdb lookups can now access SRV records.
5734 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5735 the configuration file.
5737 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5738 (-v) mode. This makes the output for a verbose queue run more intelligible.
5740 31. Added a use_postmaster feature to recipient callouts.
5742 32. Added the $body_zerocount variable, containing the number of binary zero
5743 bytes in the message body.
5745 33. The time of last modification of the "new" subdirectory is now used as the
5746 "mailbox time last read" when there is a quota error for a maildir
5749 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5751 35. Added +ignore_unknown as a special item in host lists.
5753 36. Code for decoding IPv6 addresses in host lists is now included, even if
5754 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5755 address was recognized as an IP address, but was then not correctly decoded
5756 into binary, causing unexpected and incorrect effects when compared with
5763 1. Very minor rewording of debugging text in manualroute to say "list of
5764 hosts" instead of "hostlist".
5766 2. If verify=header_syntax was set, and a header line with an unqualified
5767 address (no domain) and a large number of spaces between the end of the
5768 name and the colon was received, the reception process suffered a buffer
5769 overflow, and (when I tested it) crashed. This was caused by some obsolete
5770 code that should have been removed. The fix is to remove it!
5772 3. When running in the test harness, delay a bit after writing a bounce
5773 message to get a bit more predictability in the log output.
5775 4. Added a call to search_tidyup() just before forking a reception process. In
5776 theory, someone could use a lookup in the expansion of smtp_accept_max_
5777 per_host which, without the tidyup, could leave open a database connection.
5779 5. Added the variables $recipient_data and $sender_data which get set from a
5780 lookup success in an ACL "recipients" or "senders" condition, or a router
5781 "senders" option, similar to $domain_data and $local_part_data.
5783 6. Moved the writing of debug_print from before to after the "senders" test
5786 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5787 problems for message scanning, either using a data ACL, or using
5788 local_scan() because the Received: header was not generated till after they
5789 were called (in order to set the time as the time of reception completion).
5790 I have revised the way this works. The header is now generated after the
5791 body is received, but before the ACL or local_scan() are called. After they
5792 are run, the timestamp in the header is updated.
5798 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5799 before starting a queue runner without re-exec. This happened only when
5800 deliver_drop_privilege was set or when the Exim user was set to root. The
5801 effect of the bug was that timeouts during subsequent deliveries caused
5802 crashes instead of being properly handled. The handler is now left at its
5803 default (and expected) setting.
5805 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5806 message, again when deliver_drop_privilege is set or Exim is run as root.
5807 The bug described in (1) was not present in this case, but the tidying up
5808 of the other signals was missing. I have made the two cases consistent.
5810 3. The ignore_target_hosts setting on a manualroute router was being ignored
5811 for hosts that were looked up using the /MX notation.
5813 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5816 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5817 operated on the sender address. After changing the $sender_address to <>
5818 for the sender address verify, Exim was re-instated it as the original
5819 (before rewriting) address, but remembering that it had rewritten it, so it
5820 wasn't rewriting it again. This bug also had the effect of breaking the
5821 sender address verification caching when the sender address was rewritten.
5823 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5824 This has been changed so that if the ip literal address matches
5825 ignore_target_hosts, the router declines.
5827 7. Added expansion conditions match_domain, match_address, and match_local_
5828 part (NOT match_host).
5830 8. The placeholder for the Received: header didn't have a length field set.
5832 9. Added code to Exim itself and to exim_lock to test for a specific race
5833 condition that could lead to file corruption when using MBX delivery. The
5834 issue is with the lockfile that is created in /tmp. If this file is removed
5835 after a process has opened it but before that process has acquired a lock,
5836 there is the potential for a second process to recreate the file and also
5837 acquire a lock. This could lead to two Exim processes writing to the file
5838 at the same time. The added code performs the same test as UW imapd; it
5839 checks after acquiring the lock that its file descriptor still refers to
5840 the same named file.
5842 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5843 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5846 11. Added the smtp_active_hostname option. If used, this will typically be made
5847 to depend on the incoming interface address. Because $interface_address is
5848 not set up until the daemon has forked a reception process, error responses
5849 that can happen earlier (such as "too many connections") no longer contain
5852 12. If an expansion in a condition on a "warn" statement fails because a lookup
5853 defers, the "warn" statement is abandoned, and the next ACL statement is
5854 processed. Previously this caused the whole ACL to be aborted.
5856 13. Added the iplsearch lookup type.
5858 14. Added ident_timeout as a log selector.
5860 15. Added tls_certificate_verified as a log selector.
5862 16. Added a global option tls_require_ciphers (compare the smtp transport
5863 option of the same name). This controls incoming TLS connections.
5865 17. I finally figured out how to make tls_require_ciphers do a similar thing
5866 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5867 before starting the TLS session.
5869 18. Tabs are now shown as \t in -bP output.
5871 19. If the log selector return_path_on_delivery was set, Exim crashed when
5872 bouncing a message because it had too many Received: header lines.
5874 20. If two routers both had headers_remove settings, and the first one included
5875 a superfluous trailing colon, the final name in the first list and the
5876 first name in the second list were incorrectly joined into one item (with a
5877 colon in the middle).
5883 1. Added -C and -D options to the exinext utility, mainly to make it easier
5884 to include in the automated testing, but these could be helpful when
5885 multiple configurations are in use.
5887 2. The exinext utility was not formatting the output nicely when there was
5888 an alternate port involved in the retry record key, nor when there was a
5889 message id as well (for retries that were specific to a specific message
5890 and a specific host). It was also confused by IPv6 addresses, because of
5891 the additional colons they contain. I have fixed the IPv4 problem, and
5892 patched it up to do a reasonable job for IPv6.
5894 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5895 delivery, the log line now contains "pipelined" if PIPELINING was used.
5897 4. An SMTP transport process used to panic and die if the bind() call to set
5898 an explicit outgoing interface failed. This has been changed; it is now
5899 treated in the same way as a connect() failure.
5901 5. A reference to $sender_host_name in the part of a conditional expansion
5902 that was being skipped was still causing a DNS lookup. This no longer
5905 6. The def: expansion condition was not recognizing references to header lines
5906 that used bh_ and bheader_.
5908 7. Added the _cache feature to named lists.
5910 8. The code for checking quota_filecount in the appendfile transport was
5911 allowing one more file than it should have been.
5913 9. For compatibility with Sendmail, the command line option
5921 and sets the incoming protocol and host name (for trusted callers). The
5922 host name and its colon can be omitted when only the protocol is to be set.
5923 Note the Exim already has two private options, -pd and -ps, that refer to
5924 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5925 "s", but I don't think that's a major issue.
5927 10. A number of refactoring changes to the code, none of which should affect
5930 (a) The number of logging options was getting close to filling up the
5931 32-bit word that was used as a bit map. I have split them into two classes:
5932 those that are passed in the argument to log_write(), and those that are
5933 only ever tested independently outside of that function. These are now in
5934 separate 32-bit words, so there is plenty of room for expansion again.
5935 There is no change in the user interface or the logging behaviour.
5937 (b) When building, for example, log lines, the code previously used a
5938 macro that called string_cat() twice, in order to add two strings. This is
5939 not really sufficiently general. Furthermore, there was one instance where
5940 it was actually wrong because one of the argument was used twice, and in
5941 one call a function was used. (As it happened, calling the function twice
5942 did not affect the overall behaviour.) The macro has been replaced by a
5943 function that can join an arbitrary number of extra strings onto a growing
5946 (c) The code for expansion conditions now uses a table and a binary chop
5947 instead of a serial search (which was left over from when there were very
5948 few conditions). Also, it now recognizes conditions like "pam" even when
5949 the relevant support is not compiled in: a suitably worded error message is
5950 given if an attempt is made to use such a condition.
5952 11. Added ${time_interval:xxxxx}.
5954 12. A bug was causing one of the ddress fields not to be passed back correctly
5955 from remote delivery subprocesses. The field in question was not being
5956 subsequently used, so this caused to problems in practice.
5958 13. Added new log selectors queue_time and deliver_time.
5960 14. Might have fixed a bug in maildirsizefile handling that threw up
5961 "unexpected character" debug warnings, and recalculated the data
5962 unnecessarily. In any case, I expanded the warning message to give more
5965 15. Added the message "Restricted characters in address" to the statements in
5966 the default ACL that block characters like @ and % in local parts.
5968 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5969 Three changes have been made:
5971 (a) There was a serious bug; a negative response to MAIL caused the whole
5972 recipient domain to be cached as invalid, thereby blocking all messages
5973 to all local parts at the same domain, from all senders. This bug has
5974 been fixed. The domain is no longer cached after a negative response to
5975 MAIL if the sender used is not empty.
5977 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5980 (c) A new callout option, "use_sender" has been added for people who want
5981 the modified behaviour.
5987 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5990 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5991 indeed breaks things for older releases.
5993 3. Added additional logging to the case where there is a problem reading data
5994 from a filter that is running in a subprocess using a pipe, in order to
5995 try to track down a specific problem.
5997 4. Testing facility fudge: when running in the test harness and attempting
5998 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5999 getting "No route to host". Convert this to a timeout.
6001 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6004 6. Some OS don't have socklen_t but use size_t instead. This affects the
6005 fifth argument of getsockopt() amongst other things. This is now
6006 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6007 can be set for individual OS. I have set it for SunOS5, OSF1, and
6008 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6009 some earlier ones do not.
6011 7. Change 4.30/15 was not doing the test caselessly.
6013 8. The standard form for an IPv6 address literal was being rejected by address
6014 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6015 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6016 this, as well as the form without the "IPv6" on the front (but only when
6017 address literals are enabled, of course).
6019 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6021 10. Exim crashed if a message with an empty sender address specified by -f
6022 encountered a router with an errors_to setting. This could be provoked only
6023 by a command such as
6027 where an empty string was supplied; "<>" did not hit this bug.
6029 11. Installed PCRE release 4.5.
6031 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6032 remained set. It is now erased.
6034 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6035 times from message ids (which are base 36 rather than the normal 62).
6037 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6038 were being counted as actual protocol errors, and logged if the log
6039 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6040 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6041 and DATA following a set of rejected RCPTs do not count as protocol errors.
6042 In other words, Exim assumes they were pipelined, though this may not
6043 actually be the case. Of course, in all cases the client gets an
6044 appropriate error code.
6046 15. If a lookup fails in an ACL condition, a message about the failure may
6047 be available; it is used if testing the ACL cannot continue, because most
6048 such messages specify what the cause of the deferral is. However, some
6049 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6050 that caused an old message to be retained and used if a later statement
6051 caused a defer, replacing the real cause of the deferral.
6053 16. If an IP address had so many PTR records that the DNS lookup buffer
6054 was not large enough to hold them, Exim could crash while trying to process
6055 the truncated data. It now detects and logs this case.
6057 17. Further to 4.21/58, another change has been made: if (and only if) the
6058 first line of a message (the first header line) ends with CRLF, a bare LF
6059 in a subsequent header line has a space inserted after it, so as not to
6060 terminate the header.
6062 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6063 unnecessarily, used atoi() instead of strtol(), and didn't check the
6064 termination when getting file sizes from file names by regex.
6066 19. Completely re-implemented the support for maildirsize files, in the light
6067 of a number of problems with the previous contributed implementation
6068 (4.30/29). In particular:
6070 . If the quota is zero, the maildirsize file is maintained, but no quota is
6073 . If the maildir directory does not exist, it is created before any attempt
6074 to write a maildirsize file.
6076 . The quota value in the file is just a cache; if the quota is changed in
6077 the transport, the new value overrides.
6079 . A regular expression is available for excluding directories from the
6082 20. The autoreply transport checks the characters in options that define the
6083 message's headers; it allows continued headers, but it was checking with
6084 isspace() after an embedded newline instead of explicitly looking for a
6087 21. If all the "regular" hosts to which an address was routed had passed their
6088 expiry times, and had not reached their retry times, the address was
6089 bounced, even if fallback hosts were defined. Now Exim should go on to try
6092 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6093 equivalent code in the SMTP transport. Some hosts send humungous responses
6094 to HELO/EHLO, more than 1024 it seems.
6096 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6097 gives compiler warnings in some environments. I've now done it "properly",
6100 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6101 (because of problems with the built-in one) was declared to return uschar *
6102 instead of char *, causing compiler failure.
6104 25. Fixed a file descriptor leak when processing alias/forward files.
6106 26. Fixed a minor format string issue in dbfn.c.
6108 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6110 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6111 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6114 29. When a sender address is verified, it is cached, to save repeating the test
6115 when there is more than one recipient in a message. However, when the
6116 verification involves a callout, it is possible for different callout
6117 options to be set for different recipients. It is too complicated to keep
6118 track of this in the cache, so now Exim always runs a verification when a
6119 callout is required, relying on the callout cache for the optimization.
6120 The overhead is duplication of the address routing, but this should not be
6123 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6124 to be verified with callout=postmaster, and the main callout worked but the
6125 postmaster check failed, the verification correctly failed. However, if a
6126 subsequent RCPT command asked for sender verification *without* the
6127 postmaster check, incorrect caching caused this verification also to fail,
6130 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6131 it was not caching the DNS options (qualify_single, search_parents) that
6132 were used when the lookup failed. A subsequent lookup with different
6133 options therefore always gave the same answer, though there were cases
6134 where it should not have. (Example: a "domains = !$mx_any" option on a
6135 dnslookup router: the "domains" option is always processed without any
6136 widening, but the router might have qualify_single set.) Now Exim uses the
6137 cached value only when the same options are set.
6139 32. Added John Jetmore's "exipick" utility to the distribution.
6141 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6142 than a timeout (e.g. a certificate is required, and is not provided), an
6143 Exim server now closes the connection immediately. Previously it waited for
6144 the client to close - but if the client is SSL, it seems that they each
6145 wait for each other, leading to a delay before one of them times out.
6147 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6148 maintained 0.8.x compatibility because I don't think many are using it, and
6149 it is clearly obsolete.
6151 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6154 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6155 client certificate was expired. A simple patch fixes this, though I don't
6156 understand the full logic of why the verify callback is called multiple
6159 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6160 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6161 which causes problems with some clients (such as the Certicom SSL Plus
6162 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6163 disables the coutermeasure allowing Eudora to connect."
6165 38. Exim was not checking that a write() to a log file succeeded. This could
6166 lead to Bad Things if a log got too big, in particular if it hit a file
6167 size limit. Exim now panics and dies if it cannot write to a log file, just
6168 as it does if it cannot open a log file.
6170 39. Modified OS/Makefile-Linux so that it now contains
6172 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6174 The two -D definitions ensure that Exim is compiled with large file
6175 support, which makes it possible to handle log files that are bigger than
6178 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6179 instance) a domain was checked against a named list that involved a lookup,
6180 causing $domain_data to be set, then another domain was checked against the
6181 same list, then the first domain was re-checked, the value of $domain_data
6182 after the final check could be wrong. In particular, if the second check
6183 failed, it could be set empty. This bug probably also applied to
6186 41. The strip_trailing_dot option was not being applied to the address given
6187 with the -f command-line option.
6189 42. The code for reading a message's header from the spool was incrementing
6190 $received_count, but never initializing it. This meant that the value was
6191 incorrect (doubled) while delivering a message in the same process in which
6192 it was received. In the most common configuration of Exim, this never
6193 happens - a fresh exec is done - but it can happen when
6194 deliver_drop_privilege is set.
6196 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6197 it now includes up to 150 characters of the unexpected data in the log
6200 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6201 and building data strings. The size of both of these buffers was 10 000
6202 bytes - far larger than anybody would *ever* want, thought I. Needless to
6203 say, somebody hit the limit. I have increased the maximum line length to
6204 20 000 and the maximum data length of concatenated lines to 100 000. I have
6205 also fixed two bugs, because there was no checking on these buffers. Tsk,
6206 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6207 buffer is too small.
6209 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6210 lsearch lookups. Now it does.
6212 46. When parsing a route_list item in a manualroute router, a fixed-length
6213 buffer was used for the list of hosts. I made this 1024 bytes long,
6214 thinking that nobody would ever have a list of hosts that long. Wrong.
6215 Somebody had a whole pile of complicated expansion conditions, and the
6216 string was silently truncated, leading to an expansion error. It turns out
6217 that it is easier to change to an unlimited length (owing to other changes
6218 that have happened since this code was originally written) than to build
6219 structure for giving a limitation error. The length of the item that
6220 expands into the list of hosts is now unlimited.
6222 47. The lsearch lookup could not handle data where the length of text line was
6223 more than 4095 characters. Such lines were truncated, leading to shortened
6224 data being returned. It should now handle lines of any length.
6226 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6227 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6230 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6231 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6232 no longer happens in comment lines. A list of replacements is now placed
6233 at the head of all of the source files, except those whose only change is
6234 to replace PERL_COMMAND in the very first #! line.
6236 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6237 messages on the queue, with a bottom-up merge sort, using code contributed
6238 by Michael Haardt. This should make operations like -bp somewhat faster on
6239 large queues. It won't affect queue runners, except when queue_run_in_order
6242 51. Installed eximstats 1.31 in the distribution.
6244 52. Added support for SRV lookups to the dnslookup router.
6246 53. If an ACL referred to $message_body or $message_body_end, the value was not
6247 reset for any messages that followed in the same SMTP session.
6249 54. The store-handling optimization for building very long strings was not
6250 differentiating between the different store pools. I don't think this
6251 actually made any difference in practice, but I've tidied it.
6253 55. While running the routers to verify a sender address, $sender_address
6254 was still set to the sender address. This is wrong, because when routing to
6255 send a bounce to the sender, it would be empty. Therefore, I have changed
6256 it so that, while verifying a sender address, $sender_address is set to <>.
6257 (There is no change to what happens when verifying a recipient address.)
6259 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6260 target A or AAAA records (if not already returned) without resetting the
6261 qualify_single or search_parents options of the DNS resolver. These are
6262 inappropriate in this case because the targets of MX and SRV records must
6263 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6264 target that, when qualified, matched something in the local domain. These
6265 two options are now turned off when doing these lookups.
6267 57. It seems that at least some releases of Reiserfs (which does not have the
6268 concept of a fixed number of inodes) returns zero and not -1 for the
6269 number of available inodes. This interacted badly with check_spool_inodes,
6270 which assumed that -1 was the "no such thing" setting. What I have done is
6271 to check that the total number of inodes is greater than zero before doing
6272 the test of how many are available.
6274 58. When a "warn" ACL statement has a log_message modifier, the message is
6275 remembered, and not repeated. This is to avoid a lot of repetition when a
6276 message has many recipients that cause the same warning to be written.
6277 However, Exim was preserving the list of already written lines for an
6278 entire SMTP session, which doesn't seem right. The memory is now reset if a
6279 new message is started.
6281 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6282 debugging output unless log_rewrite was also set.
6284 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6285 of (int)(handle) when we know that handle contains (void *)(-1).
6287 61. The Exim daemon panic-logs an error return when it closes the incoming
6288 connection. However "connection reset by peer" seems to be common, and
6289 isn't really an error worthy of noting specially, so that particular error
6292 62. When Exim is trying to find all the local interfaces, it used to panic and
6293 die if the ioctl to get the interface flags failed. However, it seems that
6294 on at least one OS (Solaris 9) it is possible to have an interface that is
6295 included in the list of interfaces, but for which you get a failure error
6296 for this call. This happens when the interface is not "plumbed" into a
6297 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6298 failure of the "get flags" call assumes that the interface is down.
6300 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6301 makes life easier for people who are doing arithmetic on fields extracted
6302 from dates, where you often get leading zeros that should not be
6303 interpreted as octal.
6305 64. Added qualify_domain to the redirect router, to override the global
6308 65. If a pathologically long header line contained very many addresses (the
6309 report of this problem mentioned 10 000) and each of them was rewritten,
6310 Exim could use up a very large amount of memory. (It kept on making new
6311 copies of the header line as it rewrote, and never released the old ones.)
6312 At the expense of a bit more processing, the header rewriting function has
6313 been changed so that it no longer eats memory in this way.
6315 66. The generation of the Received: header has been moved from the time that a
6316 message starts to be received, to the time that it finishes. The timestamp
6317 in the Received: header should now be very close to that of the <= log
6318 line. There are two side-effects of this change:
6320 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6321 logged header lines no longer include the local Received: line, because
6322 it has not yet been created. The same applies to a copy of the message
6323 that is returned to a non-SMTP sender when a message is rejected.
6325 (b) When a filter file is tested using -bf, no additional Received: header
6326 is added to the test message. After some thought, I decided that this
6329 This change does not affect the value of $received_for. It is still set
6330 after address rewriting, but before local_scan() is called.
6332 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6334 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6335 gave an unhelpful panic error message, and a defer error. I have managed to
6336 change this behaviour so that it now rejects any supplied certificate,
6337 which seems right, as the list of acceptable certificates is empty.
6339 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6340 gave an unhelpful defer error. I have not managed to make this reject any
6341 supplied certificates, but the error message it gives is "no certificate
6342 supplied", which is not helpful.
6344 70. exigrep's output now also includes lines that are not associated with any
6345 message, but which match the given pattern. Implemented by a patch from
6346 Martin Sluka, which also tidied up the Perl a bit.
6348 71. Recipient callout verification, like sender verification, was using <> in
6349 the MAIL FROM command. This isn't really the right thing, since the actual
6350 sender may affect whether the remote host accepts the recipient or not. I
6351 have changed it to use the actual sender in the callout; this means that
6352 the cache record is now keyed on a recipient/sender pair, not just the
6353 recipient address. There doesn't seem to be a real danger of callout loops,
6354 since a callout by the remote host to check the sender would use <>.
6355 [SEE ABOVE: changed after hitting problems.]
6357 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6358 temporary errors. However, in the case of such a code being given after
6359 the end of a data transmission (i.e. after ".") Exim was failing to write
6360 a retry record for the message. (Yes, there was some broken host that was
6361 actually sending 8xx at this point.)
6363 73. An unknown lookup type in a host list could cause Exim to panic-die when
6364 the list was checked. (An example that provoked this was putting <; in the
6365 middle of a list instead of at the start.) If this happened during a DATA
6366 ACL check, a -D file could be left lying around. This kind of configuration
6367 error no longer causes Exim to die; instead it causes a defer error. The
6368 incident is still logged to the main and panic logs.
6370 74. Buglet left over from Exim 3 conversion. The message "too many messages
6371 in one connection" was written to the rejectlog but not the mainlog, except
6372 when address rewriting (yes!) was being logged.
6374 75. Added write_rejectlog option.
6376 76. When a system filter was run not as root (that is, when system_filter_user
6377 was set), the values of the $n variables were not being returned to the
6378 main process; thus, they were not subsequently available in the $sn
6381 77. Added +return_path_on_delivery log selector.
6383 78. A connection timeout was being treated differently from recipients deferred
6384 when testing hosts_max_try with a message that was older than the host's
6385 retry timeout. (The host should not be counted, thus allowing all hosts to
6386 be tried at least once before bouncing.) This may have been the cause of an
6387 occasionally reported bug whereby a message would remain on the queue
6388 longer than the retry timeout, but would be bounced if a delivery was
6389 forced. I say "may" because I never totally pinned down the problem;
6390 setting up timeout/retry tests is difficult. See also the next item.
6392 79. The ultimate address timeout was not being applied to errors that involved
6393 a combination of host plus message (for example, a timeout on a MAIL
6394 command). When an address resolved to a number of possible hosts, and they
6395 were not all tried for each delivery (e.g. because of hosts_max_try), a
6396 message could remain on the queue longer than the retry timeout.
6398 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6401 81. Fixed an obscure SMTP outgoing bug which required at least the following
6402 conditions: (a) there was another message waiting for the same server;
6403 (b) the server returned 5xx to all RCPT commands in the first message so
6404 that the message was not completed; (c) the server dropped the connection
6405 or gave a negative response to the RSET that Exim sends to abort the
6406 transaction. The observed case was a dropped connection after DATA that had
6407 been sent in pipelining mode. That is, the server had advertised PIPELINING
6408 but was not implementing it correctly. The effect of the bug was incorrect
6409 behaviour, such as trying another host, and this could lead to a crash.
6415 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6416 and daemon.c were passed as pointers to ints; they should have been
6417 pointers to socklen_t variables (which are typically unsigned ints).
6419 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6422 3. Fixed a really odd bug that affected only the testing scheme; patching a
6423 certain fixed string in the binary changed the value of another string that
6424 happened to be identical to the end of the original first string.
6426 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6427 name", it returns that address as the IP address. On some operating
6428 systems (e.g. Solaris), it also passes back the IP address string as the
6429 "host name". However, on others (e.g. Linux), it passes back an empty
6430 string. Exim wasn't checking for this, and was changing the host name to an
6431 empty string, assuming it had been canonicalized.
6433 5. Although rare, it is permitted to have more than one PTR record for a given
6434 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6435 all the names associated with an address, because they do in Solaris.
6436 However, it seems that they do not in Linux for data that comes from the
6437 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6438 I found this out when I moved to a new Linux workstation and tried to run
6439 the Exim test suite.
6441 To get round this problem I have changed the code so that it now does its
6442 own call to the DNS to look up PTR records when searching for a host name.
6443 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6444 addresses that are only in /etc/hosts are still found.
6446 This behaviour is, however, controlled by an option called host_lookup_
6447 order, which defaults to "bydns:byaddr". If people want to use the other
6448 order, or indeed, just use one or the other means of lookup, they can
6449 specify it in this variable.
6451 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6452 some operating systems, this comes back from gethostbyaddr() as an empty
6453 string, and this is what Exim used to test for. However, it seems that in
6454 other systems, "." is yielded. Exim now tests for this case too.
6456 7. The values of check_spool_space and check_log_space are now held internally
6457 as a number of kilobytes instead of an absolute number of bytes. If a
6458 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6459 kilobyte. This means that much larger values can be stored.
6461 8. Exim monitor: an attempt to get the action menu when not actually pointing
6462 at a message produces an empty menu entitled "No message selected". This
6463 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6464 no entries in it ("Shell widget menu has zero width and/or height"). So I
6465 have added a single, blank menu entry in this case.
6467 9. Added ${quote_local_part.
6469 10. MIME decoding is now applied to the contents of Subject: header lines when
6472 11. Now that a reference to $sender_host_address automatically causes a reverse
6473 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6474 host lookup before query-style lookups in lists that might use this
6475 variable. This has therefore been abolished, and the "net-" prefix is no
6476 longer necessary for query-style lookups.
6478 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6479 have been a typo for LFLAGS, so it has been changed.
6481 13. The install script calls Exim with "-C /dev/null" in order to find the
6482 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6483 to be output. However, since Exim outputs its version number before the
6484 error, it didn't break the script. It just looked ugly. I fixed this by
6485 always allowing "-C /dev/null" if the caller is root.
6487 14. Ignore overlarge ACL variable number when reading spool file - insurance
6488 against a later release with more variables having written the file.
6490 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6491 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6492 this, as well as the form without the "IPv6" on the front.
6494 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6495 OS/Makefile-Darwin file.
6497 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6498 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6500 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6501 variable. However, some OS use ints for inodes. I've added cast to long int
6502 to get rid of the compiler warning.
6504 19. I had forgotten to lock out "/../" in configuration file names when
6505 ALT_CONFIG_PREFIX was set.
6507 20. Routers used for verification do not need to specify transports. However,
6508 if such a router generated a host list, and callout was configured, Exim
6509 crashed, because it could not find a port number from the (non-existent)
6510 transport. It now assumes port 25 in this circumstance.
6512 21. Added the -t option to exigrep.
6514 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6515 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6516 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6517 them is set, LOOKUP_LSEARCH is forced.
6519 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6521 24. Added sender and host information to the "rejected by local_scan()" log
6522 line; previously there was no indication of these.
6524 25. Added .include_if_exists.
6526 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6527 Linux. It turns out that not all file systems support this. Apparently some
6528 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6529 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6530 not supported on the file descriptor, is now ignored when Exim is trying to
6531 sync a directory. This applies only to Linux.
6533 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6535 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6536 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6539 29. Added support for maildirsize files from supplied patch (modified a bit).
6541 30. The use of :fail: followed by an empty string could lead Exim to respond to
6542 sender verification failures with (e.g.):
6544 550 Verification failed for <xxx>
6545 550 Sender verify failed
6547 where the first response line was missing the '-' that indicates it is not
6548 the final line of the response.
6550 31. The loop for finding the name of the user that called Exim had a hardwired
6551 limit of 10; it now uses the value of finduser_retries, which is used for
6552 all other user lookups.
6554 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6557 33. Exim was neglecting to zero errno before one call of strtol() when
6558 expanding a string and expecting an integer value. On some systems this
6559 resulted in spurious "integer overflow" errors. Also, it was casting the
6560 result into an int without checking.
6562 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6563 did not work. The code looks as if it has *never* worked, though it appears
6564 to have been documented since at least release 1.62. I have made it work.
6566 35. The "timeout_DNS" error in retry rules, also documented since at least
6567 1.62, also never worked. As it isn't clear exactly what this means, and
6568 clearly it isn't a major issue, I have abolished the feature by treating it
6569 as "timeout", and writing a warning to the main and panic logs.
6571 36. The display of retry rules for -brt wasn't always showing the error code
6574 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6575 timeout_connect_A, timeout_connect_MX.
6577 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6578 to the empty sender.
6580 39. The daemon was not analysing the content of -oX till after it had closed
6581 stderr and disconnected from the controlling terminal. This meant that any
6582 syntax errors were only noted on the panic log, and the return code from
6583 the command was 0. By re-arranging the code a little, I've made the
6584 decoding happen first, so such errors now appear on stderr, and the return
6585 code is 1. However, the actual setting up of the sockets still happens in
6586 the disconnected process, so errors there are still only recorded on the
6589 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6590 connections (as happens on some IP stacks) was logged at start up time as
6591 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6592 it from "IPv6 and IPv4", which means that two separate sockets are being
6595 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6596 says whether AF_INET or AF_INET6 was passed as an argument.
6598 42. Exiwhat output was messed up when time zones were included in log
6601 43. Exiwhat now gives more information about the daemon's listening ports,
6602 and whether -tls-on-connect was used.
6604 44. The "port" option of the smtp transport is now expanded.
6606 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6607 silently ignored. Now an error message is written to the main and panic
6610 46. There's a new ACL modifier called "logwrite" which writes to a log file
6611 as soon as it is encountered.
6613 47. Added $local_user_uid and $local_user_gid at routing time.
6615 48. Exim crashed when trying to verify a sender address that was being
6618 49. Exim was recognizing only a space character after ".include". It now also
6619 recognizes a tab character.
6621 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6622 extracting the relevant information from the specification. The man page no
6623 longer contains scrambled data for the -d option, and I've added a section
6624 at the front about calling Exim under different names.
6626 51. Added "extra_headers" argument to the "mail" command in filter files.
6628 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6631 53. Installed eximstats 1.29.
6633 54. Added transport_filter_timeout as a generic transport option.
6635 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6636 Cc: header lines. This was required by RFC 822, but it not required by RFC
6639 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6640 incoming messages that did not have them. Now it does so only if the
6641 message originates locally, that is, if there is no associated remote host
6642 address. When Resent- header lines are present, this applies to the Resent-
6643 lines rather than the non-Resent- lines.
6645 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6646 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6648 58. Messages for configuration errors now include the name of the main
6649 configuration file - useful now that there may be more than one file in a
6650 list (.included file names were always shown).
6652 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6653 for those rare installations that do not start the daemon as root or run it
6654 setuid root. I've cut out the call to initgroups() if the daemon is not
6657 60. The Exim user and group can now be bound into the binary as text strings
6658 that are looked up at the start of Exim's processing.
6660 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6662 62. Added $mailstore_basename variable.
6664 63. Installed patch to sieve.c from Michael Haardt.
6666 64. When Exim failed to open the panic log after failing to open the main log,
6667 the original message it was trying to log was written to stderr and debug
6668 output, but if they were not available (the usual case in production), it
6669 was lost. Now it is written to syslog before the two lines that record the
6670 failures to open the logs.
6672 65. Users' Exim filters run in subprocesses under the user's uid. It is
6673 possible for a "deliver" command or an alias in a "personal" command to
6674 provoke an address rewrite. If logging of address rewriting is configured,
6675 this fails because the process is not running as root or exim. There may be
6676 a better way of dealing with this, but for the moment (because 4.30 needs
6677 to be released), I have disabled address rewrite logging when running a
6678 filter in a non-root, non-exim process.
6684 1. The buildconfig auxiliary program wasn't quoting the value set for
6685 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6686 not defined. This bug was masked in 4.22 by the effect that was fixed in
6689 2. Some messages that were rejected after a message id was allocated were
6690 shown as "incomplete" by exigrep. It no longer does this for messages that
6691 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6693 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6694 have allow_domain_literals set, the ID did not get logged in the <= line.
6695 Domain literals are now always recognized in Message-ID: header lines.
6697 4. The first argument for a ${extract expansion item is the key name or field
6698 number. Leading and trailing spaces in this item were not being ignored,
6699 causing some misleading effects.
6701 5. When deliver_drop_privilege was set, single queue runner processes started
6702 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6703 same command in the process it spins off) were not dropping privilege.
6705 6. When the daemon running as "exim" started a queue runner, it always
6706 re-executed Exim in the spun-off process. This is a waste of effort when
6707 deliver_drop_privilege is set. The new process now just calls the
6708 queue-runner function directly.
6714 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6717 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6718 ignored. Though the use of -oP was forcing the writing of a pid file, it
6719 was always written to the default place.
6721 3. If the message "no IP address found for host xxxx" is generated during
6722 incoming verification, it is now followed by identification of the incoming
6723 connection (so you can more easily find what provoked it).
6725 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6727 5. Added some features to "harden" Exim a bit more against certain attacks:
6729 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6730 be put in Local/Makefile. This is like the never_users runtime option,
6731 but it cannot be overridden. The default setting is "root".
6733 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6734 prefix string with which any file named in a -C command line option
6737 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6738 is retained for -C and -D only if the caller of Exim is root. Without
6739 it, the exim user may also use -C and -D and retain privilege.
6741 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6742 command line option is disabled.
6744 6. Macro names set by the -D option must start with an upper case letter, just
6745 like macro names defined in the configuration file.
6747 7. Added "dereference=" facility to LDAP.
6749 8. Two instances of the typo "uknown" in the source files are fixed.
6751 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6752 the Configure-Makefile script screwed up while processing it.
6754 10. Incorporated PCRE 4.4.
6756 11. The SMTP synchronization check was not operating right at the start of an
6757 SMTP session. For example, it could not catch a HELO sent before the client
6758 waited for the greeting. There is now a check for outstanding input at the
6759 point when the greeting is written. Because of the duplex, asynchronous
6760 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6761 way, but not yet received, when the check is performed.
6763 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6764 on TCP/IP sockets, because this apparently causes some broken clients to
6767 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6768 unchanged) from the Cygwin maintainer.
6770 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6771 was supported instead of testing for mbx. Effectively a typo.
6773 15. The spa authenticator server code was not checking that the input it
6774 received was valid base64.
6776 16. The debug output line for the "set" modifier in ACLs was not showing the
6777 name of the variable that was being set.
6779 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6781 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6782 lookup on its own. Something else was needed to trigger the lookup. For
6783 example, a match in host_lookup or the need for a host name in a host list.
6784 Now, if $sender_host_name is referenced and the host name has not yet been
6785 looked up, a lookup is performed. If the lookup fails, the variable remains
6786 empty, and $host_lookup_failed is set to "1".
6788 19. Added "eqi" as a case-independent comparison operator.
6790 20. The saslauthd authentication condition could segfault if neither service
6791 nor realm was specified.
6793 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6794 error message that was logged was misleading, and incoming SMTP
6795 connections were dropped. The message is now more accurate, and temporary
6796 errors are given to SMTP connections.
6798 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6799 (except RSET) with a 421 error, until QUIT is received. However, it was
6800 failing to send a response to QUIT.
6802 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6803 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6804 ACL is now run after the helo_try_verify_hosts code.
6806 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6807 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6808 case-independent, so other case variants are also recognized). Apparently
6809 some systems use these upper case variants.
6811 25. If more than two messages were waiting for the same host, and a transport
6812 filter was specified for the transport, Exim sent two messages over the
6813 same TCP/IP connection, and then failed with "socket operation on non-
6814 socket" when it tried to send the third.
6816 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6818 27. The extern definition of crypt16() in expand.c was not being excluded when
6819 the OS had its own crypt16() function.
6821 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6822 as a preferred synonym for return_size_limit, both as an option and as an
6825 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6827 30. Changed the default configuration ACL to relax the local part checking rule
6828 for addresses that are not in any local domains. For these addresses,
6829 slashes and pipe symbols are allowed within local parts, but the sequence
6830 /../ is explicitly forbidden.
6832 31. SPA server authentication was not clearing the challenge buffer before
6835 32. log_message in a "warn" ACL statement was writing to the reject log as
6836 well as to the main log, which contradicts the documentation and doesn't
6837 seem right (because no rejection is happening). So I have stopped it.
6839 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6840 However, I am unable to do any testing of this.
6842 34. Fixed an infelicity in the appendfile transport. When checking directories
6843 for a mailbox, to see if any needed to be created, it was accidentally
6844 using path names with one or more superfluous leading slashes; tracing
6845 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6847 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6848 contents are added to the log line that is written for every discarded
6849 recipient. (Previously a log_message setting was ignored.)
6851 36. The ${quote: operator now quotes the string if it is empty.
6853 37. The install script runs exim in order to find its version number. If for
6854 some reason other than non-existence or emptiness, which it checks, it
6855 could not run './exim', it was installing it with an empty version number,
6856 i.e. as "exim-". This error state is now caught, and the installation is
6859 38. An argument was missing from the function that creates an error message
6860 when Exim fails to connect to the socket for saslauthd authentication.
6861 This could cause Exim to crash, or give a corrupted message.
6863 39. Added isip, isip4, and isip6 to ${if conditions.
6865 40. The ACL variables $acl_xx are now saved with the message, and can be
6866 accessed later in routers, transports, and filters.
6868 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6869 strings in the file are not string-expanded.
6871 42. If a MAIL command specified a SIZE value that was too large to fit into an
6872 int variable, the check against message_size_limit failed. Such values are
6873 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6874 day this will have to be increased, but I don't think I want to be around
6875 when emails are that large.
6882 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6883 iconv() is not standard in FreeBSD.
6885 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6886 IPv6 enabled. The observed symptom was a segmentation fault on return from
6887 the function os_common_find_running_interfaces() in src/os.c.
6889 3. In the check_special_case() function in daemon.c I had used "errno" as an
6890 argument name, which causes warnings on some systems. This was basically a
6891 typo, since it was named "eno" in the comments!
6893 4. The code that waits for the clock to tick (at a resolution of some fraction
6894 of a second) so as to ensure message-id uniqueness was always waiting for
6895 at least one whole tick, when it could have waited for less. [This is
6896 almost certainly not relevant at current processor speeds, where it is
6897 unlikely to ever wait at all. But we try to future-proof.]
6899 5. The function that sleeps for a time interval that includes fractions of a
6900 second contained a race. It did not block SIGALRM between setting the
6901 timer, and suspending (a couple of lines later). If the interval was short
6902 and the sigsuspend() was delayed until after it had expired, the suspension
6903 never ended. On busy systems this could lead to processes getting stuck for
6906 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6907 process, before it forks any delivery processes. The open lookup caching
6908 mechanism meant that the open file or database connection was passed into
6909 the delivery process. The problem was that delivery processes always tidy
6910 up cached lookup data. This could cause a problem for the next delivery
6911 process started by the queue runner, because the external queue runner
6912 process does not know about the closure. So the next delivery process
6913 still has data in the lookup cache. In the case of a file lookup, there was
6914 no problem because closing a file descriptor in a subprocess doesn't affect
6915 the parent. However, if the lookup was caching a connection to a database,
6916 the connection was closed, and the second delivery process was likely to
6917 see errors such as "PGSQL: query failed: server closed the connection
6918 unexpectedly". The problem has been fixed by closing all cached lookups
6919 in a queue runner before running a delivery process.
6921 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6922 seem to have the "const" qualifier which it has on other OS. I've
6925 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6926 *of the same type* (client or server) with the same public name that an
6927 error should be diagnosed.
6929 9. When Exim looked up a host name for an IP address, but failed to find the
6930 original IP address when looking up the host name (a safety check), it
6931 output the message "<ip address> does not match any IP for NULL", which was
6932 confusing, to say the least. The bug was that the host name should have
6933 appeared instead of "NULL".
6935 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6936 user that is built into the binary, and the -C or -D options is used, root
6937 privilege is dropped before the configuration file is read. In addition,
6938 logging is switched to stderr instead of the normal log files. If the
6939 configuration then re-defines the Exim user, the unprivileged environment
6940 is probably not what is expected, so Exim logs a panic warning message (but
6943 However, if deliver_drop_privilege is set, the unprivileged state may well
6944 be exactly what is intended, so the warning has been cut out in that case,
6945 and Exim is allowed to try to write to its normal log files.
6951 1. smtp_return_error_details was not giving details for temporary sender
6952 or receiver verification errors.
6954 2. Diagnose a configuration error if two authenticators have the same public
6957 3. Exim used not to create the message log file for a message until the first
6958 delivery attempt. This could be confusing when incoming messages were held
6959 for policy or load reasons. The message log file is now created at the time
6960 the message is received, and an initial "Received" line is written to it.
6962 4. The automatically generated man page for command line options had a minor
6963 bug that caused no ill effects; however, a more serious problem was that
6964 the procedure for building the man page automatically didn't always
6965 operate. Consequently, release 4.20 contains an out-of-date version. This
6966 shouldn't happen again.
6968 5. When building Exim with embedded Perl support, the script that builds the
6969 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6970 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6972 6. The freeze_tell option was not being used for messages that were frozen on
6973 arrival, either by an ACL or by local_scan().
6975 7. Added the smtp_incomplete_transaction log selector.
6977 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6978 was accepting AUTH without a new EHLO.
6980 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6981 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6984 10. Logging of TCP/IP connections (when configured) now happens in the main
6985 daemon process instead of the child process, so that the TCP/IP connection
6986 count is more accurate (but it can never be perfect).
6988 11. The use of "drop" in a nested ACL was not being handled correctly in the
6989 outer ACL. Now, if condition failure induced by the nested "drop" causes
6990 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6991 or "require"), the connection is dropped.
6993 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6994 that yield "discard" can now be used with an "accept" or a "discard" verb,
6995 but an error is generated for any others (because I can't see a useful way
6996 to define what should happen).
6998 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6999 are now processed in the same way as lines in the Exim configuration file.
7000 In particular, continuation lines are supported.
7002 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7004 15. Added -ti meaning -t -i.
7006 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7007 domains, and warn by logging if others are found.
7009 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7010 returned from GIFCONF when Exim is trying to find the list of interfaces on
7011 a host. The code in os.c has been modified to copy each ifreq to an aligned
7012 structure in all cases.
7014 Also, in some cases, the returned ifreq's were being copied to a 'struct
7015 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7016 means the last couple of bytes of an IPv6 address could be chopped if the
7017 ifreq contained only a normal sockaddr (14 bytes storage).
7019 18. Named domain lists were not supported in the hosts_treat_as_local option.
7020 An entry such as +xxxx was not recognized, and was treated as a literal
7023 19. Ensure that header lines added by a DATA ACL are included in the reject log
7024 if the ACL subsequently rejects the message.
7026 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7027 MD5 (which is deprecated).
7029 21. When testing a filter file using -bf, Exim was writing a message when it
7030 took the sender from a "From " line in the message, but it was not doing so
7031 when it took $return_path from a Return-Path: header line. It now does.
7033 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7034 with a valid header line field name (a series of printing characters
7035 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7037 23. Changed "disc" in the source to "disk" to conform to the documentation and
7038 the book and for uniformity.
7040 24. Ignore Sendmail's -Ooption=value command line item.
7042 25. When execve() failed while trying to run a command in a pipe transport,
7043 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7044 could be confused with a return value of 69 from the command itself. This
7045 has been changed to 127, the value the shell returns if it is asked to run
7046 a non-existent command. The wording for the related log line suggests a
7047 non-existent command as the problem.
7049 26. If received_header_text expands to an empty string, do not add a Received:
7050 header line to the message. (Well, it adds a token one on the spool, but
7051 marks it "old" so that it doesn't get used or transmitted.)
7053 27. Installed eximstats 1.28 (addition of -nt option).
7055 28. There was no check for failure on the call to getsockname() in the daemon
7056 code. This can fail if there is a shortage of resources on the system, with
7057 ENOMEM, for example. A temporary error is now given on failure.
7059 29. Contrary to the C standard, it seems that in some environments, the
7060 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7061 program. Exim now does this explicitly; it affects the formatting of
7062 timestamps using strftime().
7064 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7065 complaints. I've now initialized all the variables, to avoid this.
7067 32. Header lines added by a system filter were not being "seen" during
7068 transport-time rewrites.
7070 33. The info_callback() function passed to OpenSSL is set up with type void
7071 (*)(SSL *, int, int), as described somewhere. However, when calling the
7072 function (actually a macro) that sets it up, the type void(*)() is
7073 expected. I've put in a cast to prevent warnings from picky compilers.
7075 34. If a DNS black list lookup found a CNAME record, but there were no A
7076 records associated with the domain it pointed at, Exim crashed.
7078 35. If a DNS black list lookup returned more than one A record, Exim ignored
7079 all but the first. It now scans all returned addresses if a particular IP
7080 value is being sought. In this situation, the contents of the
7081 $dnslist_value variable are a list of all the addresses, separated by a
7084 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7085 to accept a host name and all its aliases if the forward lookup for any of
7086 them yielded the IP address of the incoming connection. Now it accepts only
7087 those names whose forward lookup yields the correct IP address. Any other
7088 names are discarded. This closes a loophole whereby a rogue DNS
7089 administrator could create reverse DNS records to break through a
7090 wildcarded host restriction in an ACL.
7092 37. If a user filter or a system filter that ran in a subprocess used any of
7093 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7094 the wrong values were passed to the pipe command ($thisaddress had the
7095 value of $0, $0 had the value of $1, etc). This bug was introduced by
7096 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7098 38. Improved the line breaking for long SMTP error messages from ACLs.
7099 Previously, if there was no break point between 40 and 75 characters, Exim
7100 left the rest of the message alone. Two changes have been made: (a) I've
7101 reduced the minimum length to 35 characters; (b) if it can't find a break
7102 point between 35 and 75 characters, it looks ahead and uses the first one
7103 that it finds. This may give the occasional overlong line, but at least the
7104 remaining text gets split now.
7106 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7107 file descriptors might be low, and that setting 1000 would always raise it.
7108 It turns out that in some environments, the limit is already over 1000 and
7109 that lowering it causes trouble. So now Exim takes care not to decrease it.
7111 40. When delivering a message, the value of $return_path is set to $sender_
7112 address at the start of routing (routers may change the value). By an
7113 oversight, this default was not being set up when an address was tested by
7114 -bt or -bv, which affected the outcome if any router or filter referred to
7117 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7118 log only when "message" or "log_message" are set. However, if one of the
7119 conditions was an address verification, or a call to a nested ACL, the
7120 messages generated by the underlying test were being passed through. This
7121 no longer happens. The underlying message is available in $acl_verify_
7122 message for both "message" and "log_message" expansions, so it can be
7123 passed through if needed.
7125 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7126 new expansion $bh_ to give the encoded byte string without charset
7127 translation. Translation happens only if iconv() is available; HAVE_ICONV
7128 indicates this at build time. HEADERS_CHARSET gives the charset to
7129 translate to; headers_charset can change it in the configuration, and
7130 "headers charset" can change it in an individual filter file.
7132 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7133 that creates RFC 2047 encoded "words" labels them as that charset instead
7134 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7135 expansion operator; (ii) when Exim creates a From: line for a local
7136 message; (iii) when a header line is rewritten to include a "phrase" part.
7138 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7139 buggy, causing it to skip the first lines of messages whose message ID
7140 ended in 'D'. This would not have bitten before Exim release 4.14, because
7141 message IDs were unlikely to end in 'D' before then. The effect was to have
7142 incorrect size information for certain domains.
7144 45. #include "config.h" was missing at the start of the crypt16.c module. This
7145 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7148 46. If there was a timeout during a "random" callout check, Exim treated it as
7149 a failure of the random address, and carried on sending RSET and the real
7150 address. If the delay was just some slowness somewhere, the response to the
7151 original RCPT would be taken as a response to RSET and so on, causing
7152 mayhem of various kinds.
7154 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7155 when I implemented it. It didn't allow for the fact that some option values
7156 may legitimately be negative (e.g. size_addition), and it didn't even do
7157 the right test for positive values.
7159 48. Domain names in DNS records are case-independent. Exim always looks them up
7160 in lower case. Some resolvers return domain names in exactly the case they
7161 appear in the zone file, that is, they may contain uppercase letters. Not
7162 all resolvers do this - some return always lower case. Exim was treating a
7163 change of case by a resolver as a change of domain, similar to a widening
7164 of a domain abbreviation. This triggered its re-routing code and so it was
7165 trying to route what was effectively the same domain again. This normally
7166 caused routing to fail (because the router wouldn't handle the domain
7167 twice). Now Exim checks for this case specially, and just changes the
7168 casing of the domain that it ultimately uses when it transmits the message
7171 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7174 50. If a filter generated a file delivery with a non-absolute name (possible if
7175 no home directory exists for the router), the forbid_file option was not
7178 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7179 the existing equality matching.
7181 52. Exim was using ints instead of ino_t variables in some places where it was
7182 dealing with inode numbers.
7184 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7185 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7186 called TMPDIR, and if it finds it is different, it changes its value.
7188 54. The smtp_printf() function is now made available to local_scan() so
7189 additional output lines can be written before returning. There is also an
7190 smtp_fflush() function to enable the detection of a dropped connection.
7191 The variables smtp_input and smtp_batched_input are exported to
7194 55. Changed the default runtime configuration: the message "Unknown user"
7195 has been removed from the ACL, and instead placed on the localuser router,
7196 using the cannot_route_message feature. This means that any verification
7197 failures that generate their own messages won't get overridden. Similarly,
7198 the "Unrouteable address" message that was in the ACL for unverifiable
7199 relay addresses has also been removed.
7201 56. Added hosts_avoid_esmtp to the smtp transport.
7203 57. The exicyclog script was not checking for the esoteric option
7204 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7205 will work only if exicyclog is run under the appropriate euid.
7207 58. Following a discussion on the list, the rules by which Exim recognises line
7208 endings on incoming messages have been changed. The -dropcr and drop_cr
7209 options are now no-ops, retained only for backwards compatibility. The
7210 following line terminators are recognized: LF CRLF CR. However, special
7211 processing applies to CR:
7213 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7214 nor a local message in the state where . is a terminator.
7216 (ii) If a bare CR is encountered in a header line, an extra space is added
7217 after the line terminator so as not to end the header. The reasoning
7218 behind this is that bare CRs in header lines are most likely either
7219 to be mistakes, or people trying to play silly games.
7221 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7222 was being incorrectly given as 18 bytes larger than it should have been.
7223 This is a VOB (very old bug).
7225 60. This may never have affected anything current, but just in case it has:
7226 When the local host is found other than at the start of a list of hosts,
7227 the local host, those with the same MX, and any that follow, are discarded.
7228 When the list in question was part of a longer list of hosts, the following
7229 hosts (not currently being processed) were also being discarded. This no
7230 longer happens. I'm not sure if this situation could ever has previously
7233 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7236 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7237 header. This is something that is recommended in a new Internet Draft, and
7238 is something that is documented as being done by Sendmail. There are two
7239 possible values. For messages generated by the autoreply transport, Exim
7242 Auto-Submitted: auto-replied
7244 whereas for all other generated messages (e.g. bounces) it adds
7246 Auto-Submitted: auto-generated
7248 63. The "personal" condition in filters now includes a test for the
7249 Auto-Submitted: header. If it contains the string "auto-" the message it
7250 not considered personal.
7252 64. Added rcpt_include_affixes as a generic transport option.
7254 65. Added queue_only_override (default true).
7256 66. Added the syslog_duplication option.
7258 67. If what should have been the first header line of a message consisted of
7259 a space followed by a colon, Exim was mis-interpreting it as a header line.
7260 It isn't of course - it is syntactically invalid and should therefore be
7261 treated as the start of the message body. The misbehaviour could have
7262 caused a number of strange effects, including loss of data in subsequent
7263 header lines, and spool format errors.
7265 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7266 client host had authenticated. This control can now be exercised by an ACL
7267 for more flexibility.
7269 69. By default, callouts do not happen when testing with -bh. There is now a
7270 variant, -bhc, which does actually run the callout code, including
7271 consulting and updating the callout cache.
7273 70. Added support for saslauthd authentication, courtesy of Alexander
7276 71. If statvfs() failed on the spool or log directories while checking their
7277 size for availability, Exim confusingly gave the error "space shortage".
7278 Furthermore, in debugging mode it crashed with a floating point exception.
7279 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7280 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7281 serious problem, Exim now writes to the main and panic logs when this
7282 happens, with details of the failure. It then refuses to accept the
7283 incoming message, giving the message "spool directory problem" or "log
7284 directory problem" with a 421 code for SMTP messages.
7286 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7287 0, 1, and 2 exist, because some OS complain for execs without them (see
7288 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7289 descriptors. However, the code omitted to check that the open succeeded,
7290 causing mysterious errors if for some reason the permissions on /dev/null
7291 got screwed. Now Exim writes a message to the main and panic logs, and
7292 bombs out if it can't open /dev/null.
7294 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7295 interact so that it is all more flexible. It is supposed to remain
7296 backwards compatible. Also added extra_local_interfaces.
7298 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7299 to bomb out with an assertion failure - to the client this appears as a
7300 connection drop. This problem occurs in the part of the code that was taken
7301 from the Samba project. Fortunately, the assertion is in a very simple
7302 function, so I have fixed this by reproducing the function inline in the
7303 one place where it is called, and arranging for authentication to fail
7304 instead of killing the process with assert().
7306 75. The SPA client code was not working when the server requested OEM rather
7307 than Unicode encoding.
7309 76. Added code to make require_files with a specific uid setting more usable in
7310 the case where statting the file as root fails - usually a non-root-mounted
7311 NFS file system. When this happens and the failure is EACCES, Exim now
7312 forks a subprocess and does the per-uid checking as the relevant uid.
7314 77. Added process_log_path.
7316 78. If log_file_path was not explicitly set, a setting of check_log_space or
7317 check_log_inodes was ignored.
7319 79. If a space check for the spool or log partitions fails, the incident is now
7320 logged. Of course, in the latter case the data may get lost...
7322 80. Added the %p formatting code to string_format() so that it can be used to
7323 print addresses in debug_print(). Adjusted all the address printing in the
7324 debugging in store.c to use %p rather than %d.
7326 81. There was a concern that a line of code in smtp_in.c could overflow a
7327 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7328 initially expressed, the concern was not well-founded, because trailing
7329 spaces are removed early. However, if the trailing spaces were followed by
7330 a NULL, they did not get removed, so the overflow was possible. Two fixes
7333 (a) I re-wrote the offending code in a cleaner fashion.
7334 (b) If an incoming SMTP command contains a NULL character, it is rejected
7337 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7338 runs initgroups(), so that if the Exim user is in any additional groups,
7339 they will be used during message reception.
7345 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob