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.
176 JH/01 Bug 2351: Log failures to extract envelope addresses from message headers.
182 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
183 When redis_servers is set to a list of > 1 element, and the Redis servers
184 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
185 case of MOVED into a DEFER case instead, thus moving the query onto the
186 next server in the list. For a cluster of N elements, all N servers must
187 be defined in redis_servers.
189 GF/02 Catch and remove uninitialized value warning in exiqsumm
190 Check for existence of @ARGV before looking at $ARGV[0]
192 JH/01 Replace the store_release() internal interface with store_newblock(),
193 which internalises the check required to safely use the old one, plus
194 the allocate and data copy operations duplicated in both (!) of the
195 extant use locations.
197 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
198 modifier. This matches the restriction on the commandline.
200 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
201 Previously only the last row was returned.
203 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
204 we assumed that tags in the header were well-formed, and parsed the
205 element content after inspecting only the first char of the tag.
206 Assumptions at that stage could crash the receive process on malformed
209 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
210 While running the DKIM ACL we operate on the Permanent memory pool so that
211 variables created with "set" persist to the DATA ACL. Also (at any time)
212 DNS lookups that fail create cache records using the Permanent pool. But
213 expansions release any allocations made on the current pool - so a dnsdb
214 lookup expansion done in the DKIM ACL releases the memory used for the
215 DNS negative-cache, and bad things result. Solution is to switch to the
216 Main pool for expansions.
217 While we're in that code, add checks on the DNS cache during store_reset,
218 active in the testsuite.
219 Problem spotted, and debugging aided, by Wolfgang Breyha.
221 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
222 When none of the hosts presented to a transport match an already-open
223 connection, close it and proceed with the list. Previously we would
224 queue the message. Spotted by Lena with Yahoo, probably involving
227 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
228 Previously a spurious "250 OK id=" response was appended to the proper
231 JH/08 The "support for" informational output now, which built with Content
232 Scanning support, has a line for the malware scanner interfaces compiled
233 in. Interface can be individually included or not at build time.
235 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
236 by the template makefile "src/EDITME". The "STREAM" support for an older
237 ClamAV interface method is removed.
239 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
240 rows affected is given instead).
242 JH/11 The runtime Berkeley DB library version is now additionally output by
243 "exim -d -bV". Previously only the compile-time version was shown.
245 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
246 SMTP connection. Previously, when one had more receipients than the
247 first, an abortive onward connection was made. Move to full support for
248 multiple onward connections in sequence, handling cutthrough connection
249 for all multi-message initiating connections.
251 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
252 routers. Previously, a multi-recipient message would fail to match the
253 onward-connection opened for the first recipient, and cause its closure.
255 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
256 a timeout on read on a GnuTLS initiating connection, resulting in the
257 initiating connection being dropped. This mattered most when the callout
258 was marked defer_ok. Fix to keep the two timeout-detection methods
261 JH/15 Relax results from ACL control request to enable cutthrough, in
262 unsupported situations, from error to silently (except under debug)
263 ignoring. This covers use with PRDR, frozen messages, queue-only and
266 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
268 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
269 metadata, resulting in a crash in free().
271 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
272 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
273 Broken also in d185889f4, with init system revamp.
275 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
276 Previously we abruptly closed the connection after reading a malware-
277 found indication; now we go on to read the "scan ok" response line,
280 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
281 ACL. Previously, a crash would result.
283 JH/19 Speed up macro lookups during configuration file read, by skipping non-
284 macro text after a replacement (previously it was only once per line) and
285 by skipping builtin macros when searching for an uppercase lead character.
287 JH/20 DANE support moved from Experimental to mainline. The Makefile control
288 for the build is renamed.
290 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
291 was allocated for every new TLS startup, meaning one per message. Fix
292 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
294 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
295 reported the original. Fix to report (as far as possible) the ACL
296 result replacing the original.
298 JH/23 Fix memory leak during multi-message connections using STARTTLS under
299 OpenSSL. Certificate information is loaded for every new TLS startup,
300 and the resources needed to be freed.
302 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
304 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
307 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
308 DATA response info to the (existing) per-recipient response info for
309 the "C=" log element. It can have useful tracking info from the
310 destination system. Patch from Simon Arlott.
312 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
313 length value. Previously this would segfault.
315 HS/02 Support Avast multiline protoocol, this allows passing flags to
316 newer versions of the scanner.
318 JH/28 Ensure that variables possibly set during message acceptance are marked
319 dead before release of memory in the daemon loop. This stops complaints
320 about them when the debug_store option is enabled. Discovered specifically
321 for sender_rate_period, but applies to a whole set of variables.
322 Do the same for the queue-runner and queue-list loops, for variables set
323 from spool message files. Do the same for the SMTP per-message loop, for
324 certain variables indirectly set in ACL operations.
326 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
327 as a multi-recipient message from a mailinglist manager). The coding had
328 an arbitrary cutoff number of characters while checking for more input;
329 enforced by writing a NUL into the buffer. This corrupted long / fast
330 input. The problem was exposed more widely when more pipelineing of SMTP
331 responses was introduced, and one Exim system was feeding another.
332 The symptom is log complaints of SMTP syntax error (NUL chars) on the
333 receiving system, and refused recipients seen by the sending system
334 (propating to people being dropped from mailing lists).
335 Discovered and pinpointed by David Carter.
337 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
338 replaced by the ${authresults } expansion.
340 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
342 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
343 allows proper process termination in container environments.
345 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
346 Previously the "final dot" had a newline after it; ensure it is CR,LF.
348 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
349 and "err_perm", deprecated since 4.83 when the RFC-defined words
350 "temperror" and "permerror" were introduced.
352 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
353 transport-filters or DKIM-signing. The restriction was lost in the
354 consolidation of verify-callout and delivery SMTP handling.
355 Extend the restriction to also cover ARC-signing.
357 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
358 in defer=pass mode supply a 450 to the initiator. Previously the message
361 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
362 tls_require_ciphers is used as before.
364 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
365 "pass_unscanned". Only tmpfails from the scanner are written to
366 the paniclog, as they may require admin intervention (permission
367 denied, license issues). Other scanner errors (like decompression
368 bombs) do not cause a paniclog entry.
370 JH/36 Fix reinitialisation of DKIM logging variable between messages.
371 Previously it was possible to log spurious information in receive log
374 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
375 triggered odd behaviour from Outlook Express clients.
377 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
380 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
381 since the IETF WG has not yet settled on that versus the original
382 "bare" representation.
384 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
385 Previously the millisecond value corrupted the output.
386 Fix also for syslog_pid=no and log_selector +pid, for which the pid
387 corrupted the output.
393 JH/01 Rework error string handling in TLS interface so that the caller in
394 more cases is responsible for logging. This permits library-sourced
395 string to be attached to addresses during delivery, and collapses
396 pairs of long lines into single ones.
398 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
399 during configuration. Wildcards are allowed and expanded.
401 JH/02 Rework error string handling in DKIM to pass more info back to callers.
402 This permits better logging.
404 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
405 do not close it down and have the child transport start it up again on
406 the passed-on TCP connection. Instead, proxy the child (and any
407 subsequent ones) for TLS via a unix-domain socket channel. Logging is
408 affected: the continued delivery log lines do not have any DNSSEC, TLS
409 Certificate or OCSP information. TLS cipher information is still logged.
411 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
412 identical IP addresses on different listening ports. Will also affect
415 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
416 add noisy ifdef guards to special-case this sillyness.
417 Patch from Bernd Kuhls.
419 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
420 than 255 are no longer allowed.
422 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
423 Disable the session-cache too, which might reduce our load. Since we
424 currrectly use a new context for every connection, both as server and
425 client, there is no benefit for these.
426 GnuTLS appears to not support tickets server-side by default (we don't
427 call gnutls_session_ticket_enable_server()) but client side is enabled
428 by default on recent versions (3.1.3 +) unless the PFS priority string
431 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
432 <https://reproducible-builds.org/specs/source-date-epoch/>.
434 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
435 the check for any unsuccessful recipients did not notice the limit, and
436 erroneously found still-pending ones.
438 JH/08 Pipeline CHUNKING command and data together, on kernels that support
439 MSG_MORE. Only in-clear (not on TLS connections).
441 JH/09 Avoid using a temporary file during transport using dkim. Unless a
442 transport-filter is involved we can buffer the headers in memory for
443 creating the signature, and read the spool data file once for the
444 signature and again for transmission.
446 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
447 4.77 as the kernel support then wasn't solid, having issues in 64bit
448 mode. Now, it's been long enough. Add support for FreeBSD also.
450 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
451 case where the routing stage had gathered several addresses to send to
452 a host before calling the transport for the first, we previously failed
453 to close down TLS in the old transport process before passing the TCP
454 connection to the new process. The new one sent a STARTTLS command
455 which naturally failed, giving a failed delivery and bloating the retry
456 database. Investigation and fix prototype from Wolfgang Breyha.
458 JH/12 Fix check on SMTP command input synchronisation. Previously there were
459 false-negatives in the check that the sender had not preempted a response
460 or prompt from Exim (running as a server), due to that code's lack of
461 awareness of the SMTP input buffering.
463 PP/04 Add commandline_checks_require_admin option.
464 Exim drops privileges sanely, various checks such as -be aren't a
465 security problem, as long as you trust local users with access to their
466 own account. When invoked by services which pass untrusted data to
467 Exim, this might be an issue. Set this option in main configuration
468 AND make fixes to the calling application, such as using `--` to stop
471 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
472 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
473 responses to those, into a single TLS record each way (this usually means
474 a single packet). As a side issue, smtp_enforce_sync now works on TLS
477 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
478 affects you only if you're dancing at the edge of the param size limits.
479 If you are, and this message makes sense to you, then: raise the
480 configured limit or use OpenSSL 1.1. Nothing we can do for older
483 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
484 cmdline element to get the documented default one. Previously it was
487 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
488 are unsuable for non-message-specific reasons - eg. connection timeout,
489 banner-time rejection.
491 JH/16 Fix logging of delivery remote port, when specified by router, under
494 PP/06 Repair manualroute's ability to take options in any order, even if one
495 is the name of a transport.
498 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
500 JH/17 Change the list-building routines interface to use the expanding-string
501 triplet model, for better allocation and copying behaviour.
503 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
504 Previously it was constructed the first time a possibly-matching string
505 was met in the configuration file input during startup; now it is done
508 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
509 compatible one, to avoid the (poorly documented) possibility of a config
510 file in the working directory redirecting the DB files, possibly correpting
511 some existing file. CVE-2017-10140 assigned for BDB.
513 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
514 cache-hot. Previously, although the result was properly cached, the
515 initial verify call returned a defer.
517 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
518 the main verify for receipient in uncached-mode.
520 JH/22 Retire historical build files to an "unsupported" subdir. These are
521 defined as "ones for which we have no current evidence of testing".
523 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
524 if present. Previously it was ignored.
526 JH/24 Start using specified-initialisers in C structure init coding. This is
527 a C99 feature (it's 2017, so now considered safe).
529 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
530 if was a fixed-sized field and bitmask ops via macros; it is now more
533 PP/07 GitHub PR 56: Apply MariaDB build fix.
534 Patch provided by Jaroslav Å karvada.
536 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
537 during Coverity cleanups [4.87 JH/47]
538 Diagnosis and fix provided by Michael Fischer v. Mollard.
540 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
541 the right size to place the terminating semicolon on its own folded
542 line, the header hash was calculated to an incorrect value thanks to
543 the (relaxed) space the fold became.
545 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
546 and confused the parent.
548 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
549 which could crash as a result. This could lead to undeliverable messages.
551 JH/28 Logging: "next input sent too soon" now shows where input was truncated
554 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
555 matters on fast-turnover and PID-randomising systems, which were getting
556 out-of-order delivery.
558 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
559 a possibly-overlapping copy. The symptom was that "Remote host closed
560 connection in response to HELO" was logged instead of the actual 4xx
563 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
564 Previously only that bufferd was discarded, resulting in SYMTP command
567 JH/32 DKIM: when a message has multiple signatures matching an identity given
568 in dkim_verify_signers, run the dkim acl once for each. Previously only
569 one run was done. Bug 2189.
571 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
572 "panic the current process" to "deliberately defer". The panic log is
573 still written with the problem list name; the mail and reject logs now
574 get a temp-reject line for the message that was being handled, saying
575 something like "domains check lookup or other defer". The SMTP 451
576 message is still "Temporary local problem".
578 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
579 A crafted sequence of BDAT commands could result in in-use memory beeing
580 freed. CVE-2017-16943.
582 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
583 from SMTP input. Previously it was always done; now only done for DATA
584 and not BDAT commands. CVE-2017-16944.
586 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
587 to the message (such as an overlong header line). Previously this was
588 not done and we did not exit BDAT mode. Followon from the previous item
589 though a different problem.
595 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
596 than -2003 did; needs libidn2 in addition to libidn.
598 JH/02 The path option on a pipe transport is now expanded before use.
600 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
601 Patch provided by "Björn", documentation fix added too.
603 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
604 missing a wire-to-host endian conversion.
606 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
607 close after a BDAT command line could be taken as a following command,
608 giving a synch failure. Fix by only checking for synch immediately
609 before acknowledging the chunk.
611 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
612 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
613 macro. Patches provided by Josh Soref.
615 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
616 Previously we did not; the RFC seems ambiguous and VRFY is not listed
617 by IANA as a service extension. However, John Klensin suggests that we
620 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
621 the dkim code may be unix-mode line endings rather than smtp wire-format
622 CRLF, so prepend a CR to any bare LF.
624 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
625 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
627 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
628 the first were themselves being wrongly included in the feed into dkim
629 processing; with most chunk sizes in use this resulted in an incorrect
630 body hash calculated value.
632 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
633 DKIM signature block, for verification. Although advised against by
634 standards it is specifically not ruled illegal.
636 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
638 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
639 missing a body hash (the bh= tag).
641 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
642 It seems that HAProxy sends the Proxy Protocol information in clear and
643 only then does a TLS startup, so do the same.
645 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
646 TCP connections (such as for Spamd) unless the daemon successfully set
647 Fast Open mode on its listening sockets. This fixes breakage seen on
648 too-old kernels or those not configured for Fast Open, at the cost of
649 requiring both directions being enabled for TFO, and TFO never being used
650 by non-daemon-related Exim processes.
652 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
653 endings, at least on the first header line. Try to canonify any that get
654 past that check, despite the cost.
656 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
657 now limited to an arbitrary five deep, while parsing addresses with the
658 strip_excess_angle_brackets option enabled.
660 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
661 instead leave the unprompted TLS handshake in socket buffer for the
662 TLS library to consume.
664 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
666 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
668 JH/16 Drop variables when they go out of scope. Memory management drops a whole
669 region in one operation, for speed, and this leaves assigned pointers
670 dangling. Add checks run only under the testsuite which checks all
671 variables at a store-reset and panics on a dangling pointer; add code
672 explicitly nulling out all the variables discovered. Fixes one known
673 bug: a transport crash, where a dangling pointer for $sending_ip_address
674 originally assigned in a verify callout, is re-used.
676 PP/06 Drop '.' from @INC in various Perl scripts.
678 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
680 PP/08 Reduce a number of compilation warnings under clang; building with
681 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
682 should be warning-free.
684 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
686 HS/01 Fix portability problems introduced by PP/08 for platforms where
687 realloc(NULL) is not equivalent to malloc() [SunOS et al].
689 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
690 chunk. This allows us to accept broken chunked messages. We need a more
691 general solution here.
693 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
694 already-broken messages in the queue.
696 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
698 JH/19 Fix reference counting bug in routing-generated-address tracking.
704 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
705 supports it and a size is available (ie. the sending peer gave us one).
707 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
708 years of being deprecated). The replacements are the ACLs
709 acl_smtp_mime and acl_not_smtp_mime.
711 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
712 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
713 or even an in-clear connection were permitted. Now, if the host lookup
714 was dnssec and dane was requested then the host is only used if the
715 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
716 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
717 if one fails this test.
718 This means that a poorly-configured remote DNS will make it incommunicado;
719 but it protects against a DNS-interception attack on it.
721 JH/04 Bug 1810: make continued-use of an open smtp transport connection
722 non-noisy when a race steals the message being considered.
724 JH/05 If main configuration option tls_certificate is unset, generate a
725 self-signed certificate for inbound TLS connections.
727 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
728 in rewrites and routers.
730 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
731 and logged a warning sing 4.83; now they are a configuration file error.
733 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
734 (lacking @domain). Apply the same qualification processing as RCPT.
736 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
738 JH/10 Support ${sha256:} applied to a string (as well as the previous
741 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
742 a cutthrough deliver is pending, as we always want to make a connection.
743 This also avoids re-routing the message when later placing the cutthrough
744 connection after a verify cache hit.
745 Do not update it with the verify result either.
747 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
748 when routing results in more than one destination address.
750 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
751 signing (which inhibits the cutthrough capability). Previously only
752 the presence of an option was tested; now an expansion evaluating as
753 empty is permissible (obviously it should depend only on data available
754 when the cutthrough connection is made).
756 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
757 the relevant preceding SMTP command did not note the pipelining mode.
759 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
760 Previously they were not counted.
762 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
763 as one having no matching records. Previously we deferred the message
764 that needed the lookup.
766 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
767 distinguished as "(=".
769 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
770 for missing MX records. Previously it only worked for missing A records.
772 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
774 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
775 after the data-go-ahead and data-ack. Patch from Jason Betts.
777 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
778 even for a "none" policy. Patch from Tony Meyer.
780 JH/22 Fix continued use of a connection for further deliveries. If a port was
781 specified by a router, it must also match for the delivery to be
784 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
785 When one of the recipients of a message was unsuitable for the connection
786 (has no matching addresses), we lost track of needing to mark it
787 deferred. As a result mail would be lost.
789 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
791 JH/25 Decoding ACL controls is now done using a binary search; the source code
792 takes up less space and should be simpler to maintain. Merge the ACL
793 condition decode tables also, with similar effect.
795 JH/26 Fix problem with one_time used on a redirect router which returned the
796 parent address unchanged. A retry would see the parent address marked as
797 delivered, so not attempt the (identical) child. As a result mail would
800 JH/27 Fix a possible security hole, wherein a process operating with the Exim
801 UID can gain a root shell. Credit to http://www.halfdog.net/ for
802 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
805 JH/28 Enable {spool,log} filesystem space and inode checks as default.
806 Main config options check_{log,spool}_{inodes,space} are now
807 100 inodes, 10MB unless set otherwise in the configuration.
809 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
810 Previously it only applied to the main-section connection policy
813 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
815 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
816 by me. Added RFC7919 DH primes as an alternative.
818 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
819 are not in the system include path.
821 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
822 GnuTLS, when a session startup failed (eg because the client disconnected)
823 Exim did stdio operations after fclose. This was exposed by a recent
824 change which nulled out the file handle after the fclose.
826 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
827 signed directly by the cert-signing cert, rather than an intermediate
828 OCSP-signing cert. This is the model used by LetsEncrypt.
830 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
832 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
833 an incoming connection.
835 HS/02 Bug 1802: Do not half-close the connection after sending a request
838 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
839 fallback to "prime256v1".
841 JH/34 SECURITY: Use proper copy of DATA command in error message.
842 Could leak key material. Remotely exploitable. CVE-2016-9963.
848 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
849 and 3.4.4 - once the server is enabled to respond to an OCSP request
850 it does even when not requested, resulting in a stapling non-aware
851 client dropping the TLS connection.
853 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
854 support variable-length bit vectors. No functional change.
856 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
857 The I= interface field on outgoing lines is now after the H= remote
858 host field, same as incoming lines. There is a separate
859 outgoing_interface log selector which allows you to disable the
862 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
863 If not running log_selector +smtp_connection the mainlog would be held
864 open indefinitely after a "too many connections" event, including to a
865 deleted file after a log rotate. Leave the per net connection logging
866 leaving it open for efficiency as that will be quickly detected by the
867 check on the next write.
869 HS/01 Bug 1671: Fix post transport crash.
870 Processing the wait-<transport> messages could crash the delivery
871 process if the message IDs didn't exist for some reason. When
872 using 'split_spool_directory=yes' the construction of the spool
873 file name failed already, exposing the same netto behaviour.
875 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
876 mime_regex ACL conditions.
878 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
879 to DSN fail messages (bounces): remote IP, remote greeting, remote response
880 to HELO, local diagnostic string.
882 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
883 log line to debug. Even when configured with a tls authenticator many
884 client connections are expected to not authenticate in this way, so
885 an authenticate fail is not an error.
887 HS/02 Add the Exim version string to the process info. This way exiwhat
888 gives some more detail about the running daemon.
890 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
891 matter for fast-change records such as DNSBLs.
893 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
894 retry record, even if constant. There may be multiple transports with
895 different interface settings and the retry behaviour needs to be kept
898 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
900 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
902 JH/10 Bug 840: fix log_defer_output option of pipe transport
904 JH/11 Bug 830: use same host for all RCPTS of a message, even under
905 hosts_randomize. This matters a lot when combined with mua_wrapper.
907 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
908 ${quote_pgsql:<string>} operator.
910 JH/13 Bug 1708: avoid misaligned access in cached lookup.
912 JH/14 Change header file name for freeradius-client. Relevant if compiling
913 with Radius support; from the Gentoo tree and checked under Fedora.
915 JH/15 Bug 1712: Introduce $prdr_requested flag variable
917 JH/16 Bug 1714: Permit an empty string as expansion result for transport
918 option transport_filter, meaning no filtering.
920 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
922 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
923 defaults to "*" (all hosts). The variable is now available when not built
924 with TLS, default unset, mainly to enable keeping the testsuite sane.
925 If a server certificate is not supplied (via tls_certificate) an error is
926 logged, and clients will find TLS connections fail on startup. Presumably
927 they will retry in-clear.
928 Packagers of Exim are strongly encouraged to create a server certificate
929 at installation time.
931 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
932 with the $config_file variable.
934 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
935 in transport context, after the attempt, and per-recipient. The latter type
936 is per host attempted. The event data is the error message, and the errno
937 information encodes the lookup type (A vs. MX) used for the (first) host,
938 and the trailing two digits of the smtp 4xx response.
940 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
941 to write to mainlog (or rejectlog, paniclog) in the window between file
942 creation and permissions/ownership being changed. Particularly affects
943 installations where exicyclog is run as root, rather than exim user;
944 result is that the running daemon panics and dies.
946 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
948 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
949 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
950 "pri" and "weight". Note that the previous implicit priority given by the
951 list order is no longer honoured.
953 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
956 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
957 by defining SUPPORT_SOCKS.
959 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
960 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
961 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
962 variables are renamed to proxy_{local,external}_{address,port}.
964 JH/27 Move Internationalisation support from Experimental to mainline, enabled
965 for a build by defining SUPPORT_I18N
967 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
968 of the query string, and make ${quote_redis:} do that quoting.
970 JH/29 Move Events support from Experimental to mainline, enabled by default
971 and removable for a build by defining DISABLE_EVENT.
973 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
975 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
976 cached by the daemon.
978 JH/32 Move Redis support from Experimental to mainline, enabled for a build
979 by defining LOOKUP_REDIS. The libhiredis library is required.
981 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
982 keys are given for lookup.
984 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
985 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
986 only supported when built with TLS support. The PolarSSL SHA routines
987 are still used when the TLS library is too old for convenient support.
989 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
990 openssl_options), for security. OpenSSL forces this from version 1.1.0
991 server-side so match that on older versions.
993 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
994 allocation for $value could be released as the expansion processing
995 concluded, but leaving the global pointer active for it.
997 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
998 and to use the domains and local_parts ACL conditions.
1000 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1001 incorrectly not doubled on cutthrough transmission, hence seen as a
1002 body-termination at the receiving system - resulting in truncated mails.
1003 Commonly the sender saw a TCP-level error, and retransmitted the message
1004 via the normal store-and-forward channel. This could result in duplicates
1005 received - but deduplicating mailstores were liable to retain only the
1006 initial truncated version.
1008 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1010 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1012 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1013 we're in there, support oversigning also; bug 1309.
1015 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1017 HS/04 Add support for keep_environment and add_environment options.
1019 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1020 either intentional arithmetic overflow during PRNG, or testing config-
1023 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1024 delivery resulted in actual delivery. Cancel cutthrough before DATA
1027 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1028 rejects a recipient: pass the reject to the originator.
1030 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1031 Many were false-positives and ignorable, but it's worth fixing the
1034 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1035 for the new environment-manipulation done at startup. Move the routines
1036 from being local to tls.c to being global via the os.c file.
1038 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1039 an extract embedded as result-arg for a map, the first arg for extract
1040 is unavailable so we cannot tell if this is a numbered or keyed
1041 extraction. Accept either.
1047 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1050 JH/02 The smtp transport option "multi_domain" is now expanded.
1052 JH/03 The smtp transport now requests PRDR by default, if the server offers
1055 JH/04 Certificate name checking on server certificates, when exim is a client,
1056 is now done by default. The transport option tls_verify_cert_hostnames
1057 can be used to disable this per-host. The build option
1058 EXPERIMENTAL_CERTNAMES is withdrawn.
1060 JH/05 The value of the tls_verify_certificates smtp transport and main options
1061 default to the word "system" to access the system default CA bundle.
1062 For GnuTLS, only version 3.0.20 or later.
1064 JH/06 Verification of the server certificate for a TLS connection is now tried
1065 (but not required) by default. The verification status is now logged by
1066 default, for both outbound TLS and client-certificate supplying inbound
1069 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1072 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1073 Status Notification (bounce) messages are now MIME format per RFC 3464.
1074 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1075 under the control of the dsn_advertise_hosts option, and routers may
1076 have a dsn_lasthop option.
1078 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1079 default, modifiable by a malware= option. The list separator for
1080 the options can now be changed in the usual way. Bug 68.
1082 JH/10 The smtp_receive_timeout main option is now expanded before use.
1084 JH/11 The incoming_interface log option now also enables logging of the
1085 local interface on delivery outgoing connections.
1087 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1088 if the interface and destination host and port all match.
1090 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1093 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1094 Patch from Andrew Lewis.
1096 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1097 now supports optional time-restrictions, weighting, and priority
1098 modifiers per server. Patch originally by <rommer@active.by>.
1100 JH/16 The spamd_address main option now supports a mixed list of local
1101 and remote servers. Remote servers can be IPv6 addresses, and
1102 specify a port-range.
1104 JH/17 Bug 68: The spamd_address main option now supports an optional
1105 timeout value per server.
1107 JH/18 Bug 1581: Router and transport options headers_add/remove can
1108 now have the list separator specified.
1110 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1113 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1116 JH/21 Support for the A6 type of dns record is withdrawn.
1118 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1119 rather than the verbs used.
1121 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1122 from 255 to 1024 chars.
1124 JH/24 Verification callouts now attempt to use TLS by default.
1126 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1127 are generic router options now. The defaults didn't change.
1129 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1130 Original patch from Alexander Shikoff, worked over by JH.
1132 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1133 files if ZCAT_COMMAND is not executable.
1135 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1137 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1139 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1140 Normally benign, it bites when the pair was led to by a CNAME;
1141 modern usage is to not canonicalize the domain to a CNAME target
1142 (and we were inconsistent anyway for A-only vs AAAA+A).
1144 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1146 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1147 when evaluating $sender_host_dnssec.
1149 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1150 $sender_helo_dnssec variable.
1152 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1154 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1156 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1158 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1159 documented as working, but never had. Support all but $spam_report.
1161 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1162 added for tls authenticator.
1164 HS/03 Add perl_taintmode main config option
1170 TL/01 When running the test suite, the README says that variables such as
1171 no_msglog_check are global and can be placed anywhere in a specific
1172 test's script, however it was observed that placement needed to be near
1173 the beginning for it to behave that way. Changed the runtest perl
1174 script to read through the entire script once to detect and set these
1175 variables, reset to the beginning of the script, and then run through
1176 the script parsing/test process like normal.
1178 TL/02 The BSD's have an arc4random API. One of the functions to induce
1179 adding randomness was arc4random_stir(), but it has been removed in
1180 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1181 function when detected.
1183 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1184 cause callback expansion.
1186 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1187 syntax errors in an expansion can be treated as a string instead of
1188 logging or causing an error, due to the internal use of bool_lax
1189 instead of bool when processing it.
1191 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1192 server certificates when making smtp deliveries.
1194 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1196 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1198 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1200 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1201 Merged patch from Sebastian Wiedenroth.
1203 JH/05 Fix results-pipe from transport process. Several recipients, combined
1204 with certificate use, exposed issues where response data items split
1205 over buffer boundaries were not parsed properly. This eventually
1206 resulted in duplicates being sent. This issue only became common enough
1207 to notice due to the introduction of connection certificate information,
1208 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1210 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1211 size buffer was used, resulting in syntax errors when an expansion
1214 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1215 version 3.3.6 or later.
1217 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1218 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1219 both become "event_action", the variables become $event_name, $event_data
1220 and $event_defer_errno. There is a new variable $verify_mode, usable in
1221 routers, transports and related events. The tls:cert event is now also
1222 raised for inbound connections, if the main configuration event_action
1225 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1226 early OCSP support, but no stapling (appears to be less than 1.0.0).
1228 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1229 server certificate names available under the smtp transport option
1230 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1233 JH/10 Time-related extraction expansions from certificates now use the main
1234 option "timezone" setting for output formatting, and are consistent
1235 between OpenSSL and GnuTLS compilations. Bug 1541.
1237 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1238 encoded parameter in the incoming message. Bug 1558.
1240 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1241 include certificate info, eximon was claiming there were spoolfile
1244 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1246 JH/14 Log delivery-related information more consistently, using the sequence
1247 "H=<name> [<ip>]" wherever possible.
1249 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1250 are problematic for Debian distribution, omit them from the release
1253 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1255 JH/16 Fix string representation of time values on 64bit time_t architectures.
1258 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1259 output list separator was used.
1264 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1265 checkers that were complaining about end of non-void function with no
1268 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1269 This was a regression introduced in 4.83 by another bugfix.
1271 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1273 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1274 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1280 TF/01 Correctly close the server side of TLS when forking for delivery.
1282 When a message was received over SMTP with TLS, Exim failed to clear up
1283 the incoming connection properly after forking off the child process to
1284 deliver the message. In some situations the subsequent outgoing
1285 delivery connection happened to have the same fd number as the incoming
1286 connection previously had. Exim would try to use TLS and fail, logging
1287 a "Bad file descriptor" error.
1289 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1290 utilities have not been installed.
1292 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1293 temporary space as the ACL may create new global variables.
1295 TL/01 LDAP support uses per connection or global context settings, depending
1296 upon the detected version of the libraries at build time.
1298 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1299 to extract and use the src ip:port in logging and expansions as if it
1300 were a direct connection from the outside internet. PPv2 support was
1301 updated based on HAProxy spec change in May 2014.
1303 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1305 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1306 Properly escape header and check for NULL return.
1308 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1311 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1313 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1314 characters in header names, implemented as a verify condition.
1315 Contributed by Michael Fischer v. Mollard.
1317 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1318 results permerror and temperror. Previous values are deprecated but
1319 still accepted. In a future release, err_perm and err_temp will be
1320 completely removed, which will be a backward incompatibility if the
1321 ACL tests for either of these two old results. Patch contributed by
1322 user bes-internal on the mailing list.
1324 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1326 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1327 selectors, in both main and reject logs.
1329 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1332 JH/07 Add malware type "sock" for talking to simple daemon.
1334 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1336 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1337 routers/transports under cutthrough routing.
1339 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1340 numbers. Touch up "bool" conditional to keep the same definition.
1342 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1344 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1346 JH/12 Expand items in router/transport headers_add or headers_remove lists
1347 individually rather than the list as a whole. Bug 1452.
1349 Required for reasonable handling of multiple headers_ options when
1350 they may be empty; requires that headers_remove items with embedded
1351 colons must have them doubled (or the list-separator changed).
1353 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1354 view the policy declared in the DMARC record. Currently, $dmarc_status
1355 is a combined value of both the record presence and the result of the
1358 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1360 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1361 dnslookup router and the smtp transport (applying to the forward
1364 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1365 of ldap servers used for a specific lookup. Patch provided by Heiko
1368 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1369 New variable $lookup_dnssec_authenticated for observability.
1371 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1372 Patch submitted by Lars Timman.
1374 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1376 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1377 Requires trusted mode and valid format message id, aborts otherwise.
1378 Patch contributed by Heiko Schlichting.
1380 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1381 certextract with support for various fields. Bug 1358.
1383 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1384 is requested by default, modifiable by smtp transport option
1387 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1388 operate on certificate variables to give certificate fingerprints
1389 Also new ${sha256:cert_variable}.
1391 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1393 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1394 Christian Aistleitner.
1396 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1398 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1399 file. Patch from Wolfgang Breyha.
1401 JH/25 Expand the coverage of the delivery $host and $host_address to
1402 client authenticators run in verify callout. Bug 1476.
1404 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1405 align with daemon_smtp_ports. Bug 72.
1407 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1408 support and error reporting did not work properly.
1410 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1411 and is readable. Patch from Andrew Colin Kissa.
1413 TL/14 Enhance documentation of ${run expansion and how it parses the
1414 commandline after expansion, particularly in the case when an
1415 unquoted variable expansion results in an empty value.
1417 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1419 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1420 by renaming away from it. Addresses GH issue 15, reported by
1423 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1425 TL/15 SECURITY: prevent double expansion in math comparison functions
1426 (can expand unsanitized data). Not remotely exploitable.
1433 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1435 PP/02 Make -n do something, by making it not do something.
1436 When combined with -bP, the name of an option is not output.
1438 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1441 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1442 $sender_host_name and config options to manage this, and basic check
1445 PP/05 DSCP support for outbound connections and control modifier for inbound.
1447 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1448 (Only plugin which currently uses this is kerberos4, which nobody should
1449 be using, but we should make it available and other future plugins might
1450 conceivably use it, even though it would break NAT; stuff *should* be
1451 using channel bindings instead).
1453 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1454 name; added for Sendmail compatibility; requires admin caller.
1455 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1456 just ignore it); requires trusted caller.
1457 Also parse but ignore: -Ac -Am -X<logfile>
1460 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1462 TL/02 Add +smtp_confirmation as a default logging option.
1464 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1465 Patch by Magnus Holmgren from 2007-02-20.
1467 TL/04 Bugzilla 1281 - Spec typo.
1468 Bugzilla 1283 - Spec typo.
1469 Bugzilla 1290 - Spec grammar fixes.
1471 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1473 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1475 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1476 mailing list by by Dmitry Isaikin.
1478 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1480 JH/02 Support "G" suffix to numbers in ${if comparisons.
1482 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1484 NM/01 Bugzilla 1197 - Spec typo
1485 Bugzilla 1196 - Spec examples corrections
1487 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1489 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1490 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1493 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1494 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1496 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1497 "acl {{name}{arg}...}", and optional args on acl condition
1500 JH/05 Permit multiple router/transport headers_add/remove lines.
1502 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1504 JH/07 Avoid using a waiting database for a single-message-only transport.
1505 Performance patch from Paul Fisher. Bugzilla 1262.
1507 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1510 JH/09 Add $headers_added variable, with content from use of ACL modifier
1511 add_header (but not yet added to the message). Bugzilla 199.
1513 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1514 Pulled from Bugzilla 817 by Wolfgang Breyha.
1516 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1518 (nb: this is the same fix as in Exim 4.80.1)
1520 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1523 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1526 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1528 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1529 advertise SMTP AUTH mechanism to us, instead of a generic
1530 protocol violation error. Also, make Exim more robust to bad
1531 data from the Dovecot auth socket.
1533 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1535 When a queue runner is handling a message, Exim first routes the
1536 recipient addresses, during which it prunes them based on the retry
1537 hints database. After that it attempts to deliver the message to
1538 any remaining recipients. It then updates the hints database using
1541 So if a recipient address works intermittently, it can get repeatedly
1542 deferred at routing time. The retry hints record remains fresh so the
1543 address never reaches the final cutoff time.
1545 This is a fairly common occurrence when a user is bumping up against
1546 their storage quota. Exim had some logic in its local delivery code
1547 to deal with this. However it did not apply to per-recipient defers
1548 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1550 This change adds a proper retry rule check during routing so that the
1551 final cutoff time is checked against the message's age. We only do
1552 this check if there is an address retry record and there is not a
1553 domain retry record; this implies that previous attempts to handle
1554 the address had the retry_use_local_parts option turned on. We use
1555 this as an approximation for the destination being like a local
1556 delivery, as in LMTP.
1558 I suspect this new check makes the old local delivery cutoff check
1559 redundant, but I have not verified this so I left the code in place.
1561 TF/02 Correct gecos expansion when From: is a prefix of the username.
1563 Test 0254 submits a message to Exim with the header
1567 When I ran the test suite under the user fanf2, Exim expanded
1568 the header to contain my full name, whereas it should have added
1569 a Resent-Sender: header. It erroneously treats any prefix of the
1570 username as equal to the username.
1572 This change corrects that bug.
1574 GF/01 DCC debug and logging tidyup
1575 Error conditions log to paniclog rather than rejectlog.
1576 Debug lines prefixed by "DCC: " to remove any ambiguity.
1578 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1580 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1581 Bug spotted by Jeremy Harris; was flawed since initial commit.
1582 Would have resulted in OCSP responses post-SNI triggering an Exim
1583 NULL dereference and crash.
1585 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1587 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1588 Bug detection, analysis and fix by Samuel Thibault.
1589 Bugzilla 1331, Debian bug #698092.
1591 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1593 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1594 Server implementation by Todd Lyons, client by JH.
1595 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1596 config variable "prdr_enable" controls whether the server
1597 advertises the facility. If the client requests PRDR a new
1598 acl_data_smtp_prdr ACL is called once for each recipient, after
1599 the body content is received and before the acl_smtp_data ACL.
1600 The client is controlled by both of: a hosts_try_prdr option
1601 on the smtp transport, and the server advertisement.
1602 Default client logging of deliveries and rejections involving
1603 PRDR are flagged with the string "PRDR".
1605 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1606 fclose(). Diagnosis by Todd Lyons.
1608 PP/17 Update configure.default to handle IPv6 localhost better.
1609 Patch by Alain Williams (plus minor tweaks).
1612 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1613 This is now consistent with GnuTLS, and is now documented: the
1614 previous undocumented portable approach to treating the option as
1615 unset was to force an expansion failure. That still works, and
1616 an empty string is now equivalent.
1618 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1619 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1620 not performing validation itself.
1622 PP/20 Added force_command boolean option to pipe transport.
1623 Patch from Nick Koston, of cPanel Inc.
1625 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1628 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1630 PP/21 Fix eximon continuous updating with timestamped log-files.
1631 Broken in a format-string cleanup in 4.80, missed when I repaired the
1632 other false fix of the same issue.
1633 Report and fix from Heiko Schlichting.
1636 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1637 Report from Prashanth Katuri.
1639 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1640 It's SecureTransport, so affects any MacOS clients which use the
1641 system-integrated TLS libraries, including email clients.
1643 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1644 using a MIME ACL for non-SMTP local injection.
1645 Report and assistance in diagnosis by Warren Baker.
1647 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1649 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1651 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1652 last id that failed so it may be referenced in subsequent ACL's.
1654 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1657 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1658 ldap library initialization, allowing self-signed CA's to be
1659 used. Also properly sets require_cert option later in code by
1660 using NULL (global ldap config) instead of ldap handle (per
1661 session). Bug diagnosis and testing by alxgomz.
1663 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1664 the src/util/ subdirectory.
1666 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1667 renamed to Transport Post Delivery Action by Jeremy Harris, as
1670 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1671 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1672 redis_servers = needs to be configured which will be used by the redis
1673 lookup. Patch from Warren Baker, of The Packet Hub.
1675 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1677 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1678 hostname or reverse DNS when processing a host list. Used suggestions
1679 from multiple comments on this bug.
1681 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1683 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1684 Missed a few lines, added it to make the runtest require no keyboard
1687 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1688 contains upper case chars. Make router use caseful_local_part.
1690 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1691 support when GnuTLS has been built with p11-kit.
1697 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1699 This, or similar/improved, will also be change PP/11 of 4.82.
1705 PP/01 Handle short writes when writing local log-files.
1706 In practice, only affects FreeBSD (8 onwards).
1707 Bugzilla 1053, with thanks to Dmitry Isaikin.
1709 NM/01 Bugzilla 949 - Documentation tweak
1711 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1714 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1716 PP/02 Implemented gsasl authenticator.
1718 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1720 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1721 `pkg-config foo` for cflags/libs.
1723 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1724 with rest of GSASL and with heimdal_gssapi.
1726 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1727 `pkg-config foo` for cflags/libs for the TLS implementation.
1729 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1730 properties get this fed in as external SSF. A number of robustness
1731 and debugging improvements to the cyrus_sasl authenticator.
1733 PP/08 cyrus_sasl server now expands the server_realm option.
1735 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1736 Patch by Jeremy Harris.
1738 PP/10 Added dbmjz lookup type.
1740 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1742 PP/12 MAIL args handles TAB as well as SP, for better interop with
1743 non-compliant senders.
1744 Analysis and variant patch by Todd Lyons.
1746 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1747 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1748 Patch from Dirk Mueller <dmueller@suse.com>
1750 PP/13 tls_peerdn now print-escaped for spool files.
1751 Observed some $tls_peerdn in wild which contained \n, which resulted
1752 in spool file corruption.
1754 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1755 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1756 or write after TLS renegotiation, which otherwise led to messages
1759 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1760 as a tracking header (ie: a signed header comes before the signature).
1761 Patch from Wolfgang Breyha.
1763 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1764 comma-sep list; embedded commas doubled.
1766 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1768 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1770 Report and patch from Dmitry Banschikov.
1772 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1773 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1774 failures. We appear to now support TLS1.1+ with Exim.
1776 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1777 lets Exim select keys and certificates based upon TLS SNI from client.
1778 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1779 before an outbound SMTP session. New log_selector, +tls_sni.
1781 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1782 NULL dereference. Report and patch from Alun Jones.
1784 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1785 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1786 Not seeing resolver debug output on NetBSD, but suspect this is a
1787 resolver implementation change.
1789 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1790 Left warnings. Added "eximon gdb" invocation mode.
1792 PP/21 Defaulting "accept_8bitmime" to true, not false.
1794 PP/22 Added -bw for inetd wait mode support.
1796 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1797 locate the relevant includes and libraries. Made this the default.
1799 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1800 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1802 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1803 This may cause build issues on older platforms.
1805 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1806 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1807 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1808 Added SNI support via GnuTLS too.
1809 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1811 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1813 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1814 Applied second patch from Janne, implementing suggestion to default
1815 multiple-strings-in-record handling to match SPF spec.
1817 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1819 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1820 read-only, out of scope).
1821 Patch from Wolfgang Breyha, report from Stuart Northfield.
1823 PP/29 Fix three issues highlighted by clang analyser static analysis.
1824 Only crash-plausible issue would require the Cambridge-specific
1825 iplookup router and a misconfiguration.
1826 Report from Marcin Mirosław.
1828 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1830 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1831 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1832 As part of this, removing so much warning spew let me fix some minor
1833 real issues in debug logging.
1835 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1836 assignment on my part. Fixed.
1838 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1839 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1840 Janne Snabb (who went above and beyond: thank you).
1842 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1843 string otherwise requires a connection and a bunch more work and it's
1844 relatively easy to get wrong. Should also expose TLS library linkage
1847 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1848 64-bit ${eval} (JH/03).
1850 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1851 GNU libc to support some of the 64-bit stuff, should not lead to
1852 conflicts. Defined before os.h is pulled in, so if a given platform
1853 needs to override this, it can.
1855 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1856 protection layer was required, which is not implemented.
1857 Bugzilla 1254, patch from Wolfgang Breyha.
1859 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1860 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1861 tls_dhparam take prime identifiers. Also unbreak combination of
1862 OpenSSL+DH_params+TLSSNI.
1864 PP/39 Disable SSLv2 by default in OpenSSL support.
1870 PP/01 Solaris build fix for Oracle's LDAP libraries.
1871 Bugzilla 1109, patch from Stephen Usher.
1873 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1875 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1878 TF/02 Fix a couple more cases where we did not log the error message
1879 when unlink() failed. See also change 4.74-TF/03.
1881 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1882 lock up or crash if it happened to be inside a call to libc when it
1883 got a SIGUSR1 from exiwhat.
1885 The SIGUSR1 handler appends the current process status to the process
1886 log which is later printed by exiwhat. It used to use the general
1887 purpose logging code to do this, but several functions it calls are
1888 not safe for signals.
1890 The new output code in the SIGUSR1 handler is specific to the process
1891 log, and simple enough that it's easy to inspect for signal safety.
1892 Removing some special cases also simplifies the general logging code.
1893 Removing the spurious timestamps from the process log simplifies
1896 TF/04 Improved ratelimit ACL condition.
1898 The /noupdate option has been deprecated in favour of /readonly which
1899 has clearer semantics. The /leaky, /strict, and /readonly update modes
1900 are mutually exclusive. The update mode is no longer included in the
1901 database key; it just determines when the database is updated. (This
1902 means that when you upgrade Exim will forget old rate measurements.)
1904 Exim now checks that the per_* options are used with an update mode that
1905 makes sense for the current ACL. For example, when Exim is processing a
1906 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1907 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1908 must specify per_mail/readonly. If you omit the update mode it defaults to
1909 /leaky where that makes sense (as before) or /readonly where required.
1911 The /noupdate option is now undocumented but still supported for
1912 backwards compatibility. It is equivalent to /readonly except that in
1913 ACLs where /readonly is required you may specify /leaky/noupdate or
1914 /strict/noupdate which are treated the same as /readonly.
1916 A useful new feature is the /count= option. This is a generalization
1917 of the per_byte option, so that you can measure the throughput of other
1918 aggregate values. For example, the per_byte option is now equivalent
1919 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1921 The per_rcpt option has been generalized using the /count= mechanism
1922 (though it's more complicated than the per_byte equivalence). When it is
1923 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1924 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1925 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1926 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1927 Note that using per_rcpt with a non-readonly update mode in more than
1928 one ACL will cause the recipients to be double-counted. (The per_mail
1929 and per_byte options don't have this problem.)
1931 The handling of very low rates has changed slightly. If the computed rate
1932 is less than the event's count (usually one) then this event is the first
1933 after a long gap. In this case the rate is set to the same as this event's
1934 count, so that the first message of a spam run is counted properly.
1936 The major new feature is a mechanism for counting the rate of unique
1937 events. The new per_addr option counts the number of different
1938 recipients that someone has sent messages to in the last time period. It
1939 behaves like per_rcpt if all the recipient addresses are different, but
1940 duplicate recipient addresses do not increase the measured rate. Like
1941 the /count= option this is a general mechanism, so the per_addr option
1942 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1943 example, measure the rate that a client uses different sender addresses
1944 with the options per_mail/unique=$sender_address. There are further
1945 details in the main documentation.
1947 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1949 TF/06 Removed a few PCRE remnants.
1951 TF/07 Automatically extract Exim's version number from tags in the git
1952 repository when doing development or release builds.
1954 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1955 Bugzilla 879. Patch from Paul Fisher.
1957 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1958 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1961 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1963 PP/05 Variable $av_failed, true if the AV scanner deferred.
1964 Bugzilla 1078. Patch from John Horne.
1966 PP/06 Stop make process more reliably on build failure.
1967 Bugzilla 1087. Patch from Heiko Schlittermann.
1969 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1970 Bugzilla 1089. Patch from Heiko Schlittermann.
1972 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1973 Bugzilla 1131. Patch from Holger Weiß.
1975 PP/09 Handle IPv6 addresses with SPF.
1976 Bugzilla 860. Patch from Wolfgang Breyha.
1978 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1980 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1983 PP/11 match_* no longer expand right-hand-side by default.
1984 New compile-time build option, EXPAND_LISTMATCH_RHS.
1985 New expansion conditions, "inlist", "inlisti".
1987 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1989 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1991 PP/14 fix log_write() format string regression from TF/03.
1992 Bugzilla 1152. Patch from Dmitry Isaikin.
1998 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2000 PP/02 Harmonised TLS library version reporting; only show if debugging.
2001 Layout now matches that introduced for other libraries in 4.74 PP/03.
2003 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2005 PP/04 New "dns_use_edns0" global option.
2007 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2010 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2011 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2013 TK/01 Updated PolarSSL code to 0.14.2.
2014 Bugzilla 1097. Patch from Andreas Metzler.
2016 PP/07 Catch divide-by-zero in ${eval:...}.
2017 Fixes bugzilla 1102.
2019 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2022 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2023 format-string attack -- SECURITY: remote arbitrary code execution.
2025 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2026 time unintentionally subject to list matching rules, letting the header
2027 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2028 arbitrary string expansion). This allowed for information disclosure.
2030 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2031 INT_MIN/-1 -- value coerced to INT_MAX.
2037 NM/01 Workaround for PCRE version dependency in version reporting
2040 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2041 This fixes portability to compilers other than gcc, notably
2042 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2044 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2045 makefiles for portability to HP-UX and POSIX correctness.
2047 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2048 Also via indented variable definition in the Makefile.
2049 (Debugging by Oliver Heesakkers).
2051 PP/02 Restore caching of spamd results with expanded spamd_address.
2052 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2054 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2055 Improves build reliability. Fix from: Frank Elsner
2057 NM/02 Fix wide character breakage in the rfc2047 coding
2058 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2060 NM/03 Allow underscore in dnslist lookups
2061 Fixes bug 1026. Patch from Graeme Fowler
2063 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2064 Code patches from Adam Ciarcinski of NetBSD.
2066 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2069 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2070 is logged, to avoid truncation. Patch from John Horne.
2072 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2073 Patch from Jakob Hirsch.
2075 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2076 SQL string expansion failure details.
2077 Patch from Andrey Oktyabrski.
2079 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2080 Patch from Simon Arlott.
2082 PP/09 New lookups functionality failed to compile on old gcc which rejects
2083 extern declarations in function scope.
2084 Patch from Oliver Fleischmann
2086 PP/10 Use sig_atomic_t for flags set from signal handlers.
2087 Check getgroups() return and improve debugging.
2088 Fixed developed for diagnosis in bug 927 (which turned out to be
2091 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2092 Patch from Mark Zealey.
2094 PP/12 Bugzilla 1056: Improved spamd server selection.
2095 Patch from Mark Zealey.
2097 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2098 Based on patch from Heiko Schlittermann.
2100 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2101 Patch from Uwe Doering, sign-off by Michael Haardt.
2103 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2104 variable declaration deep within a block. Bug and patch from
2107 PP/15 lookups-Makefile IRIX compatibility coercion.
2109 PP/16 Make DISABLE_DKIM build knob functional.
2111 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2112 Patch by Simon Arlott
2114 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2115 variable argument macros. Our copy now differs from upstream.
2121 TF/01 Failure to get a lock on a hints database can have serious
2122 consequences so log it to the panic log.
2124 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2125 controlled using the smtp_confirmation log selector.
2127 TF/03 Include the error message when we fail to unlink a spool file.
2129 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2130 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2131 for maintaining out-of-tree patches for some time.
2133 PP/01 Bugzilla 139: Documentation and portability issues.
2134 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2135 Handle per-OS dynamic-module compilation flags.
2137 PP/02 Let /dev/null have normal permissions.
2138 The 4.73 fixes were a little too stringent and complained about the
2139 permissions on /dev/null. Exempt it from some checks.
2140 Reported by Andreas M. Kirchwitz.
2142 PP/03 Report version information for many libraries, including
2143 Exim version information for dynamically loaded libraries. Created
2144 version.h, now support a version extension string for distributors
2145 who patch heavily. Dynamic module ABI change.
2147 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2148 privilege escalation vulnerability whereby the Exim run-time user
2149 can cause root to append content of the attacker's choosing to
2152 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2155 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2156 If dropping privileges for untrusted macros, we disabled normal logging
2157 on the basis that it would fail; for the Exim run-time user, this is not
2158 the case, and it resulted in successful deliveries going unlogged.
2159 Fixed. Reported by Andreas Metzler.
2165 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2166 only prepend for the Resent-* case. Fixes regression introduced in
2167 Exim 4.70 by NM/22 for Bugzilla 607.
2169 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2170 increasing numbers of administrators be bitten by this.
2172 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2174 PP/03 Bugzilla 994: added openssl_options main configuration option.
2176 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2178 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2180 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2182 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2183 without a peer certificate, leading to a segfault because of an
2184 assumption that peers always have certificates. Be a little more
2185 paranoid. Problem reported by Martin Tscholak.
2187 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2188 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2189 NB: ClamAV planning to remove STREAM in "middle of 2010".
2190 CL also introduces -bmalware, various -d+acl logging additions and
2191 more caution in buffer sizes.
2193 PP/09 Implemented reverse_ip expansion operator.
2195 PP/10 Bugzilla 937: provide a "debug" ACL control.
2197 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2199 PP/12 Bugzilla 973: Implement --version.
2201 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2203 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2205 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2207 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2208 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2209 ignore trailing whitespace.
2211 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2213 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2216 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2217 Notification from Dr Andrew Aitchison.
2219 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2220 ExtendedDetectionInfo response format.
2221 Notification from John Horne.
2223 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2226 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2227 XSL and documented dependency on system catalogs, with examples of how
2230 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2233 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2234 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2235 configuration file which is writeable by the Exim user or group.
2237 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2238 of configuration files to cover files specified with the -C option if
2239 they are going to be used with root privileges, not just the default
2242 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2243 option (effectively making it always true).
2245 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2246 files to be used while preserving root privileges.
2248 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2249 that rogue child processes cannot use them.
2251 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2252 run-time user, instead of root.
2254 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2255 Exim run-time user without dropping privileges.
2257 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2258 result string, instead of calling string_vformat() twice with the same
2261 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2262 for other users. Others should always drop root privileges if they use
2263 -C on the command line, even for a whitelisted configure file.
2265 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2267 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2273 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2274 $data_path, and $header_path variables; fixed documentation bugs and
2277 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2278 exipick to access non-standard spools, including the "frozen" queue
2281 NM/01 Bugzilla 965: Support mysql stored procedures.
2282 Patch from Alain Williams
2284 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2286 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2287 Patch from Andreas Metzler
2289 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2290 Patch from Kirill Miazine
2292 NM/05 Bugzilla 671: Added umask to procmail example.
2294 JJ/03 installed exipick 20100323.0, fixing doc bug
2296 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2297 directory. Notification and patch from Dan Rosenberg.
2299 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2301 TK/02 Improve log output when DKIM signing operation fails.
2303 MH/01 Treat the transport option dkim_domain as a colon separated
2304 list, not as a single string, and sign the message with each element,
2305 omitting multiple occurences of the same signer.
2307 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2308 Bugzilla 985, 986. Patch by Simon Arlott
2310 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2311 Patch by Simon Arlott
2313 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2314 MBX locking. Notification from Dan Rosenberg.
2320 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2322 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2324 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2326 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2328 NM/04 Bugzilla 915: Flag broken perl installation during build.
2334 TK/01 Added patch by Johannes Berg that expands the main option
2335 "spamd_address" if it starts with a dollar sign.
2337 TK/02 Write list of recipients to X-Envelope-Sender header when building
2338 the mbox-format spool file for content scanning (suggested by Jakob
2341 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2342 (http://www.dcc-servers.net/) support via dccifd. Activated by
2343 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2345 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2346 by Mark Daniel Reidel <mr@df.eu>.
2348 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2349 When building exim an external PCRE library is now needed -
2350 PCRE is a system library on the majority of modern systems.
2351 See entry on PCRE_LIBS in EDITME file.
2353 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2354 conversation. Added nologin parameter to request.
2355 Patch contributed by Kirill Miazine.
2357 TF/01 Do not log submission mode rewrites if they do not change the address.
2359 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2361 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2362 log files in place. Contributed by Roberto Lima.
2364 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2366 TF/03 Bugzilla 615: When checking the local_parts router precondition
2367 after a local_part_suffix or local_part_prefix option, Exim now
2368 does not use the address's named list lookup cache, since this
2369 contains cached lookups for the whole local part.
2371 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2372 Robert Millan. Documentation is in experimental-spec.txt.
2374 TF/04 Bugzilla 668: Fix parallel build (make -j).
2376 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2378 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2379 Patch provided by Jan Srzednicki.
2381 TF/05 Leading white space used to be stripped from $spam_report which
2382 wrecked the formatting. Now it is preserved.
2384 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2385 that they are available at delivery time.
2387 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2389 TF/08 TLS error reporting now respects the incoming_interface and
2390 incoming_port log selectors.
2392 TF/09 Produce a more useful error message if an SMTP transport's hosts
2393 setting expands to an empty string.
2395 NM/06 Bugzilla 744: EXPN did not work under TLS.
2396 Patch provided by Phil Pennock.
2398 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2399 Patch provided by Richard Godbee.
2401 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2402 acl_smtp_notquit, added index entry.
2404 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2405 Patch provided by Eugene Bujak.
2407 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2408 accept(). Patch provided by Maxim Dounin.
2410 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2411 Patch provided by Phil Pennock.
2413 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2415 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2416 Patch provided by Brad "anomie" Jorsch.
2418 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2419 Patch provided by Dean Brooks.
2421 TK/05 Add native DKIM support (does not depend on external libraries).
2423 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2424 Patch provided by Graeme Fowler.
2426 NM/16 Bugzilla 851: Documentation example syntax fix.
2428 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2430 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2433 NM/19 Bugzilla 745: TLS version reporting.
2434 Patch provided by Phil Pennock.
2436 NM/20 Bugzilla 167: bool: condition support.
2437 Patch provided by Phil Pennock.
2439 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2440 clients. Patch provided by Phil Pennock.
2442 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2443 Patch provided by Brad "anomie" Jorsch.
2445 NM/23 Bugzilla 687: Fix misparses in eximstats.
2446 Patch provided by Heiko Schlittermann.
2448 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2449 Patch provided by Heiko Schlittermann.
2451 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2452 plus update to original patch.
2454 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2456 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2457 Patch provided by David Brownlee.
2459 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2461 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2463 NM/30 Bugzilla 888: TLS documentation bugfixes.
2465 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2467 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2468 Unlike the original bugzilla I have changed all shell scripts in src tree.
2470 NM/33 Bugzilla 898: Transport filter timeout fix.
2471 Patch by Todd Rinaldo.
2473 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2474 Patch by Serge Demonchaux.
2476 NM/35 Bugzilla 39: Base64 decode bug fixes.
2477 Patch by Jakob Hirsch.
2479 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2481 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2483 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2485 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2491 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2492 ALT-N's libdkim that I have put here:
2493 http://duncanthrax.net/exim-experimental/
2495 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2496 were called 'true' and it seems that C99 defines that as a reserved
2497 keyword to be used with 'bool' variable types. That means you could
2498 not include C99-style headers which use bools without triggering
2499 build errors in sieve.c.
2501 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2502 as mailq or other aliases. Changed the --help handling significantly
2503 to do whats expected. exim_usage() emits usage/help information.
2505 SC/01 Added the -bylocaldomain option to eximstats.
2507 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2509 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2511 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2517 PH/01 Another patch from the Sieve maintainer.
2519 PH/02 When an IPv6 address is converted to a string for single-key lookup
2520 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2521 dots are used instead of colons so that keys in lsearch files need not
2522 contain colons. This was done some time before quoting was made available
2523 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2524 (notated using the quote facility) so as to distinguish them from IPv4
2525 keys. This meant that lookups for IP addresses in host lists did not work
2526 for iplsearch lookups.
2528 This has been fixed by arranging for IPv6 addresses to be expressed with
2529 colons if the lookup type is iplsearch. This is not incompatible, because
2530 previously such lookups could never work.
2532 The situation is now rather anomalous, since one *can* have colons in
2533 ordinary lsearch keys. However, making the change in all cases is
2534 incompatible and would probably break a number of configurations.
2536 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2539 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2540 conversion specification without a maximum field width, thereby enabling
2541 a rogue spamd server to cause a buffer overflow. While nobody in their
2542 right mind would setup Exim to query an untrusted spamd server, an
2543 attacker that gains access to a server running spamd could potentially
2544 exploit this vulnerability to run arbitrary code as the Exim user.
2546 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2547 $primary_hostname instead of what libspf2 thinks the hosts name is.
2549 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2550 a directory entry by the name of the lookup key. Previously, if a
2551 symlink pointed to a non-existing file or a file in a directory that
2552 Exim lacked permissions to read, a lookup for a key matching that
2553 symlink would fail. Now it is enough that a matching directory entry
2554 exists, symlink or not. (Bugzilla 503.)
2556 PH/03 The body_linecount and body_zerocount variables are now exported in the
2559 PH/04 Added the $dnslist_matched variable.
2561 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2562 This means they are set thereafter only if the connection becomes
2565 PH/06 Added the client_condition to authenticators so that some can be skipped
2566 by clients under certain conditions.
2568 PH/07 The error message for a badly-placed control=no_multiline_responses left
2569 "_responses" off the end of the name.
2571 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2573 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2574 (without spaces) instead of just copying the configuration text.
2576 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2578 PH/11 Added $max_received_linelength.
2580 PH/12 Added +ignore_defer and +include_defer to host lists.
2582 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2583 way in which PCRE > 7.0 is built.
2585 PH/14 Implemented queue_only_load_latch.
2587 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2588 MAIL command. The effect was to mangle the value on 64-bit systems.
2590 PH/16 Another patch from the Sieve maintainer.
2592 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2594 PH/18 If a system quota error occurred while trying to create the file for
2595 a maildir delivery, the message "Mailbox is full" was not appended to the
2596 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2597 only to a quota excession during the actual writing of the file.
2599 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2600 characters?) which causes problems in log lines. The DN values are now
2601 passed through string_printing() before being added to log lines.
2603 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2604 and InterBase are left for another time.)
2606 PH/21 Added message_body_newlines option.
2608 PH/22 Guard against possible overflow in moan_check_errorcopy().
2610 PH/23 POSIX allows open() to be a macro; guard against that.
2612 PH/24 If the recipient of an error message contained an @ in the local part
2613 (suitably quoted, of course), incorrect values were put in $domain and
2614 $local_part during the evaluation of errors_copy.
2620 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2621 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2624 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2625 issue a MAIL command.
2627 PH/02 In an ACL statement such as
2629 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2631 if a client was not listed at all, or was listed with a value other than
2632 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2633 the condition was not true (as it should be), so access was not denied.
2634 The bug was that the ! inversion was incorrectly passed on to the second
2635 item. This has been fixed.
2637 PH/03 Added additional dnslists conditions == and =& which are different from
2638 = and & when the dns lookup returns more than one IP address.
2640 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2641 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2643 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2644 FSYNC, which compiles an option called disable_fsync that allows for
2645 bypassing fsync(). The documentation is heavily laced with warnings.
2647 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2649 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2650 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2651 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2652 including adding "make clean"; (3) Added -fPIC when compiling the test
2653 dynamically loaded module, to get rid of a warning.
2655 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2656 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2657 = 0s. The bug is otherwise harmless.
2659 PH/07 There was a bug in the dovecot authenticator such that the value of
2660 $auth1 could be overwritten, and so not correctly preserved, after a
2661 successful authentication. This usually meant that the value preserved by
2662 the server_setid option was incorrect.
2664 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2666 PH/09 Installed PCRE release 7.0.
2668 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2669 run for batched SMTP input. It is now run at the start of every message
2670 in the batch. While fixing this I discovered that the process information
2671 (output by running exiwhat) was not always getting set for -bs and -bS
2672 input. This is fixed, and it now also says "batched" for BSMTP.
2674 PH/11 Added control=no_pipelining.
2676 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2677 patch, slightly modified), and move the expansion of helo_data till after
2678 the connection is made in the smtp transport (so it can use these
2681 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2683 PH/14 Added log_selector = +pid.
2685 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2687 PH/16 Add ${if forany and ${if forall.
2689 PH/17 Added dsn_from option to vary the From: line in DSNs.
2691 PH/18 Flush SMTP output before performing a callout, unless control =
2692 no_callout_flush is set.
2694 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2695 was true (the default) a successful delivery failed to delete the retry
2696 item, thus causing premature timeout of the address. The bug is now
2699 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2701 PH/21 Long custom messages for fakedefer and fakereject are now split up
2702 into multiline responses in the same way that messages for "deny" and
2703 other ACL rejections are.
2705 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2706 with slight modification.
2708 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2709 draft, changing the syntax and factoring some duplicate code".
2711 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2712 for deliveries of the second and subsequent messages over the same SMTP
2715 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2716 ${reduce, with only minor "tidies".
2718 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2720 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2721 expansion side effects.
2723 PH/27 When a message times out after an over-quota error from an Exim-imposed
2724 quota, the bounce message says "mailbox is full". This message was not
2725 being given when it was a system quota that was exceeded. It now should
2728 MH/03 Made $recipients available in local_scan(). local_scan() already has
2729 better access to the recipient list through recipients_list[], but
2730 $recipients can be useful in postmaster-provided expansion strings.
2732 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2733 in the case of a MAIL command with additional options following the
2734 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2735 were accidentally chopped off.
2737 PH/29 SMTP synchronization checks are implemented when a command is read -
2738 there is a check that no more input is waiting when there shouldn't be
2739 any. However, for some commands, a delay in an ACL can mean that it is
2740 some time before the response is written. In this time, more input might
2741 arrive, invalidly. So now there are extra checks after an ACL has run for
2742 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2743 pipelining has not been advertised.
2745 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2747 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2748 returned in the SMTP response when smtp_return_error_details was false.
2749 This has been fixed.
2751 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2752 instead of the C I/O that was originally supplied, because problems were
2753 reported on Solaris.
2755 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2756 Exim which did not show up earlier: it was assuming that a call to
2757 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2758 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2759 was a macro that became an assignment, so it seemed to work. This has
2760 changed to a proper function call with a void return, hence the compile
2761 error. Exim's code has been fixed.
2763 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2766 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2768 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2770 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2771 because it turns out to be too big or there is a timeout) there is no
2772 "Completed" line in the log. When some messages of this type were
2773 selected by exigrep, they were listed as "not completed". Others were
2774 picked up by some special patterns. I have improved the selection
2775 criteria to be more general.
2777 PH/38 The host_find_failed option in the manualroute router can now be set
2778 to "ignore", to completely ignore a host whose IP address cannot be
2779 found. If all hosts are ignored, the behaviour is controlled by the new
2780 host_all_ignored option.
2782 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2783 homing or because of multiple MX records with /mx) generated more than
2784 one IP address, and the following item turned out to be the local host,
2785 all the secondary addresses of the first item were incorrectly removed
2786 from the list, along with the local host and any following hosts (which
2787 is what is supposed to happen).
2789 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2790 whoever called Exim into the -H file. In the case of the daemon it was
2791 behaving confusingly. When first started, it used values for whoever
2792 started the daemon, but after a SIGHUP it used the Exim user (because it
2793 calls itself on a restart). I have changed the code so that it now always
2796 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2797 message are rejected with the same error (e.g. no authentication or bad
2798 sender address), and a DATA command is nevertheless sent (as can happen
2799 with PIPELINING or a stupid MUA), the error message that was given to the
2800 RCPT commands is included in the rejection of the DATA command. This is
2801 intended to be helpful for MUAs that show only the final error to their
2804 PH/42 Another patch from the Sieve maintainer.
2806 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2807 Eximstats - Fixed some broken HTML links and added missing column headers
2809 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2810 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2812 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2814 PH/43 Yet another patch from the Sieve maintainer.
2816 PH/44 I found a way to check for a TCP/IP connection going away before sending
2817 the response to the final '.' that terminates a message, but only in the
2818 case where the client has not sent further data following the '.'
2819 (unfortunately, this is allowed). However, in many cases there won't be
2820 any further data because there won't be any more messages to send. A call
2821 to select() can be used: if it shows that the input is "ready", there is
2822 either input waiting, or the socket has been closed. An attempt to read
2823 the next input character can distinguish the two cases. Previously, Exim
2824 would have sent an OK response which the client would never have see.
2825 This could lead to message repetition. This fix should cure that, at
2826 least in a lot of common cases.
2828 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2829 advertised in response to EHLO.
2835 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2836 fixed by 4.65/MH/01 (is this a record?) are fixed:
2838 (i) An empty string was always treated as zero by the numeric comparison
2839 operators. This behaviour has been restored.
2841 (ii) It is documented that the numeric comparison operators always treat
2842 their arguments as decimal numbers. This was broken in that numbers
2843 starting with 0 were being interpreted as octal.
2845 While fixing these problems I realized that there was another issue that
2846 hadn't been noticed. Values of message_size_limit (both the global option
2847 and the transport option) were treated as octal if they started with 0.
2848 The documentation was vague. These values are now always treated as
2849 decimal, and I will make that clear in the documentation.
2855 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2856 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2859 MH/01 Don't check that the operands of numeric comparison operators are
2860 integers when their expansion is in "skipping" mode (fixes bug
2861 introduced by 4.64-PH/07).
2863 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2864 child addresses, Exim now panics and dies. Previously, because the count
2865 is held in a short int, deliveries were likely to be lost. As such a
2866 large number of recipients for a single message is ridiculous
2867 (performance will be very, very poor), I have chosen to impose a limit
2868 rather than extend the field.
2874 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2875 leftover -K file (the existence of which was triggered by #402).
2876 While we were at it, introduced process PID as part of the -K
2877 filename. This should rule out race conditions when creating
2880 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2881 processing considerably. Previous code took too long for large mails,
2882 triggering a timeout which in turn triggers #401.
2884 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2885 in the DK code in transports.c. sendfile() is not really portable,
2886 hence the _LINUX specificness.
2888 TF/01 In the add_headers option to the mail command in an Exim filter,
2889 there was a bug that Exim would claim a syntax error in any
2890 header after the first one which had an odd number of characters
2893 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2894 callout verification, Exim cached a "reject" for the entire domain. This
2895 is correct for most verifications, but it is not correct for a recipient
2896 verification with use_sender or use_postmaster set, because in that case
2897 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2898 case of MAIL FROM:<> rejection from other early rejections (e.g.
2899 rejection of HELO). When verifying a recipient using a non-null MAIL
2900 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2901 Whatever the result of the callout, the value of the domain cache is
2902 left unchanged (for any other kind of callout, getting as far as trying
2903 RCPT means that the domain itself is ok).
2905 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2908 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2909 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2910 session, the daemon ignores ECONNECT errors and logs others; it now
2911 ignores EPIPE as well.
2913 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2914 (quoted-printable decoding).
2916 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2917 later the small subsequent patch to fix an introduced bug.
2919 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2921 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2923 PH/08 An error is now given if message_size_limit is specified negative.
2925 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2926 to be given (somewhat) arbitrary names.
2928 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2931 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2932 miscellaneous code fixes
2934 PH/10 Added the log_reject_target ACL modifier to specify where to log
2937 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2938 hostname. This is wrong, because it relates to the incoming message (and
2939 probably the interface on which it is arriving) and not to the outgoing
2940 callout (which could be using a different interface). This has been
2941 changed to use the value of the helo_data option from the smtp transport
2942 instead - this is what is used when a message is actually being sent. If
2943 there is no remote transport (possible with a router that sets up host
2944 addresses), $smtp_active_hostname is used.
2946 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2947 tweaks were necessary in order to get it to work (see also 21 below):
2948 (a) The code assumed that strncpy() returns a negative number on buffer
2949 overflow, which isn't the case. Replaced with Exim's string_format()
2951 (b) There were several signed/unsigned issues. I just did the minimum
2952 hacking in of casts. There is scope for a larger refactoring.
2953 (c) The code used strcasecmp() which is not a standard C function.
2954 Replaced with Exim's strcmpic() function.
2955 (d) The code set only $1; it now sets $auth1 as well.
2956 (e) A simple test gave the error "authentication client didn't specify
2957 service in request". It would seem that Dovecot has changed its
2958 interface. Fortunately there's a specification; I followed it and
2959 changed what the client sends and it appears to be working now.
2961 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2964 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2965 address A is aliased to B and C, where B exists and C does not. Without
2966 -v the output is "A verified" because verification stops after a
2967 successful redirection if more than one address is generated. However,
2968 with -v the child addresses are also verified. Exim was outputting "A
2969 failed to verify" and then showing the successful verification for C,
2970 with its parentage. It now outputs "B failed to verify", showing B's
2971 parentage before showing the successful verification of C.
2973 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2974 look up a TXT record in a specific list after matching in a combined
2977 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2978 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2979 they consult the DNS. I had assumed they would set it the way they
2980 wanted; and indeed my experiments on Linux seem to show that in some
2981 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2982 To be on the safe side, however, I have now made the interface to
2983 host_find_byname() similar to host_find_bydns(), with an argument
2984 containing the DNS resolver options. The host_find_byname() function now
2985 sets these options at its start, just as host_find_bydns() does. The smtp
2986 transport options dns_qualify_single and dns_search_parents are passed to
2987 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2988 of host_find_byname() use the default settings of RES_DEFNAMES
2989 (qualify_single) but not RES_DNSRCH (search_parents).
2991 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2992 spool_read_header() do less string testing, by means of a preliminary
2993 switch on the second character of optional "-foo" lines. (This is
2994 overdue, caused by the large number of possibilities that now exist.
2995 Originally there were few.) While I was there, I also converted the
2996 str(n)cmp tests so they don't re-test the leading "-" and the first
2997 character, in the hope this might squeeze out yet more improvement.
2999 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3000 flag allowing group syntax was set by the header_syntax check but not
3001 turned off, possible causing trouble later; (2) The flag was not being
3002 set at all for the header_verify test, causing "group"-style headers to
3003 be rejected. I have now set it in this case, and also caused header_
3004 verify to ignore an empty address taken from a group. While doing this, I
3005 came across some other cases where the code for allowing group syntax
3006 while scanning a header line wasn't quite right (mostly, not resetting
3007 the flag correctly in the right place). These bugs could have caused
3008 trouble for malformed header lines. I hope it is now all correct.
3010 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3011 with the "reply" argument non-NULL. The code, however (which originally
3012 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3013 but it didn't always do it. This confused somebody who was copying the
3014 code for some other use. I have removed all the tests.
3016 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3017 feature that was used to support insecure browsers during the U.S. crypto
3018 embargo. It requires special client support, and Exim is probably the
3019 only MTA that supported it -- and would never use it because real RSA is
3020 always available. This code has been removed, because it had the bad
3021 effect of slowing Exim down by computing (never used) parameters for the
3022 RSA_EXPORT functionality.
3024 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3025 authenticator to fail if there's a tab character in the incoming data
3026 (there should never be unless someone is messing about, as it's supposed
3027 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3028 if the connection is using TLS or if the remote IP is the same as the
3029 local IP, and the "valid-client-cert option" if a client certificate has
3032 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3033 authenticators. This can be used for authorization after authentication
3034 succeeds. (In the case of plaintext, it servers for both authentication
3037 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3038 if any retry times were supplied.
3040 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3041 connection, where there is no client IP address to check. In this
3042 situation, the verify now always succeeds.
3044 PH/25 Applied John Jetmore's -Mset patch.
3046 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3048 PH/27 In a string expansion for a processed (not raw) header when multiple
3049 headers of the same name were present, leading whitespace was being
3050 removed from all of them, but trailing whitespace was being removed only
3051 from the last one. Now trailing whitespace is removed from each header
3052 before concatenation. Completely empty headers in a concatenation (as
3053 before) are ignored.
3055 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3056 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3058 PH/29 [Removed. This was a change that I later backed out, and forgot to
3059 correct the ChangeLog entry (that I had efficiently created) before
3060 committing the later change.]
3062 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3063 address errors (4xx response to RCPT) over the same connection as other
3064 messages routed to the same hosts. Such deliveries are always "forced",
3065 so retry times are not inspected. This resulted in far too many retries
3066 for the affected addresses. The effect occurred only when there were more
3067 hosts than the hosts_max_try setting in the smtp transport when it had
3068 the 4xx errors. Those hosts that it had tried were not added to the list
3069 of hosts for which the message was waiting, so if all were tried, there
3070 was no problem. Two fixes have been applied:
3072 (i) If there are any address or message errors in an SMTP delivery, none
3073 of the hosts (tried or untried) are now added to the list of hosts
3074 for which the message is waiting, so the message should not be a
3075 candidate for sending over the same connection that was used for a
3076 successful delivery of some other message. This seems entirely
3077 reasonable: after all the message is NOT "waiting for some host".
3078 This is so "obvious" that I'm not sure why it wasn't done
3079 previously. Hope I haven't missed anything, but it can't do any
3080 harm, as the worst effect is to miss an optimization.
3082 (ii) If, despite (i), such a delivery is accidentally attempted, the
3083 routing retry time is respected, so at least it doesn't keep
3084 hammering the server.
3086 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3087 in ${readsocket because some servers need this prod.
3089 PH/32 Added some extra debug output when updating a wait-xxx database.
3091 PH/33 The hint "could be header name not terminated by colon", which has been
3092 given for certain expansion errors for a long time, was not being given
3093 for the ${if def:h_colon_omitted{... case.
3095 PH/34 The spec says: "With one important exception, whenever a domain list is
3096 being scanned, $domain contains the subject domain." There was at least
3097 one case where this was not true.
3099 PH/35 The error "getsockname() failed: connection reset by peer" was being
3100 written to the panic log as well as the main log, but it isn't really
3101 panic-worthy as it just means the connection died rather early on. I have
3102 removed the panic log writing for the ECONNRESET error when getsockname()
3105 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3106 runs only) independently of the message's sender address. This meant
3107 that, if the 4xx error was in fact related to the sender, a different
3108 message to the same recipient with a different sender could confuse
3109 things. In particular, this can happen when sending to a greylisting
3110 server, but other circumstances could also provoke similar problems.
3111 I have changed the default so that the retry time for these errors is now
3112 based a combination of the sender and recipient addresses. This change
3113 can be overridden by setting address_retry_include_sender=false in the
3116 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3117 remote server are returned as part of bounce messages. This was not
3118 happening for LMTP over a pipe (the lmtp transport), but now it is the
3119 same for both kinds of LMTP.
3121 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3122 in header lines that were in fact CNAMEs. This is no longer the case.
3124 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3125 and queue runs started by the daemon processed all messages. This has
3126 been fixed so that -R and -S can now usefully be given with -q<time>.
3128 PH/40 Import PCRE release 6.7 (fixes some bugs).
3130 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3132 PH/42 Give an error if -q is specified more than once.
3134 PH/43 Renamed the variables $interface_address and $interface_port as
3135 $received_ip_address and $received_port, to make it clear that these
3136 values apply to message reception, and not to the outgoing interface when
3137 a message is delivered. (The old names remain recognized, of course.)
3139 PH/44 There was no timeout on the connect() call when using a Unix domain
3140 socket in the ${readsocket expansion. There now is.
3142 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3143 be meaningful with "accept".
3145 SC/01 Eximstats V1.43
3146 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3148 SC/02 Eximstats V1.44
3149 Use a glob alias rather than an array ref in the generated
3150 parser. This improves both readability and performance.
3152 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3153 Collect SpamAssassin and rejection statistics.
3154 Don't display local sender or destination tables unless
3155 there is data to show.
3156 Added average volumes into the top table text output.
3158 SC/04 Eximstats V1.46
3159 Collect data on the number of addresses (recipients)
3160 as well as the number of messages.
3162 SC/05 Eximstats V1.47
3163 Added 'Message too big' to the list of mail rejection
3164 reasons (thanks to Marco Gaiarin).
3166 SC/06 Eximstats V1.48
3167 Mainlog lines which have GMT offsets and are too short to
3168 have a flag are now skipped.
3170 SC/07 Eximstats V1.49 (Alain Williams)
3171 Added the -emptyok flag.
3173 SC/08 Eximstats V1.50
3174 Fixes for obtaining the IP address from reject messages.
3176 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3177 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3178 whitespace changes from 4.64-PH/27
3180 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3183 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3184 are found, allow negative numbers in numeric criteria)
3186 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3188 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3189 to match changes made in 4.64-PH/43
3191 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3193 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3194 "domain" is actually an IP address, and force a failure. This locks out
3195 those revolvers/nameservers that support "A-for-A" lookups, in
3196 contravention of the specifications.
3198 PH/48 When a host name was looked up from an IP address, and the subsequent
3199 forward lookup of the name timed out, the host name was left in
3200 $sender_host_name, contrary to the specification.
3202 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3203 restricted to single-key lookups, Exim was not diagnosing an error if
3204 * or *@ was used with a query-style lookup.
3206 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3208 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3209 long ago, but noone interested enough thought of it. Let's just say that
3210 the "1.1" means that there are some new functions that weren't there at
3211 some point in the past.
3213 PH/51 Error processing for expansion failure of helo_data from an smtp
3214 transport during callout processing was broken.
3216 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3217 tested/used via the -bh/-bhc/-bs options.
3219 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3220 bug, fixed in subsequent PCRE releases).
3222 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3223 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3225 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3231 SC/01 Use a glob alias rather than an array ref in eximstats generated
3232 parser. This improves both readability and performance.
3234 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3235 Don't display local sender or destination tables in eximstats unless
3236 there is data to show.
3237 Added average volumes into the eximstats top table text output.
3239 SC/03 Collect data on the number of addresses (recipients) as well
3240 as the number of messages in eximstats.
3242 TF/01 Correct an error in the documentation for the redirect router. Exim
3243 does (usually) call initgroups() when daemonizing.
3245 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3246 with consistent privilege compared to when running as a daemon.
3248 TF/03 Note in the spec that $authenticated_id is not set for local
3249 submissions from trusted users.
3251 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3252 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3254 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3255 by adding some example configuration directives to the default
3256 configuration file. A little bit of work is required to uncomment the
3257 directives and define how usernames and passwords are checked, but
3258 there is now a framework to start from.
3260 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3261 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3262 without this. I don't know how relevant this is to other LDAP libraries.
3264 PH/02 Add the verb name to the "unknown ACL verb" error.
3266 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3268 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3270 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3271 directory not expanded when it should be if an expanded home directory
3272 was set for the address (which is overridden by the transport).
3274 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3277 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3278 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3279 because it is too late at that time, and has no effect.
3281 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3282 security issue with \' (bugzilla #107). I could not use the
3283 PQescapeStringConn() function, because it needs a PGconn value as one of
3286 PH/08 When testing addresses using -bt, indicate those final addresses that
3287 are duplicates that would not cause an additional delivery. At least one
3288 person was confused, thinking that -bt output corresponded to deliveries.
3289 (Suppressing duplicates isn't a good idea as you lose the information
3290 about possibly different redirections that led to the duplicates.)
3292 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3293 systems where poll() doesn't work, in particular OS X.
3295 PH/10 Added more information to debugging output for retry time not reached.
3297 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3298 operations in malware.c.
3300 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3303 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3304 syslog_duplication set false, log lines that would normally be written
3305 both the the main log and to the reject log were not written to syslog at
3308 PH/14 In the default configuration, change the use of "message" in ACL warn
3309 statements to "add_header".
3311 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3312 not followed by a command (e.g. "seen endif").
3314 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3315 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3318 PH/17 Added extra conditions to the default value of delay_warning_condition
3322 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3323 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3324 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3327 The Auto-Submitted: and various List- headers are standardised, whereas I
3328 don't think Precedence: ever was.
3330 PH/18 Refactored debugging code in route_finduser() to show more information,
3331 in particular, the error code if getpwnam() issues one.
3333 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3334 This is apparently needed in addition to the PH/07 change above to avoid
3335 any possible encoding problems.
3337 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3338 but not after initializing Perl.
3340 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3341 output them only if debugging. By default they are written stderr,
3342 apparently, which is not desirable.
3344 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3347 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3350 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3352 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3353 authenticated or an ident call has been made. Suppress the default
3354 values for $authenticated_id and $authenticated_sender (but permit -oMai
3355 and -oMas) when testing with -bh.
3357 PH/24 Re-jigged the order of the tests in the default configuration so that the
3358 tests for valid domains and recipients precede the DNS black list and CSA
3359 tests, on the grounds that those ones are more expensive.
3361 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3362 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3363 This bug exists in every version of Exim that I still have, right back to
3366 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3367 However, an attempt to turn on case-sensitivity in a regex key by
3368 including (?-i) didn't work because the subject string was already
3369 lowercased, and the effects were non-intuitive. It turns out that a
3370 one-line patch can be used to allow (?-i) to work as expected.
3376 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3377 other effects) broke the use of negated acl sub-conditions.
3379 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3382 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3383 "Deny" causes Exim to reject the incoming connection with a 554 error.
3384 Unfortunately, if there is a major crisis, such as a disk failure,
3385 tcp-wrappers gives "deny", whereas what one would like would be some
3386 kind of temporary error. A kludge has been added to help with this.
3387 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3388 554 error is used if errno is still zero or contains ENOENT (which occurs
3389 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3392 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3394 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3395 errors. Otherwise a message that provokes a temporary error (when other
3396 messages do not) can cause a whole host to time out.
3398 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3399 addresses were routed directly to files or pipes from a redirect router.
3400 File deliveries just didn't batch; pipe deliveries might have suffered
3403 PH/06 A failure to get a lock for a hints database would erroneously always say
3404 "Failed to get write lock", even when it was really a read lock.
3406 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3407 of 0600. This has been changed to use the value of the lockfile_mode
3408 option (which defaults to 0600).
3410 PH/08 Applied small patch from the Sieve maintainer.
3412 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3413 folder from quota calculations, a direct delivery into this folder messed
3414 up the contents of the maildirsize file. This was because the regex was
3415 used only to exclude .Trash (or whatever) when the size of the mailbox
3416 was calculated. There was no check that a delivery was happening into an
3417 excluded directory. This bug has been fixed by ignoring all quota
3418 processing for deliveries into excluded directories.
3420 PH/10 Added the maildirfolder_create_regex option to appendfile.
3426 PH/01 The code for finding all the local interface addresses on a FreeBSD
3427 system running IPv6 was broken. This may well have applied to all BSD
3428 systems, as well as to others that have similar system calls. The broken
3429 code found IPv4 interfaces correctly, but gave incorrect values for the
3430 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3431 that it would not match correctly against @[] and not recognize the IPv6
3434 PH/02 The ipliteral router was not recognizing addresses of the form user@
3435 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3437 PH/03 Added disable_ipv6.
3439 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3440 decoded form, because it is most often used to construct To: headers
3441 lines in autoreplies, and the decoded form may well be syntactically
3442 invalid. However, $reply_address has leading white space removed, and all
3443 newlines turned into spaces so that the autoreply transport does not
3446 PH/05 If group was specified without a user on a router, and no group or user
3447 was specified on a transport, the group from the router was ignored.
3449 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3450 for visible compile-time settings that can be used to change these
3451 numbers, for those that want even more. Backwards compatibility with old
3452 spool files has been maintained. However, going back to a previous Exim
3453 release will lost any variables that are in spool files.
3455 PH/07 Two small changes when running in the test harness: increase delay when
3456 passing a TCP/IP connection to a new process, in case the original
3457 process has to generate a bounce, and remove special handling of
3458 127.0.0.2 (sic), which is no longer necessary.
3460 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3461 be the same on different OS.
3463 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3466 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3467 whether --show-vars was specified or not
3469 JJ/02 exipick: Added support for new ACL variable spool format introduced
3472 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3473 syntactically invalid From: or Reply-to: line, and a filter used this to
3474 generate an autoreply, and therefore failed to obtain an address for the
3475 autoreply, Exim could try to deliver to a non-existent relative file
3476 name, causing unrelated and misleading errors. What now happens is that
3477 it logs this as a hard delivery error, but does not attempt to create a
3480 PH/11 The exinext utility has a -C option for testing purposes, but although
3481 the given file was scanned by exinext itself; it wasn't being passed on
3482 when Exim was called.
3484 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3485 an end-of-file indication when reading a command response.
3487 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3488 compiled. In many other places in Exim, IPv6 addresses are always
3489 recognized, so I have changed this. It also means that IPv4 domain
3490 literals of the form [IPV4:n.n.n.n] are now always recognized.
3492 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3493 used if the router is not running as root, for example, when verifying at
3494 ACL time, or when using -bh. The debugging output from this situation was
3495 non-existent - all you got was a failure to exec. I have made two
3498 (a) Failures to set uid/gid, the current directory, or a process leader
3499 in a subprocess such as that created by queryprogram now generate
3500 suitable debugging output when -d is set.
3502 (b) The queryprogram router detects when it is not running as root,
3503 outputs suitable debugging information if -d is set, and then runs
3504 the subprocess without attempting to change uid/gid.
3506 PH/15 Minor change to Makefile for building test_host (undocumented testing
3509 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3510 additional section of a DNS packet that returns MX or SRV records.
3511 Instead, it always explicitly searches for A/AAAA records. This avoids
3512 major problems that occur when a DNS server includes only records of one
3513 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3514 fixed another bug: if SRV records were looked up and the corresponding
3515 address records were *not* found in the additional section, the port
3516 values from the SRV records were lost.
3518 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3519 using the correct key (the original address) when searching the retry
3520 rules in order to find which one to use for generating the retry hint.
3522 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3523 adding the default one. Similarly, if it contains a Reply-To: header, the
3524 errors_reply_to option, if set, is not used.
3526 PH/19 When calculating a retry time, Exim used to measure the "time since
3527 failure" by looking at the "first failed" field in the retry record. Now
3528 it does not use this if it is later than than the arrival time of the
3529 message. Instead it uses the arrival time. This makes for better
3530 behaviour in cases where some deliveries succeed, thus re-setting the
3531 "first failed" field. An example is a quota failure for a huge message
3532 when small messages continue to be delivered. Without this change, the
3533 "time since failure" will always be short, possible causing more frequent
3534 delivery attempts for the huge message than are intended.
3535 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3537 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3538 $1, $2, $3) because the numerical variables can be reset during some
3539 expansion items (e.g. "match"), thereby losing the authentication data.
3541 PH/21 Make -bV show the size of off_t variables so that the test suite can
3542 decide whether to run tests for quotas > 2G.
3544 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3545 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3546 filecount value is greater than 2G or if a quota value is greater than 2G
3547 on a system where the size of off_t is not greater than 4, a panic error
3550 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3551 never match. The debug and -bh output now contains an explicit error
3552 message indicating a malformed IPv4 address or mask.
3554 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3555 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3556 PH/23 above applies.
3558 PH/25 Do not write to syslog when running in the test harness. The only
3559 occasion when this arises is a failure to open the main or panic logs
3560 (for which there is an explicit test).
3562 PH/26 Added the /no_tell option to "control=freeze".
3564 PH/27 If a host name lookup failed very early in a connection, for example, if
3565 the IP address matched host_lookup and the reverse lookup yielded a name
3566 that did not have a forward lookup, an error message of the form "no IP
3567 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3568 could be logged. Now it outputs the IP address instead of "NULL".
3570 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3571 allows the sender and the authenticated sender to be set when
3572 submitting a message from within Exim. Since child_open_exim() is
3573 documented for local_scan(), the new function should be too.
3575 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3576 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3577 results in an empty string is now treated as unset.
3579 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3581 PH/31 Added sender_verify_fail logging option.
3583 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3584 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3585 tidied the source and removed it altogether.
3587 PH/33 When a queue run was abandoned because the load average was too high, a
3588 log line was always written; now it is written only if the queue_run log
3589 selector is set. In addition, the log line for abandonment now contains
3590 information about the queue run such as the pid. This is always present
3591 in "start" and "stop" lines but was omitted from the "abandon" line.
3593 PH/34 Omit spaces between a header name and the colon in the error message that
3594 is given when verify = headers_syntax fails (if there are lots of them,
3595 the message gets confusing).
3597 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3598 names, as there are now some PTR records that contain slashes. This check
3599 is only to protect against broken name servers that fall over on strange
3600 characters, so the fact that it applies to all lookups doesn't matter.
3602 PH/36 Now that the new test suite is complete, we can remove some of the
3603 special code in Exim that was needed for the old test suite. For example,
3604 sorting DNS records because real resolvers return them in an arbitrary
3605 order. The new test suite's fake resolver always returns records in the
3608 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3609 bounces) except when queue_only is set, to avoid logging races between
3610 the different processes.
3612 PH/38 Panic-die if .include specifies a non-absolute path.
3614 PH/39 A tweak to the "H" retry rule from its user.
3616 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3617 a label. They prevented compilation on older perls.
3619 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3620 a warning to be raised on newish perls.
3622 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3623 on queue. Changes to match documented behaviour of showing count of
3624 messages matching specified criteria.
3626 PH/40 Changed the default ident timeout from 30s to 5s.
3628 PH/41 Added support for the use of login_cap features, on those BSD systems
3629 that have them, for controlling the resources used by pipe deliveries.
3631 PH/42 The content-scanning code uses fopen() to create files in which to put
3632 message data. Previously it was not paying any attention to the mode of
3633 the files. Exim runs with umask(0) because the rest of the code creates
3634 files with open(), and sets the required mode explicitly. Thus, these
3635 files were ending up world-writeable. This was not a big issue, because,
3636 being within the spool directory, they were not world-accessible. I have
3637 created a function called modefopen, which takes an additional mode
3638 argument. It sets umask(777), creates the file, chmods it to the required
3639 mode, then resets the umask. All the relevant calls to fopen() in the
3640 content scanning code have been changed to use this function.
3642 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3643 to 24 hours. This avoids potential overflow problems when processing G
3644 and H retry rules. I suspect nobody ever tinkers with this value.
3646 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3648 PH/45 When the plaintext authenticator is running as a client, the server's
3649 challenges are checked to ensure they are valid base64 strings. By
3650 default, the authentication attempt is cancelled if an invalid string is
3651 received. Setting client_ignore_invalid_base64 true ignores these errors.
3652 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3653 they are received. Thus, the responses can be made to depend on the
3654 challenges. If an invalid string is ignored, an empty string is placed in
3657 PH/46 Messages that are created by the autoreply transport now contains a
3658 References: header, in accordance with RFCs 2822 and 3834.
3660 PH/47 Added authenticated_sender_force to the smtp transport.
3662 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3664 PH/49 Installed latest patch from the Sieve maintainer.
3666 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3667 was also set, the appendfile transport was unnecessarily scanning a
3668 directory of message files (e.g. for maildir delivery) to find the count
3669 of files (along with the size), even though it did not need this
3670 information. It now does the scan only if it needs to find either the
3671 size of the count of files.
3673 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3675 PH/52 Two bugs concerned with error handling when the smtp transport is
3678 (i) Exim was not creating retry information for temporary errors given
3679 for individual recipients after the DATA command when the smtp transport
3680 was used in LMTP mode. This meant that they could be retried too
3681 frequently, and not timed out correctly.
3683 (ii) Exim was setting the flag that allows error details to be returned
3684 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3685 recipients that were returned after the DATA command.
3687 PH/53 This is related to PH/52, but is more general: for any failing address,
3688 when detailed error information was permitted to be returned to the
3689 sender, but the error was temporary, then after the final timeout, only
3690 "retry timeout exceeded" was returned. Now it returns the full error as
3691 well as "retry timeout exceeded".
3693 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3694 do this, and (what is worse) MTAs that accept it.
3696 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3697 will now be deprecated.
3699 PH/56 New os.c-cygwin from the Cygwin maintainer.
3701 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3702 formats (previously only available in exim formats via -bpr, -bpru,
3703 and -bpra. Now also available in native and exiqgrep formats)
3705 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3706 with very large, slow to parse queues
3708 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3710 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3712 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3713 responses to authentication challenges, though it was showing the
3714 challenges; (ii) I've removed the CR characters from the debug output for
3717 PH/58 Allow for the insertion of a newline as well as a space when a string
3718 is turned into more than one encoded-word during RFC 2047 encoding. The
3719 Sieve code now uses this.
3721 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3722 data_4xx, lost_connection, tls_required.
3724 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3725 message was being sent as an SMTP response.
3727 PH/61 Add -l and -k options to exicyclog.
3729 PH/62 When verifying, if an address was redirected to one new address, so that
3730 verification continued, and the new address failed or deferred after
3731 having set something in $address_data, the value of $address_data was not
3732 passed back to the ACL. This was different to the case when no
3733 redirection occurred. The value is now passed back in both cases.
3735 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3736 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3737 use login_cap.h, so on its own it isn't the distinguishing feature. The
3738 new name refers directly to the setclassresources() function.
3740 PH/65 Added configuration files for NetBSD3.
3742 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3744 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3745 is preferred over IPv4.
3747 PH/68 The bounce_return_message and bounce_return_body options were not being
3748 honoured for bounces generated during the reception of non-SMTP messages.
3749 In particular, this applied to messages rejected by the ACL. This bug has
3750 been fixed. However, if bounce_return_message is true and bounce_return_
3751 body is false, the headers that are returned for a non-SMTP message
3752 include only those that have been read before the error was detected.
3753 (In the case of an ACL rejection, they have all been read.)
3755 PH/69 The HTML version of the specification is now built in a directory called
3756 spec_html instead of spec.html, because the latter looks like a path with
3757 a MIME-type, and this confuses some software.
3759 PH/70 Catch two compiler warnings in sieve.c.
3761 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3762 function verify_get_ident() calls ip_connect() to connect a socket, but
3763 if the "connect()" function timed out, ip_connect() used to close the
3764 socket. However, verify_get_ident() also closes the socket later, and in
3765 between Exim writes to the log, which may get opened at this point. When
3766 the socket was closed in ip_connect(), the log could get the same file
3767 descriptor number as the socket. This naturally causes chaos. The fix is
3768 not to close the socket in ip_connect(); the socket should be closed by
3769 the function that creates it. There was only one place in the code where
3770 this was missing, in the iplookup router, which I don't think anybody now
3771 uses, but I've fixed it anyway.
3773 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3774 well as to direct DNS lookups. Otherwise the handling of names in host
3775 lists is inconsistent and therefore confusing.
3781 PH/01 Two changes to the default runtime configuration:
3783 (1) Move the checks for relay_from_hosts and authenticated clients from
3784 after to before the (commented out) DNS black list checks.
3786 (2) Add control=submission to the relay_from_hosts and authenticated
3787 clients checks, on the grounds that messages accepted by these
3788 statements are most likely to be submissions.
3790 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3792 (1) Generate an error if the third argument for the ${prvs expansion is
3795 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3798 (3) Reset the variables that are obtained from the first argument of
3799 ${prvscheck and used in the second argument before leaving the code,
3800 because their memory is reclaimed, so using them afterwards may do
3803 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3804 one (it's much easier than Tom thought :-).
3806 (5) Because of (4), we can now allow for the use of $prvscheck_result
3807 inside the third argument.
3809 PH/03 For some reason, the default setting of PATH when running a command from
3810 a pipe transport was just "/usr/bin". I have changed it to
3813 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3814 anything to be listed in the output from -bV.
3816 PH/05 When a filter generated an autoreply, the entire To: header line was
3817 quoted in the delivery log line, like this:
3819 => >A.N.Other <ano@some.domain> <original@ddress> ...
3821 This has been changed so that it extracts the operative address. There
3822 may be more than one such address. If so, they are comma-separated, like
3825 => >ano@some.domain,ona@other.domain <original@ddress> ...
3827 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3828 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3829 not being looked up in the rDNS to get a host name, Exim was showing the
3830 IP address twice in Received: lines, even though the IP addresses were
3831 identical. For example:
3833 Received: from [1.2.3.4] (helo=[1.2.3.4])
3835 However, if the real host name was known, it was omitting the HELO data
3836 if it matched the actual IP address. This has been tidied up so that it
3837 doesn't show the same IP address twice.
3839 PH/07 When both +timestamp and +memory debugging was on, the value given by
3840 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3841 called by the debug printing, thereby overwriting the timestamp buffer.
3842 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3844 PH/08 When the original message was included in an autoreply transport, it
3845 always said "this is a copy of the message, including all the headers",
3846 even if body_only or headers_only was set. It now gives an appropriate
3849 PH/09 Applied a patch from the Sieve maintainer which:
3851 o fixes some comments
3852 o adds the (disabled) notify extension core
3853 o adds some debug output for the result of if/elsif tests
3854 o points to the current vacation draft in the documentation
3855 and documents the missing references header update
3859 o fixes a bug in processing the envelope test (when testing
3860 multiple envelope elements, the last element determined the
3863 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3864 Electronic Mail") by including:
3866 Auto-submitted: auto-generated
3868 in the messages that it generates (bounce messages and others, such as
3869 warnings). In the case of bounce messages for non-SMTP messages, there was
3870 also a typo: it was using "Auto_submitted" (underscore instead of
3871 hyphen). Since every message generated by Exim is necessarily in response
3872 to another message, thes have all been changed to:
3874 Auto-Submitted: auto-replied
3876 in accordance with these statements in the RFC:
3878 The auto-replied keyword:
3880 - SHOULD be used on messages sent in direct response to another
3881 message by an automatic process,
3883 - MUST NOT be used on manually-generated messages,
3885 - MAY be used on Delivery Status Notifications (DSNs) and Message
3886 Disposition Notifications (MDNs),
3888 - MUST NOT be used on messages generated by automatic or periodic
3889 processes, except for messages which are automatic responses to
3892 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3893 to the default Received: header definition.
3895 PH/12 Added log selector acl_warn_skipped (default on).
3897 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3898 variables because (a) they are in the wrong storage pool and (b) even if
3899 they were copied, it wouldn't work properly because of the caching.
3901 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3902 checking when decoding. Apparently there are clients that generate
3903 overlong encoded strings. Why am I not surprised?
3905 PH/15 If the first argument of "${if match_address" was not empty, but did not
3906 contain an "@" character, Exim crashed. Now it writes a panic log message
3907 and treats the condition as false.
3909 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3911 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3912 extension "envelope-auth". The code is finished and in agreement with
3913 other implementations, but there is no documentation so far and in fact,
3914 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3915 not changing the active code.
3917 Print executed "if" and "elsif" statements when debugging is used. This
3918 helps a great deal to understand what a filter does.
3920 Document more things not specified clearly in RFC3028. I had all this
3921 sorted out, when out of a sudden new issues came to my mind. Oops."
3923 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3926 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3927 canonical form (as documented). However, after a host name lookup from
3928 the IP address, check_host() was doing a simple string comparison with
3929 addresses acquired from the DNS when checking that the found name did
3930 have the original IP as one of its addresses. Since any found IPv6
3931 addresses are likely to be in abbreviated form, the comparison could
3932 fail. Luckily, there already exists a function for doing the comparison
3933 by converting both addresses to binary, so now that is used instead of
3934 the text comparison.
3936 PH/20 There was another similar case to PH/19, when a complete host name was
3937 given in a host list; looking up its IP address could give an abbreviated
3938 form, whereas the current host's name might or might not be abbreviated.
3939 The same fix has been applied.
3945 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3946 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3949 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3950 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3952 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3954 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3955 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3956 Also, for "Auto-Submitted", treat anything other than "no" as
3957 non-personal, in accordance with RFC 3834. (Previously it treated
3958 anything starting "auto-" as non-personal.)
3960 TF/01 The control=submission/name=... option had a problem with syntax
3961 errors if the name included a slash character. The /name= option
3962 now slurps the rest of the string, so it can include any characters
3963 but it must come last in the list of options (after /sender_retain
3966 PH/05 Some modifications to the interface to the fake nameserver for the new
3974 TK/01 Added the "success_on_redirect" address verification option. See
3975 NewStuff for rationale and an example.
3977 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3979 PH/02 Patch to exigrep to allow it to work on syslog lines.
3981 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3982 fread() to skip over the body file's header line, because in Cygwin the
3983 header line is locked and is inaccessible.
3985 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3986 co-exist for some time) to make it clear that it is the Exim ID that is
3987 referenced, not the Message-ID: header line.
3989 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3990 string_format() function, because snprintf() does not exist on all
3993 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3994 ${stat: expansion item.
3996 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3997 protocol synchronization error", to keep the pedants happy.
3999 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4000 well as for IRIX systems, when gcc is being used. See the host.c source
4003 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4005 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4008 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4009 transport and to the smtp transport in LMTP mode.
4011 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4013 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4014 run a filter in a subprocess. This could lead to confusion in subsequent
4015 lookups in the parent process. There should also be a search_tidyup() at
4016 the end of the subprocess.
4018 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4019 only if the host matched helo_try_verify_hosts, which caused the
4020 verification to occur when the EHLO/HELO command was issued. The ACL just
4021 tested the remembered result. Now, if a previous verification attempt has
4022 not happened, "verify = helo" does it there and then.
4024 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4026 TK/03 Fix log output including CR from clamd.
4028 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4029 exist provoked a memory error which could cause a segfault.
4031 PH/15 Installed PCRE 6.2
4033 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4035 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4036 of the problem. Specifically, suggested +O2 rather than +O1 for the
4039 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4041 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4042 record, the port was not correctly logged when the outgoing_port log
4043 selector was set (it logged the transort's default port).
4045 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4046 fallback_hosts, and "hosts" in the smtp transport.
4048 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4049 host errors such as "Connection refused".
4051 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4052 authentication with radiusclient 0.4.9:
4054 - Error returned from rc_read_config was caught wrongly
4055 - Username/password not passed on to radius server due to wrong length.
4057 The presumption is that some radiusclient API changes for 4.51/PH/17
4058 were not taken care of correctly. The code is still untested by me (my
4059 Linux distribution still has 0.3.2 of radiusclient), but it was
4060 contributed by a Radius user.
4062 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4063 expanding the "port" option of the smtp transport.
4065 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4066 while reading a MIME header. Thanks to Tom Hughes for a patch.
4068 PH/24 Include config.h inside local_scan.h so that configuration settings are
4071 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4072 that in an ACL for RCPT (for example), you can examine exactly what was
4075 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4076 commands, but it was not correctly comparing the address with the actual
4077 client host address. Thus, it would show the EHLO address in Received:
4078 header lines when this was not necessary.
4080 PH/27 Added the % operator to ${eval:}.
4082 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4083 it should be ignoring failures (because with -C, for example, it has lost
4084 privilege). It wasn't ignoring creation failures other than "already
4087 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4090 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4091 because an input line was too long, either on its own, or by virtue of
4092 too many continuations, the temporary file was not being removed, and the
4093 return code was incorrect.
4095 PH/31 Missing "BOOL" in function definition in filtertest.c.
4097 PH/32 Applied Sieve patches from the maintainer.
4099 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4101 PH/33 Added "verify = not_blind".
4103 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4104 Local/Makefile (with some defaults set). These are used in built scripts
4105 such as exicyclog, but they have never been used in the exim_install
4106 script (though there are many overriding facilities there). I have
4107 arranged that the exim_install script now takes note of these two
4110 PH/35 Installed configuration files for Dragonfly.
4112 PH/36 When a locally submitted message by a trusted user did not contain a
4113 From: header, and the sender address was obtained from -f or from an SMTP
4114 MAIL command, and the trusted user did not use -F to supply a sender
4115 name, $originator_name was incorrectly used when constructing a From:
4116 header. Furthermore, $originator_name was used for submission mode
4117 messages from external hosts without From: headers in a similar way,
4118 which is clearly wrong.
4120 PH/37 Added control=suppress_local_fixups.
4122 PH/38 When log_selector = +received_sender was set, and the addition of the
4123 sender made the log line's construction buffer exactly full, or one byte
4124 less than full, an overflow happened when the terminating "\n" was
4127 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4128 when the result of a list match is failure because a DNS lookup failed.
4130 PH/40 RM_COMMAND is now used in the building process.
4132 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4133 the "build-* directories that it finds.
4135 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4136 address in a domain literal was a prefix of an interface address.
4138 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4139 when verifying a sender address, unless rewrite_headers is false.
4141 PH/44 Wrote a long comment about why errors_to addresses are verified as
4142 recipients, not senders.
4144 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4145 the ratelimit ACL was added.
4147 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4149 PH/46 Added extra information about PostgreSQL errors to the error string.
4151 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4152 avoiding the need to install special zones in a real server. This is
4153 backwards compatible; if it can't find the fake resolver, it drops back.
4154 Thus, both old and new test suites can be run.
4156 TF/02 Added util/ratelimit.pl
4158 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4159 clock is set back in time.
4161 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4162 Candler <B.Candler@pobox.com>.
4164 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4165 OK for addresses that are the result of redirections.
4167 PH/48 A number of further additions for the benefit of the new test suite,
4168 including a fake gethostbyname() that interfaces to the fake DNS resolver
4171 TF/06 The fix for widen_domains has also been applied to qualify_single and
4172 search_parents which are the other dnslookup options that can cause
4175 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4178 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4180 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4181 service" if it encountered an MX record with an empty target hostname.
4182 The message is now "an MX or SRV record indicated no SMTP service".
4184 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4185 if the DNS of the sending site is misconfigured. This is quite a
4186 common situation. This change restores the behaviour of treating a
4187 helo verification defer as a failure.
4189 PH/51 If self=fail was set on a router, the bounce message did not include the
4190 actual error message.
4196 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4198 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4199 command itself ended in error, the underlying message about the transport
4200 filter timeout was being overwritten with the pipe command error. Now the
4201 underlying error message should be appended to the second error message.
4203 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4205 PH/02 Reduce the amount of output that "make" produces by default. Full output
4206 can still be requested.
4208 PH/03 The warning log line about a condition test deferring for a "warn" verb
4209 was being output only once per connection, rather than after each
4210 occurrence (because it was using the same function as for successful
4211 "warn" verbs). This seems wrong, so I have changed it.
4213 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4214 it should not have, which might have caused a crash in the right
4215 circumstances, but probably never did.
4217 PH/04 Installed a modified version of Tony Finch's patch to make submission
4218 mode fix the return path as well as the Sender: header line, and to
4219 add a /name= option so that you can make the user's friendly name appear
4222 TF/03 Added the control = fakedefer ACL modifier.
4224 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4225 Mark Lowes for thorough testing.
4227 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4229 TK/03 Merged latest SRS patch from Miles Wilton.
4231 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4232 with the definition in sysexits.h (which is #included earlier).
4233 Fortunately, Exim does not actually use EX_OK. The code used to try to
4234 preserve the sysexits.h value, by assuming that macro definitions were
4235 scanned for macro replacements. I have been disabused of this notion,
4236 so now the code just undefines EX_OK before #including unistd.h.
4238 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4239 in the smtp transport. When a block could not be written in a single
4240 write() function, the timeout was being re-applied to each part-write.
4241 This seems wrong - if the receiver was accepting one byte at a time it
4242 would take for ever. The timeout is now adjusted when this happens. It
4243 doesn't have to be particularly precise.
4245 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4246 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4248 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4249 without a domain if the check to <postmaster@domain> fails.
4251 SC/01 Eximstats: added -xls and the ability to specify output files
4252 (patch written by Frank Heydlauf).
4254 SC/02 Eximstats: use FileHandles for outputting results.
4256 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4258 SC/04 Eximstats: fixed display of large numbers with -nvr option
4260 SC/05 Eximstats: fixed merging of reports with empty tables.
4262 SC/06 Eximstats: added the -include_original_destination flag
4264 SC/07 Eximstats: removed tabs and trailing whitespace.
4266 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4268 TK/06 MBOX spool code: Add real "From " MBOX separator line
4269 so the .eml file is really in mbox format (even though
4270 most programs do not really care). Patch from Alex Miller.
4272 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4273 The latter is generated from $received_to and is only set if the
4274 message has one envelope recipient. SA can use these headers,
4275 obviously out-of-the-box. Patch from Alex Miller.
4277 PH/08 The ${def test on a variable was returning false if the variable's
4278 value was "0", contrary to what the specification has always said!
4279 The result should be true unless the variable is empty.
4281 PH/09 The syntax error of a character other than { following "${if
4282 def:variable_name" (after optional whitespace) was not being diagnosed.
4283 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4284 accidental colon was present, for example, could give incorrect results.
4286 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4287 result is used (not including appendfile, where other changes are about
4290 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4291 This involved changing a lot of size variables from int to off_t. It
4292 should work with maildirs and everything.
4294 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4295 spamd dying while we are connected to it.
4297 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4300 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4301 function for BDB 4.3.
4303 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4305 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4306 the sources for PCRE have been re-arranged and more files are now
4309 PH/15 The code I had for printing potentially long long variables in PH/11
4310 above was not the best (it lost precision). The length of off_t variables
4311 is now inspected at build time, and an appropriate printing format (%ld
4312 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4313 to be "long long int" or "long int". This is needed for the internal
4314 formatting function string_vformat().
4316 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4317 the configuration file to be ":syslog", then the script "guesses" where
4318 the logs files are, rather than using the compiled in default. In our
4319 case the guess is not the same as the compiled default, so the script
4320 suddenly stopped working when I started to use syslog. The patch checks
4321 to see if log_file_path is "". If so, it attempts to read it from exim
4322 with no configuration file to get the compiled in version, before it
4323 falls back to the previous guessing code."
4325 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4326 implementing BATV in an Exim configuration. See NewStuff for the gory
4329 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4330 Makefile that are specific to HP-UX.
4332 PH/18 If the "use_postmaster" option was set for a recipient callout together
4333 with the "random" option, the postmaster address was used as the MAIL
4334 FROM address for the random test, but not for the subsequent recipient
4335 test. It is now used for both.
4337 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4338 patch removes a few documentation additions to RFC 3028, because the
4339 latest draft now contains them. It adds the new en;ascii-case comparator
4340 and a new error check for 8bit text in MIME parts. Comparator and
4341 require names are now matched exactly. I enabled the subaddress
4342 extension, but it is not well tested yet (read: it works for me)."
4344 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4345 rework some of the code of TK/09 above to avoid the hardwired use of
4346 "%lld" and "long long". Replaced the call to snprintf() with a call to
4349 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4350 records point to non-existent hosts", "retry timeout exceeded", and
4351 "retry time not reached for any host after a long failure period".
4353 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4354 experimental DomainKeys support:
4356 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4357 (2) On an error such as an illegally used "control", the wrong name for
4358 the control was given.
4360 These problems did NOT occur unless DomainKeys support was compiled.
4362 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4364 PH/24 Added ${if match_ip condition.
4366 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4367 databases so that it will be absolutely obvious if a crash occurs in the
4368 DB library. This is a regular occurrence (often caused by mis-matched
4371 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4372 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4373 picked up on a user's system that detects such things. There doesn't seem
4374 to be a gcc warning option for this - only an attribute that has to be
4375 put on the function's prototype. It seems that in Fedora Core 4 they have
4376 set this on a number of new functions. No doubt there will be more in due
4379 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4380 specify a transport. However, if an address that was verified by such a
4381 router was the subject of a callout, Exim crashed because it tried to
4382 read the rcpt_include_affixes from the non-existent transport. Now it
4383 just assumes that the setting of that option is false. This bug was
4384 introduced by 4.51/PH/31.
4386 PH/28 Changed -d+all to exclude +memory, because that information is very
4387 rarely of interest, but it makes the output a lot bigger. People tend to
4388 do -d+all out of habit.
4390 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4391 code in os-type was giving problems when libc.so lives in lib64, like on
4394 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4395 aren't the modern standard, and it seems that some systems' include files
4396 don't always have them. Exim was already checking for some of the newer
4397 ones like T_AAAA, and defining it itself. I've added checks for all the
4398 record types that Exim uses.
4400 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4401 not automatically generating a new one, as it is supposed to. This
4402 prevented TLS from working. If the file did exist, but contained invalid
4403 data, a new version was generated, as expected. It was only the case of a
4404 non-existent file that was broken.
4406 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4407 with a change in libdomainkeys > 0.64.
4409 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4410 from DNS. If the selector record carries the flag, it now has
4411 precedence over the domain-wide flag.
4413 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4415 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4416 the use of a port name that isn't defined in /etc/services) occurred, the
4417 message was deferred as in a normal delivery, and thus remained on the
4418 spool, instead of being failed because of the mua_wrapper setting. This
4419 is now fixed, and I tidied up some of the mua_wrapper messages at the
4422 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4423 the messages in a hash of arrays rather than using individual hashes.
4424 This is a bit cleaner and results in dramatic memory savings, albeit
4425 at a slight CPU cost.
4427 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4428 as requested by Marc Sherman.
4430 SC/10 Eximstats: added histograms for user specified patterns as requested
4433 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4435 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4436 fopen() in the content-scanning modules that did not already have it.
4442 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4443 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4445 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4447 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4449 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4450 to test Sieve filters that use "vacation".
4452 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4453 that changes the way the GnuTLS parameters are stored in the cache file.
4454 The new format can be generated externally. For backward compatibility,
4455 if the data in the cache doesn't make sense, Exim assumes it has read an
4456 old-format file, and it generates new data and writes a new file. This
4457 means that you can't go back to an older release without removing the
4460 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4461 work if there are any delivery delays because "one_time" forces the
4462 parent to be marked "delivered", so its unseen clone is never tried
4463 again. For this reason, Exim now forbids the simultaneous setting of
4466 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4467 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4468 entry if you want to know the details. The fix, however, neglected to
4469 consider the case when local delivery batching is involved. The test for
4470 "previously delivered" was not happening when checking to see if an
4471 address could be batched with a previous (undelivered) one; under
4472 certain circumstances this could lead to multiple deliveries to the same
4475 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4476 in its include files, and this causes problems building Exim.
4478 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4479 header_syntax) but Exim was just ignoring anything given after a slash.
4480 In particular, this caused confusion with an attempt to use "verify =
4481 reverse_host_lookup/defer_ok". An error is now given when options are
4482 supplied for verify items that do not have them. (Maybe reverse_host_
4483 lookup should have a defer_ok option, but that's a different point.)
4485 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4486 defined by RFC 821) to 2048, because there were problems with some AUTH
4487 commands, and RFC 1869 says the size should be increased for extended
4488 SMTP commands that take arguments.
4490 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4493 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4494 "unknown" error; now it says that the functionality isn't in the binary.
4496 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4497 an address' error message when a string expansion fails (syntax or
4498 whatever). Otherwise the password may appear in the log. Following change
4499 PH/42 below, there is no longer a chance of it appearing in a bounce
4502 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4504 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4505 of its addresses was ever tried. (Bugzilla bug #2.)
4507 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4508 the result incorrectly in the debug output. (It correctly added a newline
4509 to what was transported.)
4511 TF/01 Added $received_time.
4513 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4514 commented out examples of how to interface to a virus scanner and to
4515 SpamAssassin. Also added commented examples of av_scanner and
4516 spamd_address settings.
4518 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4519 and controls are allowed in which ACLs. There were a couple of minor
4520 errors. Some of the entries in the conditions table (which is a table of
4521 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4522 negation of where the condition IS allowed.
4524 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4526 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4527 header file does not have a version number, so I've had to invent a new
4528 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4529 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4530 radiusclient), but it was contributed by a Radius user.
4532 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4533 files or directories, for OpenSSL.
4535 PH/19 When an Exim process that is running as root has to create an Exim log
4536 file, it does so in a subprocess that runs as exim:exim so as to get the
4537 ownership right at creation (otherwise, other Exim processes might see
4538 the file with the wrong ownership). There was no test for failure of this
4539 fork() call, which would lead to the process getting stuck as it waited
4540 for a non-existent subprocess. Forks do occasionally fail when resources
4541 run out. I reviewed all the other calls to fork(); they all seem to check
4544 PH/20 When checking for unexpected SMTP input at connect time (before writing
4545 the banner), Exim was not dealing correctly with a non-positive return
4546 from the read() function. If the client had disconnected by this time,
4547 the result was a log entry for a synchronization error with an empty
4548 string after "input=" when read() returned zero. If read() returned -1
4549 (an event I could not check), uninitialized data bytes were printed.
4550 There were reports of junk text (parts of files, etc) appearing after
4553 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4555 PH/22 Added support for macro redefinition, and (re)definition in between
4556 driver and ACL definitions.
4558 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4559 forgetting to use the resulting value; it was using the unexpanded value.
4561 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4562 hadn't been configured. The fix is from Juergen Kreileder, who
4563 understands it better than I do:
4565 "Here's what I see happening with three configured cyrus_sasl
4566 authenticators configured (plain, login, cram-md5):
4568 On startup auth_cyrus_sasl_init() gets called for each of these.
4569 This means three calls to sasl_listmech() without a specified mech_list.
4570 => SASL tests which mechs of all available mechs actually work
4571 => three warnings about OTP not working
4572 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4574 With the patch, sasl_listmech() also gets called three times. But now
4575 SASL's mech_list option is set to the server_mech specified in the the
4576 authenticator. Or in other words, the answer from sasl_listmech()
4577 gets limited to just the mech you're testing for (which is different
4579 => the return list contains just 'plain' or 'login', 'cram-md5' or
4580 nothing depending on the value of ob->server_mech.
4582 I've just tested the patch: Authentication still works fine,
4583 unavailable mechs specified in the exim configuration are still
4584 caught, and the auth.log warnings about OTP are gone."
4586 PH/25 When debugging is enabled, the contents of the command line are added
4587 to the debugging output, even when log_selector=+arguments is not
4590 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4591 answer is "GNU", and only if the return is "GNU/something" is the answer
4594 PH/27 $acl_verify_message is now set immediately after the failure of a
4595 verification in an ACL, and so is available in subsequent modifiers. In
4596 particular, the message can be preserved by coding like this:
4598 warn !verify = sender
4599 set acl_m0 = $acl_verify_message
4601 Previously, $acl_verify_message was set only while expanding "message"
4602 and "log_message" when a very denied access.
4604 PH/28 Modified OS/os.c-Linux with
4606 -#ifndef OS_LOAD_AVERAGE
4607 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4609 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4610 nomenclature these days.)
4612 PH/29 Installed patch from the Sieve maintainer that adds the options
4613 sieve_useraddress and sieve_subaddress to the redirect router.
4615 PH/30 In these circumstances:
4616 . Two addresses routed to the same list of hosts;
4617 . First host does not offer TLS;
4618 . First host accepts first address;
4619 . First host gives temporary error to second address;
4620 . Second host offers TLS and a TLS session is established;
4621 . Second host accepts second address.
4622 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4623 and peerdn, if requested) that were in fact used only for the second
4626 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4627 attention to any local part prefix or suffix that was matched by the
4628 router that accepted the address. It now behaves in the same way as it
4629 does for delivery: the affixes are removed from the local part unless
4630 rcpt_include_affixes is set on the transport.
4632 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4633 timeout during the DATA phase of an incoming message.
4635 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4636 applied a patch sanctioned by the Sieve maintainer.
4638 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4639 the uid or gid is negative. A case of a negative gid caused this to be
4640 noticed. The fix allows for either to be negative.
4642 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4643 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4646 PH/36 The change PH/12 above was broken. Fixed it.
4648 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4649 the grounds that routing the same address twice would always produce the
4650 same answer. This might have been true once, but it is certainly no
4651 longer true now. Routing a child address may depend on the previous
4652 routing that produced that child. Some complicated redirection strategies
4653 went wrong when messages had multiple recipients, and made Exim's
4654 behaviour dependent on the order in which the addresses were given.
4656 I have moved the duplicate checking until after the routing is complete.
4657 Exim scans the addresses that are assigned to local and remote
4658 transports, and removes any duplicates. This means that more work will be
4659 done, as duplicates will always all be routed, but duplicates are
4660 presumably rare, so I don't expect this is of any significance.
4662 For deliveries to pipes, files, and autoreplies, the duplicate checking
4663 still happens during the routing process, since they are not going to be
4666 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4667 It corrects a timeout issue with spamd. This is Ian's comment: "The
4668 background is that sometimes spamd either never reads data from a
4669 connection it has accepted, or it never writes response data. The exiscan
4670 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4671 blindly assumes that writes won't block so it may never time out."
4673 PH/39 Allow G after quota size as well as K and M.
4675 PH/40 The value set for $authenticated_id in an authenticator may not contain
4676 binary zeroes or newlines because the value is written to log lines and
4677 to spool files. There was no check on this. Now the value is run through
4678 the string_printing() function so that such characters are converted to
4679 printable escape sequences.
4681 PH/41 $message_linecount is a new variable that contains the total number of
4682 lines in the message. Compare $body_linecount, which is the count for the
4685 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4686 bounce and delay warning messages, except in certain special cases, which
4689 (a) An SMTP error message from a remote host;
4690 (b) A message specified in a :fail: redirection;
4691 (c) A message specified in a "fail" command in a system filter;
4692 (d) A message specified in a FAIL return from the queryprogram router;
4693 (e) A message specified by the cannot_route_message router option.
4695 In these cases only, Exim does include the error details in bounce and
4696 warning messages. There are also a few cases where bland messages such
4697 as "unrouteable address" or "local delivery error" are given.
4699 PH/43 $value is now also set for the "else" part of a ${run expansion.
4701 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4702 being worked on, but at least Exim now implements the latest version to
4705 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4706 process to complete was treated as a delivery failure, a timeout while
4707 writing the message to the pipe was logged, but erroneously treated as a
4708 successful delivery. Such timeouts include transport filter timeouts. For
4709 consistency with the overall process timeout, these timeouts are now
4710 treated as errors, giving rise to delivery failures by default. However,
4711 there is now a new Boolean option for the pipe transport called
4712 timeout_defer, which, if set TRUE, converts the failures into defers for
4713 both kinds of timeout. A transport filter timeout is now identified in
4716 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4717 systems where "make" and "gmake" are different, calling "gmake" at top
4718 level broke things. I've arranged for the value of $(MAKE) to be passed
4719 from the Makefile to this script so that it can call the same version of
4723 A note about Exim versions 4.44 and 4.50
4724 ----------------------------------------
4726 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4727 changes of various kinds. As a consequence, a big documentation update was
4728 needed. This delayed the release for rather longer than seemed good, especially
4729 in the light of a couple of (minor) security issues. Therefore, the changes
4730 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4731 release. So 4.44 and 4.50 are in effect two different branches that both start
4734 I have left the 4.50 change log unchanged; it contains all the changes since
4735 4.43. The change log for 4.44 is below; many of its items are identical to
4736 those for 4.50. This seems to be the most sensible way to preserve the
4737 historical information.
4743 1. Minor wording change to the doc/README.SIEVE file.
4745 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4746 computation of the current number of files was incorrect.
4748 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4749 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4752 4. Give more explanation in the error message when the command for a transport
4753 filter fails to execute.
4755 5. There are several places where Exim runs a non-Exim command in a
4756 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4757 was being done only for the command run by the queryprogram router. It is
4758 now done for all such subprocesses. The other cases are: ${run, transport
4759 filters, and the commands run by the lmtp and pipe transports.
4761 6. Added CONFIGURE_GROUP build-time option.
4763 7. Some older OS have a limit of 256 on the maximum number of file
4764 descriptors. Exim was using setrlimit() to set 1000 as a large value
4765 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4766 systems. I've change it so that if it can't get 1000, it tries for 256.
4768 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4769 was an oversight, and furthermore, ever since the addition of extra
4770 controls (e.g. 4.43/32), the checks on when to allow different forms of
4771 "control" were broken. There should now be diagnostics for all cases when a
4772 control that does not make sense is encountered.
4774 9. Added the /retain_sender option to "control=submission".
4776 10. $recipients is now available in the predata ACL (oversight).
4778 11. Tidy the search cache before the fork to do a delivery from a message
4779 received from the command line. Otherwise the child will trigger a lookup
4780 failure and thereby defer the delivery if it tries to use (for example) a
4781 cached ldap connection that the parent has called unbind on.
4783 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4784 of $address_data from the recipient verification was clobbered by the
4785 sender verification.
4787 13. The value of address_data from a sender verification is now available in
4788 $sender_address_data in subsequent conditions in the ACL statement.
4790 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4792 15. Added a new option "connect=<time>" to callout options, to set a different
4795 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4796 was its contents. (It was OK if the option was not defined at all.)
4798 17. A "Completed" log line is now written for messages that are removed from
4799 the spool by the -Mrm option.
4801 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4802 information about exactly what failed.
4804 19. Added -dd to debug only the daemon process.
4806 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4807 handles timeouts, both on the server side and network timeouts. Renamed the
4808 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4810 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4811 It is now set to "smtps".
4813 22. $host_address is now set to the target address during the checking of
4814 ignore_target_hosts.
4816 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4817 being passed; this would have caused $sender_host_name to have been used if
4818 matching the list had actually called for a host name (not very likely,
4819 since this list is usually IP addresses). A host name is now passed as
4822 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4823 code that specifies a non-restarting handler (typically sigaction() in
4824 modern systems) in an attempt to fix a rare and obscure crash bug.
4826 25. Narrowed the window for a race in the daemon that could cause it to ignore
4827 SIGCHLD signals. This is not a major problem, because they are used only to
4828 wake it up if nothing else does.
4830 26. A malformed maildirsize file could cause Exim to calculate negative values
4831 for the mailbox size or file count. Odd effects could occur as a result.
4832 The maildirsize information is now recalculated if the size or filecount
4835 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4836 support for a long time. Removed HAVE_SYS_VFS_H.
4838 28. Installed the latest version of exipick from John Jetmore.
4840 29. In an address list, if the pattern was not a regular expression, an empty
4841 subject address (from a bounce message) matched only if the pattern was an
4842 empty string. Non-empty patterns were not even tested. This was the wrong
4843 because it is perfectly reasonable to use an empty address as part of a
4844 database query. An empty address is now tested by patterns that are
4845 lookups. However, all the other forms of pattern expect the subject to
4846 contain a local part and a domain, and therefore, for them, an empty
4847 address still always fails if the pattern is not itself empty.
4849 30. Exim went into a mad DNS loop when attempting to do a callout where the
4850 host was specified on an smtp transport, and looking it up yielded more
4851 than one IP address.
4853 31. Re-factored the code for checking spool and log partition space into a
4854 function that finds that data and another that does the check. The former
4855 is then used to implement four new variables: $spool_space, $log_space,
4856 $spool_inodes, and $log_inodes.
4858 32. The RFC2047 encoding function was originally intended for short strings
4859 such as real names; it was not keeping to the 75-character limit for
4860 encoded words that the RFC imposes. It now respects the limit, and
4861 generates multiple encoded words if necessary. To be on the safe side, I
4862 have increased the buffer size for the ${rfc2047: expansion operator from
4865 33. It is now permitted to omit both strings after an "if" condition; if the
4866 condition is true, the result is "true". As before, when the second string
4867 is omitted, a false condition yields an empty string. This makes it less
4868 cumbersome to write custom ACL and router conditions.
4870 34. Failure to deliver a bounce message always caused it to be frozen, even if
4871 there was an errors_to setting on the router. The errors_to setting is now
4874 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4875 canonical form (fully expanded) before being placed in
4876 $sender_host_address.
4878 36. The table in the code that translates DNS record types into text (T_A to
4879 "A" for instance) was missing entries for NS and CNAME. It is just possible
4880 that this could have caused confusion if both these types were looked up
4881 for the same domain, because the text type is used as part of Exim's
4882 per-process caching. But the chance of anyone hitting this buglet seems
4885 37. The dnsdb lookup has been extended in a number of ways.
4887 (1) There is a new type, "zns", which walks up the domain tree until it
4888 finds some nameserver records. It should be used with care.
4890 (2) There is a new type, "mxh", which is like "mx" except that it returns
4891 just the host names, not the priorities.
4893 (3) It is now possible to give a list of domains (or IP addresses) to be
4894 looked up. The behaviour when one of the lookups defers can be
4895 controlled by a keyword.
4897 (4) It is now possible to specify the separator character for use when
4898 multiple records are returned.
4900 38. The dnslists ACL condition has been extended: it is now possible to supply
4901 a list of IP addresses and/or domains to be looked up in a particular DNS
4904 39. Added log_selector=+queue_time_overall.
4906 40. When running the queue in the test harness, wait just a tad after forking a
4907 delivery process, to get repeatability of debugging output.
4909 41. Include certificate and key file names in error message when GnuTLS fails
4910 to set them up, because the GnuTLS error message doesn't include the name
4911 of the failing file when there is a problem reading it.
4913 42. Allow both -bf and -bF in the same test run.
4915 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4917 44. The "Exiscan patch" is now merged into the mainline Exim source.
4919 45. Sometimes the final signoff response after QUIT could fail to get
4920 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4921 < 0 before doing a fflush(). This bug looks as though it goes back to the
4922 introduction of TLS in release 3.20, but "sometimes" must have been rare
4923 because the tests only now provoked it.
4925 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4926 (this can affect the format of dates).
4928 47. exim_tidydb, when checking for the continued existence of a message for
4929 which it has found a message-specific retry record, was not finding
4930 messages that were in split spool directories. Consequently, it was
4931 deleting retry records that should have stayed in existence.
4933 48. Steve fixed some bugs in eximstats.
4935 49. The SPA authentication driver was not abandoning authentication and moving
4936 on to the next authenticator when an expansion was forced to fail,
4937 contradicting the general specification for all authenticators. Instead it
4938 was generating a temporary error. It now behaves as specified.
4940 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4941 (the order specifies the preference for clients). The order is now AES256,
4942 AES128, 3DES, ARCFOUR128.
4944 51. Small patch to Sieve code - explicitly set From: when generating an
4947 52. Exim crashed if a remote delivery caused a very long error message to be
4948 recorded - for instance if somebody sent an entire SpamAssassin report back
4949 as a large number of 550 error lines. This bug was coincidentally fixed by
4950 increasing the size of one of Exim's internal buffers (big_buffer) that
4951 happened as part of the Exiscan merge. However, to be on the safe side, I
4952 have made the code more robust (and fixed the comments that describe what
4955 53. Now that there can be additional text after "Completed" in log lines (if
4956 the queue_time_overall log selector is set), a one-byte patch to exigrep
4957 was needed to allow it to recognize "Completed" as not the last thing in
4960 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4961 patch that reportedly fixes this has been added. I am not expert enough to
4962 create a test for it. This is what the patch creator wrote:
4964 "I found a little strange behaviour of ldap code when working with
4965 Windows 2003 AD Domain, where users was placed in more than one
4966 Organization Units. When I tried to give exim partial DN, the exit code
4967 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4968 But simultaneously result of request was absolutely normal ldap result,
4969 so I produce this patch..."
4971 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4972 REFERENCE, so I have modified the code to exclude the patch when that macro
4975 55. Some experimental protocols are using DNS PTR records for new purposes. The
4976 keys for these records are domain names, not reversed IP addresses. The
4977 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4978 leaves it alone. Component reversal etc. now happens only for IP addresses.
4981 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4983 57. Double the size of the debug message buffer (to 2048) so that more of very
4984 long debug lines gets shown.
4986 58. The exicyclog utility now does better if the number of log files to keep
4987 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4989 59. Two changes related to the smtp_active_hostname option:
4991 (1) $smtp_active_hostname is now available as a variable.
4992 (2) The default for smtp_banner uses $smtp_active_hostname instead
4993 of $primary_hostname.
4995 60. The host_aton() function is supposed to be passed a string that is known
4996 to be a valid IP address. However, in the case of IPv6 addresses, it was
4997 not checking this. This is a hostage to fortune. Exim now panics and dies
4998 if the condition is not met. A case was found where this could be provoked
4999 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5000 components; fortuitously, this particular loophole had already been fixed
5001 by change 4.50/55 above.
5003 If there are any other similar loopholes, the new check in host_aton()
5004 itself should stop them being exploited. The report I received stated that
5005 data on the command line could provoke the exploit when Exim was running as
5006 exim, but did not say which command line option was involved. All I could
5007 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5008 running as the user.
5011 61. There was a buffer overflow vulnerability in the SPA authentication code
5012 (which came originally from the Samba project). I have added a test to the
5013 spa_base64_to_bits() function which I hope fixes it.
5016 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5017 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5019 63. The daemon start-up calls getloadavg() while still root for those OS that
5020 need the first call to be done as root, but it missed one case: when
5021 deliver_queue_load_max is set with deliver_drop_privilege. This is
5022 necessary for the benefit of the queue runner, because there is no re-exec
5023 when deliver_drop_privilege is set.
5025 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5026 This has been fixed.
5028 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5029 was in use, was not putting the data itself into the right store pool;
5030 consequently, it could be overwritten for a subsequent message in the same
5031 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5034 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5036 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5037 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5038 the function were treating the return as a boolean value, which happened to
5039 work because 0=false and not-0=true, but is not correct code.
5041 68. The host_aton() function was not handling scoped IPv6 addresses (those
5042 with, for example, "%eth0" on the end) correctly.
5044 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5045 negated items (that is, ~something) in unsigned ints. Some compilers
5046 apparently mutter when there is no cast.
5048 70. If an address verification called from an ACL failed, and did not produce a
5049 user-specific message (i.e. there was only a "system" message), nothing was
5050 put in $acl_verify_message. In this situation, it now puts the system
5053 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5054 session; change 4.31/43 added the unwanted input to the log line - except
5055 that it did not do this in the start of session case. It now does.
5057 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5058 This is wrong and can cause the other end to generate a synchronization
5059 error if it is another Exim or anything else that does the synchronization
5060 check. A QUIT command is no longer sent after a timeout.
5062 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5063 during host lookups.
5065 74. The defer_ok option of callout verification was not working if it was used
5066 when verifying addresses in header lines, that is, for this case:
5068 verify = header_sender/callout=defer_ok
5070 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5071 those file descriptors could be used for SMTP connections. If anything
5072 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5073 could be sent to the SMTP client, causing chaos. The daemon now opens
5074 stdin, stdout, and stderr to /dev/null when it puts itself into the
5077 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5078 log by default. The user can override this with suitable Perl magic.
5080 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5081 the log message when discard triggers, was not working for the DATA ACL or
5082 for the non-SMTP ACL.
5084 78. Error message wording change in sieve.c.
5086 79. If smtp_accept_max_per_host was set, the number of connections could be
5087 restricted to fewer than expected, because the daemon was trying to set up
5088 a new connection before checking whether the processes handling previous
5089 connections had finished. The check for completed processes is now done
5090 earlier. On busy systems, this bug wouldn't be noticed because something
5091 else would have woken the daemon, and it would have reaped the completed
5094 80. If a message was submitted locally by a user whose login name contained one
5095 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5096 It caused a spool format error. I have fixed the spool reading code. A
5097 related problem was that the "from" clause in the Received: line became
5098 illegal because of the space(s). It is now covered by ${quote_local_part.
5100 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5103 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5105 83. Patch from Sieve maintainer to fix unterminated string problem in
5106 "vacation" handling.
5108 84. Some minor changes to the Linux configuration files to help with other
5109 OS variants using glibc.
5111 85. One more patch for Sieve to update vacation handling to latest spec.
5114 ----------------------------------------------------
5115 See the note above about the 4.44 and 4.50 releases.
5116 ----------------------------------------------------
5122 1. Change 4.43/35 introduced a bug that caused file counts to be
5123 incorrectly computed when quota_filecount was set in an appendfile
5126 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5127 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5130 3. Give more explanation in the error message when the command for a transport
5131 filter fails to execute.
5133 4. There are several places where Exim runs a non-Exim command in a
5134 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5135 was being done only for the command run by the queryprogram router. It is
5136 now done for all such subprocesses. The other cases are: ${run, transport
5137 filters, and the commands run by the lmtp and pipe transports.
5139 5. Some older OS have a limit of 256 on the maximum number of file
5140 descriptors. Exim was using setrlimit() to set 1000 as a large value
5141 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5142 systems. I've change it so that if it can't get 1000, it tries for 256.
5144 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5145 was an oversight, and furthermore, ever since the addition of extra
5146 controls (e.g. 4.43/32), the checks on when to allow different forms of
5147 "control" were broken. There should now be diagnostics for all cases when a
5148 control that does not make sense is encountered.
5150 7. $recipients is now available in the predata ACL (oversight).
5152 8. Tidy the search cache before the fork to do a delivery from a message
5153 received from the command line. Otherwise the child will trigger a lookup
5154 failure and thereby defer the delivery if it tries to use (for example) a
5155 cached ldap connection that the parent has called unbind on.
5157 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5158 of $address_data from the recipient verification was clobbered by the
5159 sender verification.
5161 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5162 was its contents. (It was OK if the option was not defined at all.)
5164 11. A "Completed" log line is now written for messages that are removed from
5165 the spool by the -Mrm option.
5167 12. $host_address is now set to the target address during the checking of
5168 ignore_target_hosts.
5170 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5171 being passed; this would have caused $sender_host_name to have been used if
5172 matching the list had actually called for a host name (not very likely,
5173 since this list is usually IP addresses). A host name is now passed as
5176 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5177 code that specifies a non-restarting handler (typically sigaction() in
5178 modern systems) in an attempt to fix a rare and obscure crash bug.
5180 15. Narrowed the window for a race in the daemon that could cause it to ignore
5181 SIGCHLD signals. This is not a major problem, because they are used only to
5182 wake it up if nothing else does.
5184 16. A malformed maildirsize file could cause Exim to calculate negative values
5185 for the mailbox size or file count. Odd effects could occur as a result.
5186 The maildirsize information is now recalculated if the size or filecount
5189 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5190 support for a long time. Removed HAVE_SYS_VFS_H.
5192 18. Updated exipick to current release from John Jetmore.
5194 19. Allow an empty sender to be matched against a lookup in an address list.
5195 Previously the only cases considered were a regular expression, or an
5198 20. Exim went into a mad DNS lookup loop when doing a callout where the
5199 host was specified on the transport, if the DNS lookup yielded more than
5202 21. The RFC2047 encoding function was originally intended for short strings
5203 such as real names; it was not keeping to the 75-character limit for
5204 encoded words that the RFC imposes. It now respects the limit, and
5205 generates multiple encoded words if necessary. To be on the safe side, I
5206 have increased the buffer size for the ${rfc2047: expansion operator from
5209 22. Failure to deliver a bounce message always caused it to be frozen, even if
5210 there was an errors_to setting on the router. The errors_to setting is now
5213 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5214 canonical form (fully expanded) before being placed in
5215 $sender_host_address.
5217 24. Updated eximstats to version 1.33
5219 25. Include certificate and key file names in error message when GnuTLS fails
5220 to set them up, because the GnuTLS error message doesn't include the name
5221 of the failing file when there is a problem reading it.
5223 26. Expand error message when OpenSSL has problems setting up cert/key files.
5226 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5227 (this can affect the format of dates).
5229 28. exim_tidydb, when checking for the continued existence of a message for
5230 which it has found a message-specific retry record, was not finding
5231 messages that were in split spool directories. Consequently, it was
5232 deleting retry records that should have stayed in existence.
5234 29. eximstats updated to version 1.35
5235 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5236 1.35 - bugfix such that pie charts by volume are generated correctly
5238 30. The SPA authentication driver was not abandoning authentication and moving
5239 on to the next authenticator when an expansion was forced to fail,
5240 contradicting the general specification for all authenticators. Instead it
5241 was generating a temporary error. It now behaves as specified.
5243 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5244 (the order specifies the preference for clients). The order is now AES256,
5245 AES128, 3DES, ARCFOUR128.
5247 31. Small patch to Sieve code - explicitly set From: when generating an
5250 32. Exim crashed if a remote delivery caused a very long error message to be
5251 recorded - for instance if somebody sent an entire SpamAssassin report back
5252 as a large number of 550 error lines. This bug was coincidentally fixed by
5253 increasing the size of one of Exim's internal buffers (big_buffer) that
5254 happened as part of the Exiscan merge. However, to be on the safe side, I
5255 have made the code more robust (and fixed the comments that describe what
5258 33. Some experimental protocols are using DNS PTR records for new purposes. The
5259 keys for these records are domain names, not reversed IP addresses. The
5260 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5261 leaves it alone. Component reversal etc. now happens only for IP addresses.
5264 34. The host_aton() function is supposed to be passed a string that is known
5265 to be a valid IP address. However, in the case of IPv6 addresses, it was
5266 not checking this. This is a hostage to fortune. Exim now panics and dies
5267 if the condition is not met. A case was found where this could be provoked
5268 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5269 components; fortuitously, this particular loophole had already been fixed
5270 by change 4.50/55 or 4.44/33 above.
5272 If there are any other similar loopholes, the new check in host_aton()
5273 itself should stop them being exploited. The report I received stated that
5274 data on the command line could provoke the exploit when Exim was running as
5275 exim, but did not say which command line option was involved. All I could
5276 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5277 running as the user.
5280 35. There was a buffer overflow vulnerability in the SPA authentication code
5281 (which came originally from the Samba project). I have added a test to the
5282 spa_base64_to_bits() function which I hope fixes it.
5285 36. The daemon start-up calls getloadavg() while still root for those OS that
5286 need the first call to be done as root, but it missed one case: when
5287 deliver_queue_load_max is set with deliver_drop_privilege. This is
5288 necessary for the benefit of the queue runner, because there is no re-exec
5289 when deliver_drop_privilege is set.
5291 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5292 was in use, was not putting the data itself into the right store pool;
5293 consequently, it could be overwritten for a subsequent message in the same
5294 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5297 38. Sometimes the final signoff response after QUIT could fail to get
5298 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5299 < 0 before doing a fflush(). This bug looks as though it goes back to the
5300 introduction of TLS in release 3.20, but "sometimes" must have been rare
5301 because the tests only now provoked it.
5307 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5308 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5309 Now it returns an integer. A number of calls were still expecting a T/F
5310 return. Fortuitously, in all cases, the tests worked in OK situations,
5311 which is the norm. However, things would have gone wrong on any write
5312 failures on the smtp file descriptor. This function is used when sending
5313 messages over SMTP and also when doing verify callouts.
5315 2. When Exim is called to do synchronous delivery of a locally submitted
5316 message (the -odf or -odi options), it no longer closes stderr before doing
5319 3. Implemented the mua_wrapper option.
5321 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5323 5. Implemented the functions header_remove(), header_testname(),
5324 header_add_at_position(), and receive_remove_recipient(), and exported them
5327 6. If an ACL "warn" statement specified the addition of headers, Exim already
5328 inserted X-ACL-Warn: at the start if there was no header name. However, it
5329 was not making this test for the second and subsequent header lines if
5330 there were newlines in the string. This meant that an invalid header could
5331 be inserted if Exim was badly configured.
5333 7. Allow an ACL "warn" statement to add header lines at the start or after all
5334 the Received: headers, as well as at the end.
5336 8. Added the rcpt_4xx retry error code.
5338 9. Added postmaster_mailfrom=xxx to callout verification option.
5340 10. Added mailfrom=xxxx to the callout verification option, for verify=
5343 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5344 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5346 12. Inserted some casts to stop certain compilers warning when using pointer
5347 differences as field lengths or precisions in printf-type calls (mostly
5348 affecting debugging statements).
5350 13. Added optional readline() support for -be (dynamically loaded).
5352 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5353 same clock tick as a message's arrival, so that its received time was the
5354 same as the "first fail" time on the retry record, and that message
5355 remained on the queue past the ultimate address timeout, every queue runner
5356 would try a delivery (because it was past the ultimate address timeout) but
5357 after another failure, the ultimate address timeout, which should have then
5358 bounced the address, did not kick in. This was a "< instead of <=" error;
5359 in most cases the first failure would have been in the next clock tick
5360 after the received time, and all would be well.
5362 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5363 being recognized when the domain list was tested by the match_domain
5364 condition in an expansion string.
5366 16. Added the ${str2b64: operator.
5368 17. Exim was always calling setrlimit() to set a large limit for the number of
5369 processes, without checking whether the existing limit was already
5370 adequate. (It did check for the limit on file descriptors.) Furthermore,
5371 errors from getrlimit() and setrlimit() were being ignored. Now they are
5372 logged to the main and panic logs, but Exim does carry on, to try to do its
5373 job under whatever limits there are.
5375 18. Imported PCRE 5.0.
5377 19. Trivial typo in log message " temporarily refused connection" (the leading
5380 20. If the log selector return_path_on_delivery was set and an address was
5381 redirected to /dev/null, the delivery process crashed because it assumed
5382 that a return path would always be set for a "successful" delivery. In this
5383 case, the whole delivery is bypassed as an optimization, and therefore no
5386 21. Internal re-arrangement: the function for sending a challenge and reading
5387 a response while authentication was assuming a zero-terminated challenge
5388 string. It's now changed to take a pointer and a length, to allow for
5389 binary data in such strings.
5391 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5393 23. Exim was not respecting finduser_retries when seeking the login of the
5394 uid under which it was called; it was always trying 10 times. (The default
5395 setting of finduser_retries is zero.) Also, it was sleeping after the final
5396 failure, which is pointless.
5398 24. Implemented tls_on_connect_ports.
5400 25. Implemented acl_smtp_predata.
5402 26. If the domain in control=submission is set empty, Exim assumes that the
5403 authenticated id is a complete email address when it generates From: or
5404 Sender: header lines.
5406 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5407 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5408 chown and chgrp in /bin and hostname in /usr/bin.
5410 28. Exim was keeping the "process log" file open after each use, just as it
5411 does for the main log. This opens the possibility of it remaining open for
5412 long periods when the USR1 signal hits a daemon. Occasional processlog
5413 errors were reported, that could have been caused by this. Anyway, it seems
5414 much more sensible not to leave this file open at all, so that is what now
5417 29. The long-running daemon process does not normally write to the log once it
5418 has entered its main loop, and it closes the log before doing so. This is
5419 so that log files can straightforwardly be renamed and moved. However,
5420 there are a couple of unusual error situations where the daemon does write
5421 log entries, and I had neglected to close the log afterwards.
5423 30. The text of an SMTP error response that was received during a remote
5424 delivery was being truncated at 512 bytes. This is too short for some of
5425 the long messages that one sometimes sees. I've increased the limit to
5428 31. It is now possible to make retry rules that apply only when a message has a
5429 specific sender, in particular, an empty sender.
5431 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5432 it possible to be selective about when SMTP synchronization is enforced.
5434 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5436 32. Implemented hosts_connection_nolog.
5438 33. Added an ACL for QUIT.
5440 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5443 35. Added mailbox_size and mailbox_filecount to appendfile.
5445 36. Added control = no_multiline_responses to ACLs.
5447 37. There was a bug in the logic of the code that waits for the clock to tick
5448 in the case where the clock went backwards by a substantial amount such
5449 that the microsecond fraction of "now" was more than the microsecond
5450 fraction of "then" (but the whole seconds number was less).
5452 38. Added support for the libradius Radius client library this is found on
5453 FreeBSD (previously only the radiusclient library was supported).
5459 1. When certain lookups returned multiple values in the form name=value, the
5460 quoting of the values was not always being done properly. Specifically:
5461 (a) If the value started with a double quote, but contained no whitespace,
5463 (b) If the value contained whitespace other than a space character (i.e.
5464 tabs or newlines or carriage returns) it was not quoted.
5465 This fix has been applied to the mysql and pgsql lookups by writing a
5466 separate quoting function and calling it from the lookup code. The fix
5467 should probably also be applied to nisplus, ibase and oracle lookups, but
5468 since I cannot test any of those, I have not disturbed their existing code.
5470 2. A hit in the callout cache for a specific address caused a log line with no
5471 reason for rejecting RCPT. Now it says "Previous (cached) callout
5472 verification failure".
5474 3. There was an off-by-one bug in the queryprogram router. An over-long
5475 return line was truncated at 256 instead of 255 characters, thereby
5476 overflowing its buffer with the terminating zero. As well as fixing this, I
5477 have increased the buffer size to 1024 (and made a note to document this).
5479 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5480 when Exim is waiting for an SMTP response from a remote server, Exim
5481 restarts its select() call on the socket, thereby resetting its timeout.
5482 This is not a problem when such interrupts are rare. Somebody set up a cron
5483 job to run exiwhat every 2 minutes, which is less than the normal select()
5484 timeout (5 or 10 minutes). This meant that the select() timeout never
5485 kicked in because it was always reset. I have fixed this by comparing the
5486 time when an interrupt arrives with the time at the start of the first call
5487 to select(). If more time than the timeout has elapsed, the interrupt is
5488 treated as a timeout.
5490 5. Some internal re-factoring in preparation for the addition of Sieve
5491 extensions (by MH). In particular, the "personal" test is moved to a
5492 separate function, and given an option for scanning Cc: and Bcc: (which is
5493 not set for Exim filters).
5495 6. When Exim created an email address using the login of the caller as the
5496 local part (e.g. when creating a From: or Sender: header line), it was not
5497 quoting the local part when it contained special characters such as @.
5499 7. Installed new OpenBSD configuration files.
5501 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5502 try to make them clearer.
5504 9. Callout options, other than the timeout value, were being ignored when
5505 verifying sender addresses in header lines. For example, when using
5507 verify = header_sender/callout=no_cache
5509 the cache was (incorrectly) being used.
5511 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5512 only the Cygwin environment.
5514 11. When return_path_on_delivery was set as a log selector, if different remote
5515 addresses in the same message used different return paths and parallel
5516 remote delivery occurred, the wrong values would sometimes be logged.
5517 (Whenever a remote delivery process finished, the return path value from
5518 the most recently started remote delivery process was logged.)
5520 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5521 lines when AUTH and/or TLS are in use. This is the "received protocol"
5522 field. Exim used to use "asmtp" for authenticated SMTP, without any
5523 indication (in the protocol name) for TLS use. Now it follows the RFC and
5524 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5525 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5526 names appear in log lines as well as in Received: header lines.
5528 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5529 extensions, and comparison tests, and to fix some bugs.
5531 14. Changes to the "personal" filter test:
5533 (1) The test was buggy in that it was just doing the equivalent of
5534 "contains" tests on header lines. For example, if a user's address was
5535 anne@some.where, the "personal" test would incorrectly be true for
5537 To: susanne@some.where
5539 This test is now done by extracting each address from the header in turn,
5540 and checking the entire address. Other tests that are part of "personal"
5541 are now done using regular expressions (for example, to check local parts
5542 of addresses in From: header lines).
5544 (2) The list of non-personal local parts in From: addresses has been
5545 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5546 taken from the Sieve specification recommendations.
5548 (3) If the message contains any header line starting with "List-" it is
5549 treated as non-personal.
5551 (4) The test for "circular" in the Subject: header line has been removed
5552 because it now seems ill-conceived.
5554 15. Minor typos in src/EDITME comments corrected.
5556 16. Installed latest exipick from John Jetmore.
5558 17. If headers_add on a router specified a text string that was too long for
5559 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5560 of string_sprintf() is now avoided.
5562 18. $message_body_size was not set (it was always zero) when running the DATA
5563 ACL and the local_scan() function.
5565 19. For the "mail" command in an Exim filter, no default was being set for
5566 the once_repeat time, causing a random time value to be used if "once" was
5567 specified. (If the value happened to be <= 0, no repeat happened.) The
5568 default is now 0s, meaning "never repeat". The "vacation" command was OK
5569 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5570 (I found it when inspecting the code).
5572 20. There is now an overall timeout for performing a callout verification. It
5573 defaults to 4 times the callout timeout, which applies to individual SMTP
5574 commands during the callout. The overall timeout applies when there is more
5575 than one host that can be tried. The timeout is checked before trying the
5576 next host. This prevents very long delays if there are a large number of
5577 hosts and all are timing out (e.g. when the network connections are timing
5578 out). The value of the overall timeout can be changed by specifying an
5579 additional sub-option for "callout", called "maxwait". For example:
5581 verify = sender/callout=5s,maxwait=20s
5583 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5584 to the end before writing, but this should make it even safer).
5586 22. Exim was forgetting that it had advertised PIPELINING for the second and
5587 subsequent messages on an SMTP connection. It was also not resetting its
5588 memory on STARTTLS and an internal HELO.
5590 23. When Exim logs an SMTP synchronization error within a session, it now
5591 records whether PIPELINING has been advertised or not.
5593 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5594 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5595 rather than long int.
5597 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5599 26. Added the never_mail option to autoreply.
5605 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5606 crash if the getsockname() call failed; this can happen if a connection is
5607 closed very soon after it is established. The problem was simply in the
5608 order in which certain operations were done, causing Exim to try to write
5609 to the SMTP stream before it had set up the file descriptor. The bug has
5610 been fixed by making things happen in the correct order.
5616 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5617 before the connection was closed, thus losing the rejection response.
5619 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5620 some early Solaris releases, but causes trouble in current releases where
5621 socklen_t is defined.
5623 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5626 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5629 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5630 that releasing the top of it at the end releases what was used for sub-
5631 expansions (unless the block got too big). However, discard this block if
5632 the first thing is a variable or header, so that we can use its block when
5633 it is dynamic (useful for very large $message_headers, for example).
5635 6. Lookups now cache *every* query, not just the most recent. A new, separate
5636 store pool is used for this. It can be recovered when all lookup caches are
5637 flushed. Lookups now release memory at the end of their result strings.
5638 This has involved some general refactoring of the lookup sources.
5640 7. Some code has been added to the store_xxx() functions to reduce the amount
5641 of flapping under certain conditions.
5643 8. log_incoming_interface used to affect only the <= reception log lines. Now
5644 it causes the local interface and port to be added to several more SMTP log
5645 lines, for example "SMTP connection from", and rejection lines.
5647 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5649 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5651 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5652 could be overwritten at the end of the current message (or the start of a
5653 new message if it was set in a HELO ACL). The value is now preserved for
5654 the duration of the SMTP connection.
5656 12. If a transport had a headers_rewrite setting, and a matching header line
5657 contained an unqualified address, that address was qualified, even if it
5658 did not match any rewriting rules. The underlying bug was that the values
5659 of the flags that permit the existence of unqualified sender and recipient
5660 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5661 non-local messages, and by -bnq for local messages) were not being
5662 preserved with the message after it was received.
5664 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5665 "next input=" as part of the text comprising the host identity instead of
5666 the correct text. The code was using the same buffer for two different
5667 strings. However, depending on which order the printing function evaluated
5668 its arguments, the bug did not always show up. Under Linux, for example, my
5669 test suite worked just fine.
5671 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5672 doesn't work with some older versions of Perl. It has been changed to "my",
5673 which in any case is probably the better facility to use.
5675 15. A really picky compiler found some instances of statements for creating
5676 error messages that either had too many or two few arguments for the format
5679 16. The size of the buffer for calls to the DNS resolver has been increased
5680 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5681 for addresses that have a lot of PTR records. This alleviates a problem; it
5682 does not fully solve it.
5684 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5685 buffer now truncates the list and logs the incident, which is the same
5686 action as happens when Exim is looking up a host name and its aliases.
5687 Previously in this situation something unpredictable would happen;
5688 sometimes it was "internal error: store_reset failed".
5690 18. If a server dropped the connection unexpectedly when an Exim client was
5691 using GnuTLS and trying to read a response, the client delivery process
5692 crashed while trying to generate an error log message.
5694 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5695 string, for example:
5697 warn message = H1: something\nH2: something
5699 the text was added as a single header line from Exim's point of view
5700 though it ended up OK in the delivered message. However, searching for the
5701 second and subsequent header lines using $h_h2: did not work. This has been
5702 fixed. Similarly, if a system filter added multiple headers in this way,
5703 the routers could not see them.
5705 20. Expanded the error message when iplsearch is called with an invalid key to
5706 suggest using net-iplsearch in a host list.
5708 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5709 ACLs are no longer actually imposed (and a message to that effect is
5712 22. If a "gecos" field in a passwd entry contained escaped characters, in
5713 particular, if it contained a \" sequence, Exim got it wrong when building
5714 a From: or a Sender: header from that name. A second bug also caused
5715 incorrect handling when an unquoted " was present following a character
5716 that needed quoting.
5718 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5719 was not being matched caselessly.
5721 24. Arranged for all hyphens in the exim.8 source to be escaped with
5724 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5725 properly. Recipient callout cache records were still being keyed to include
5726 the sender, even when use_sender was set false. This led to far more
5727 callouts that were necessary. The sender is no longer included in the key
5728 when use_sender is false.
5730 26. Added "control = submission" modifier to ACLs.
5732 27. Added the ${base62d: operator to decode base 62 numbers.
5734 28. dnsdb lookups can now access SRV records.
5736 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5737 the configuration file.
5739 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5740 (-v) mode. This makes the output for a verbose queue run more intelligible.
5742 31. Added a use_postmaster feature to recipient callouts.
5744 32. Added the $body_zerocount variable, containing the number of binary zero
5745 bytes in the message body.
5747 33. The time of last modification of the "new" subdirectory is now used as the
5748 "mailbox time last read" when there is a quota error for a maildir
5751 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5753 35. Added +ignore_unknown as a special item in host lists.
5755 36. Code for decoding IPv6 addresses in host lists is now included, even if
5756 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5757 address was recognized as an IP address, but was then not correctly decoded
5758 into binary, causing unexpected and incorrect effects when compared with
5765 1. Very minor rewording of debugging text in manualroute to say "list of
5766 hosts" instead of "hostlist".
5768 2. If verify=header_syntax was set, and a header line with an unqualified
5769 address (no domain) and a large number of spaces between the end of the
5770 name and the colon was received, the reception process suffered a buffer
5771 overflow, and (when I tested it) crashed. This was caused by some obsolete
5772 code that should have been removed. The fix is to remove it!
5774 3. When running in the test harness, delay a bit after writing a bounce
5775 message to get a bit more predictability in the log output.
5777 4. Added a call to search_tidyup() just before forking a reception process. In
5778 theory, someone could use a lookup in the expansion of smtp_accept_max_
5779 per_host which, without the tidyup, could leave open a database connection.
5781 5. Added the variables $recipient_data and $sender_data which get set from a
5782 lookup success in an ACL "recipients" or "senders" condition, or a router
5783 "senders" option, similar to $domain_data and $local_part_data.
5785 6. Moved the writing of debug_print from before to after the "senders" test
5788 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5789 problems for message scanning, either using a data ACL, or using
5790 local_scan() because the Received: header was not generated till after they
5791 were called (in order to set the time as the time of reception completion).
5792 I have revised the way this works. The header is now generated after the
5793 body is received, but before the ACL or local_scan() are called. After they
5794 are run, the timestamp in the header is updated.
5800 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5801 before starting a queue runner without re-exec. This happened only when
5802 deliver_drop_privilege was set or when the Exim user was set to root. The
5803 effect of the bug was that timeouts during subsequent deliveries caused
5804 crashes instead of being properly handled. The handler is now left at its
5805 default (and expected) setting.
5807 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5808 message, again when deliver_drop_privilege is set or Exim is run as root.
5809 The bug described in (1) was not present in this case, but the tidying up
5810 of the other signals was missing. I have made the two cases consistent.
5812 3. The ignore_target_hosts setting on a manualroute router was being ignored
5813 for hosts that were looked up using the /MX notation.
5815 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5818 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5819 operated on the sender address. After changing the $sender_address to <>
5820 for the sender address verify, Exim was re-instated it as the original
5821 (before rewriting) address, but remembering that it had rewritten it, so it
5822 wasn't rewriting it again. This bug also had the effect of breaking the
5823 sender address verification caching when the sender address was rewritten.
5825 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5826 This has been changed so that if the ip literal address matches
5827 ignore_target_hosts, the router declines.
5829 7. Added expansion conditions match_domain, match_address, and match_local_
5830 part (NOT match_host).
5832 8. The placeholder for the Received: header didn't have a length field set.
5834 9. Added code to Exim itself and to exim_lock to test for a specific race
5835 condition that could lead to file corruption when using MBX delivery. The
5836 issue is with the lockfile that is created in /tmp. If this file is removed
5837 after a process has opened it but before that process has acquired a lock,
5838 there is the potential for a second process to recreate the file and also
5839 acquire a lock. This could lead to two Exim processes writing to the file
5840 at the same time. The added code performs the same test as UW imapd; it
5841 checks after acquiring the lock that its file descriptor still refers to
5842 the same named file.
5844 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5845 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5848 11. Added the smtp_active_hostname option. If used, this will typically be made
5849 to depend on the incoming interface address. Because $interface_address is
5850 not set up until the daemon has forked a reception process, error responses
5851 that can happen earlier (such as "too many connections") no longer contain
5854 12. If an expansion in a condition on a "warn" statement fails because a lookup
5855 defers, the "warn" statement is abandoned, and the next ACL statement is
5856 processed. Previously this caused the whole ACL to be aborted.
5858 13. Added the iplsearch lookup type.
5860 14. Added ident_timeout as a log selector.
5862 15. Added tls_certificate_verified as a log selector.
5864 16. Added a global option tls_require_ciphers (compare the smtp transport
5865 option of the same name). This controls incoming TLS connections.
5867 17. I finally figured out how to make tls_require_ciphers do a similar thing
5868 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5869 before starting the TLS session.
5871 18. Tabs are now shown as \t in -bP output.
5873 19. If the log selector return_path_on_delivery was set, Exim crashed when
5874 bouncing a message because it had too many Received: header lines.
5876 20. If two routers both had headers_remove settings, and the first one included
5877 a superfluous trailing colon, the final name in the first list and the
5878 first name in the second list were incorrectly joined into one item (with a
5879 colon in the middle).
5885 1. Added -C and -D options to the exinext utility, mainly to make it easier
5886 to include in the automated testing, but these could be helpful when
5887 multiple configurations are in use.
5889 2. The exinext utility was not formatting the output nicely when there was
5890 an alternate port involved in the retry record key, nor when there was a
5891 message id as well (for retries that were specific to a specific message
5892 and a specific host). It was also confused by IPv6 addresses, because of
5893 the additional colons they contain. I have fixed the IPv4 problem, and
5894 patched it up to do a reasonable job for IPv6.
5896 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5897 delivery, the log line now contains "pipelined" if PIPELINING was used.
5899 4. An SMTP transport process used to panic and die if the bind() call to set
5900 an explicit outgoing interface failed. This has been changed; it is now
5901 treated in the same way as a connect() failure.
5903 5. A reference to $sender_host_name in the part of a conditional expansion
5904 that was being skipped was still causing a DNS lookup. This no longer
5907 6. The def: expansion condition was not recognizing references to header lines
5908 that used bh_ and bheader_.
5910 7. Added the _cache feature to named lists.
5912 8. The code for checking quota_filecount in the appendfile transport was
5913 allowing one more file than it should have been.
5915 9. For compatibility with Sendmail, the command line option
5923 and sets the incoming protocol and host name (for trusted callers). The
5924 host name and its colon can be omitted when only the protocol is to be set.
5925 Note the Exim already has two private options, -pd and -ps, that refer to
5926 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5927 "s", but I don't think that's a major issue.
5929 10. A number of refactoring changes to the code, none of which should affect
5932 (a) The number of logging options was getting close to filling up the
5933 32-bit word that was used as a bit map. I have split them into two classes:
5934 those that are passed in the argument to log_write(), and those that are
5935 only ever tested independently outside of that function. These are now in
5936 separate 32-bit words, so there is plenty of room for expansion again.
5937 There is no change in the user interface or the logging behaviour.
5939 (b) When building, for example, log lines, the code previously used a
5940 macro that called string_cat() twice, in order to add two strings. This is
5941 not really sufficiently general. Furthermore, there was one instance where
5942 it was actually wrong because one of the argument was used twice, and in
5943 one call a function was used. (As it happened, calling the function twice
5944 did not affect the overall behaviour.) The macro has been replaced by a
5945 function that can join an arbitrary number of extra strings onto a growing
5948 (c) The code for expansion conditions now uses a table and a binary chop
5949 instead of a serial search (which was left over from when there were very
5950 few conditions). Also, it now recognizes conditions like "pam" even when
5951 the relevant support is not compiled in: a suitably worded error message is
5952 given if an attempt is made to use such a condition.
5954 11. Added ${time_interval:xxxxx}.
5956 12. A bug was causing one of the ddress fields not to be passed back correctly
5957 from remote delivery subprocesses. The field in question was not being
5958 subsequently used, so this caused to problems in practice.
5960 13. Added new log selectors queue_time and deliver_time.
5962 14. Might have fixed a bug in maildirsizefile handling that threw up
5963 "unexpected character" debug warnings, and recalculated the data
5964 unnecessarily. In any case, I expanded the warning message to give more
5967 15. Added the message "Restricted characters in address" to the statements in
5968 the default ACL that block characters like @ and % in local parts.
5970 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5971 Three changes have been made:
5973 (a) There was a serious bug; a negative response to MAIL caused the whole
5974 recipient domain to be cached as invalid, thereby blocking all messages
5975 to all local parts at the same domain, from all senders. This bug has
5976 been fixed. The domain is no longer cached after a negative response to
5977 MAIL if the sender used is not empty.
5979 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5982 (c) A new callout option, "use_sender" has been added for people who want
5983 the modified behaviour.
5989 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5992 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5993 indeed breaks things for older releases.
5995 3. Added additional logging to the case where there is a problem reading data
5996 from a filter that is running in a subprocess using a pipe, in order to
5997 try to track down a specific problem.
5999 4. Testing facility fudge: when running in the test harness and attempting
6000 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6001 getting "No route to host". Convert this to a timeout.
6003 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6006 6. Some OS don't have socklen_t but use size_t instead. This affects the
6007 fifth argument of getsockopt() amongst other things. This is now
6008 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6009 can be set for individual OS. I have set it for SunOS5, OSF1, and
6010 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6011 some earlier ones do not.
6013 7. Change 4.30/15 was not doing the test caselessly.
6015 8. The standard form for an IPv6 address literal was being rejected by address
6016 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6017 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6018 this, as well as the form without the "IPv6" on the front (but only when
6019 address literals are enabled, of course).
6021 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6023 10. Exim crashed if a message with an empty sender address specified by -f
6024 encountered a router with an errors_to setting. This could be provoked only
6025 by a command such as
6029 where an empty string was supplied; "<>" did not hit this bug.
6031 11. Installed PCRE release 4.5.
6033 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6034 remained set. It is now erased.
6036 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6037 times from message ids (which are base 36 rather than the normal 62).
6039 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6040 were being counted as actual protocol errors, and logged if the log
6041 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6042 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6043 and DATA following a set of rejected RCPTs do not count as protocol errors.
6044 In other words, Exim assumes they were pipelined, though this may not
6045 actually be the case. Of course, in all cases the client gets an
6046 appropriate error code.
6048 15. If a lookup fails in an ACL condition, a message about the failure may
6049 be available; it is used if testing the ACL cannot continue, because most
6050 such messages specify what the cause of the deferral is. However, some
6051 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6052 that caused an old message to be retained and used if a later statement
6053 caused a defer, replacing the real cause of the deferral.
6055 16. If an IP address had so many PTR records that the DNS lookup buffer
6056 was not large enough to hold them, Exim could crash while trying to process
6057 the truncated data. It now detects and logs this case.
6059 17. Further to 4.21/58, another change has been made: if (and only if) the
6060 first line of a message (the first header line) ends with CRLF, a bare LF
6061 in a subsequent header line has a space inserted after it, so as not to
6062 terminate the header.
6064 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6065 unnecessarily, used atoi() instead of strtol(), and didn't check the
6066 termination when getting file sizes from file names by regex.
6068 19. Completely re-implemented the support for maildirsize files, in the light
6069 of a number of problems with the previous contributed implementation
6070 (4.30/29). In particular:
6072 . If the quota is zero, the maildirsize file is maintained, but no quota is
6075 . If the maildir directory does not exist, it is created before any attempt
6076 to write a maildirsize file.
6078 . The quota value in the file is just a cache; if the quota is changed in
6079 the transport, the new value overrides.
6081 . A regular expression is available for excluding directories from the
6084 20. The autoreply transport checks the characters in options that define the
6085 message's headers; it allows continued headers, but it was checking with
6086 isspace() after an embedded newline instead of explicitly looking for a
6089 21. If all the "regular" hosts to which an address was routed had passed their
6090 expiry times, and had not reached their retry times, the address was
6091 bounced, even if fallback hosts were defined. Now Exim should go on to try
6094 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6095 equivalent code in the SMTP transport. Some hosts send humungous responses
6096 to HELO/EHLO, more than 1024 it seems.
6098 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6099 gives compiler warnings in some environments. I've now done it "properly",
6102 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6103 (because of problems with the built-in one) was declared to return uschar *
6104 instead of char *, causing compiler failure.
6106 25. Fixed a file descriptor leak when processing alias/forward files.
6108 26. Fixed a minor format string issue in dbfn.c.
6110 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6112 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6113 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6116 29. When a sender address is verified, it is cached, to save repeating the test
6117 when there is more than one recipient in a message. However, when the
6118 verification involves a callout, it is possible for different callout
6119 options to be set for different recipients. It is too complicated to keep
6120 track of this in the cache, so now Exim always runs a verification when a
6121 callout is required, relying on the callout cache for the optimization.
6122 The overhead is duplication of the address routing, but this should not be
6125 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6126 to be verified with callout=postmaster, and the main callout worked but the
6127 postmaster check failed, the verification correctly failed. However, if a
6128 subsequent RCPT command asked for sender verification *without* the
6129 postmaster check, incorrect caching caused this verification also to fail,
6132 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6133 it was not caching the DNS options (qualify_single, search_parents) that
6134 were used when the lookup failed. A subsequent lookup with different
6135 options therefore always gave the same answer, though there were cases
6136 where it should not have. (Example: a "domains = !$mx_any" option on a
6137 dnslookup router: the "domains" option is always processed without any
6138 widening, but the router might have qualify_single set.) Now Exim uses the
6139 cached value only when the same options are set.
6141 32. Added John Jetmore's "exipick" utility to the distribution.
6143 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6144 than a timeout (e.g. a certificate is required, and is not provided), an
6145 Exim server now closes the connection immediately. Previously it waited for
6146 the client to close - but if the client is SSL, it seems that they each
6147 wait for each other, leading to a delay before one of them times out.
6149 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6150 maintained 0.8.x compatibility because I don't think many are using it, and
6151 it is clearly obsolete.
6153 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6156 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6157 client certificate was expired. A simple patch fixes this, though I don't
6158 understand the full logic of why the verify callback is called multiple
6161 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6162 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6163 which causes problems with some clients (such as the Certicom SSL Plus
6164 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6165 disables the coutermeasure allowing Eudora to connect."
6167 38. Exim was not checking that a write() to a log file succeeded. This could
6168 lead to Bad Things if a log got too big, in particular if it hit a file
6169 size limit. Exim now panics and dies if it cannot write to a log file, just
6170 as it does if it cannot open a log file.
6172 39. Modified OS/Makefile-Linux so that it now contains
6174 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6176 The two -D definitions ensure that Exim is compiled with large file
6177 support, which makes it possible to handle log files that are bigger than
6180 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6181 instance) a domain was checked against a named list that involved a lookup,
6182 causing $domain_data to be set, then another domain was checked against the
6183 same list, then the first domain was re-checked, the value of $domain_data
6184 after the final check could be wrong. In particular, if the second check
6185 failed, it could be set empty. This bug probably also applied to
6188 41. The strip_trailing_dot option was not being applied to the address given
6189 with the -f command-line option.
6191 42. The code for reading a message's header from the spool was incrementing
6192 $received_count, but never initializing it. This meant that the value was
6193 incorrect (doubled) while delivering a message in the same process in which
6194 it was received. In the most common configuration of Exim, this never
6195 happens - a fresh exec is done - but it can happen when
6196 deliver_drop_privilege is set.
6198 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6199 it now includes up to 150 characters of the unexpected data in the log
6202 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6203 and building data strings. The size of both of these buffers was 10 000
6204 bytes - far larger than anybody would *ever* want, thought I. Needless to
6205 say, somebody hit the limit. I have increased the maximum line length to
6206 20 000 and the maximum data length of concatenated lines to 100 000. I have
6207 also fixed two bugs, because there was no checking on these buffers. Tsk,
6208 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6209 buffer is too small.
6211 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6212 lsearch lookups. Now it does.
6214 46. When parsing a route_list item in a manualroute router, a fixed-length
6215 buffer was used for the list of hosts. I made this 1024 bytes long,
6216 thinking that nobody would ever have a list of hosts that long. Wrong.
6217 Somebody had a whole pile of complicated expansion conditions, and the
6218 string was silently truncated, leading to an expansion error. It turns out
6219 that it is easier to change to an unlimited length (owing to other changes
6220 that have happened since this code was originally written) than to build
6221 structure for giving a limitation error. The length of the item that
6222 expands into the list of hosts is now unlimited.
6224 47. The lsearch lookup could not handle data where the length of text line was
6225 more than 4095 characters. Such lines were truncated, leading to shortened
6226 data being returned. It should now handle lines of any length.
6228 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6229 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6232 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6233 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6234 no longer happens in comment lines. A list of replacements is now placed
6235 at the head of all of the source files, except those whose only change is
6236 to replace PERL_COMMAND in the very first #! line.
6238 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6239 messages on the queue, with a bottom-up merge sort, using code contributed
6240 by Michael Haardt. This should make operations like -bp somewhat faster on
6241 large queues. It won't affect queue runners, except when queue_run_in_order
6244 51. Installed eximstats 1.31 in the distribution.
6246 52. Added support for SRV lookups to the dnslookup router.
6248 53. If an ACL referred to $message_body or $message_body_end, the value was not
6249 reset for any messages that followed in the same SMTP session.
6251 54. The store-handling optimization for building very long strings was not
6252 differentiating between the different store pools. I don't think this
6253 actually made any difference in practice, but I've tidied it.
6255 55. While running the routers to verify a sender address, $sender_address
6256 was still set to the sender address. This is wrong, because when routing to
6257 send a bounce to the sender, it would be empty. Therefore, I have changed
6258 it so that, while verifying a sender address, $sender_address is set to <>.
6259 (There is no change to what happens when verifying a recipient address.)
6261 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6262 target A or AAAA records (if not already returned) without resetting the
6263 qualify_single or search_parents options of the DNS resolver. These are
6264 inappropriate in this case because the targets of MX and SRV records must
6265 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6266 target that, when qualified, matched something in the local domain. These
6267 two options are now turned off when doing these lookups.
6269 57. It seems that at least some releases of Reiserfs (which does not have the
6270 concept of a fixed number of inodes) returns zero and not -1 for the
6271 number of available inodes. This interacted badly with check_spool_inodes,
6272 which assumed that -1 was the "no such thing" setting. What I have done is
6273 to check that the total number of inodes is greater than zero before doing
6274 the test of how many are available.
6276 58. When a "warn" ACL statement has a log_message modifier, the message is
6277 remembered, and not repeated. This is to avoid a lot of repetition when a
6278 message has many recipients that cause the same warning to be written.
6279 However, Exim was preserving the list of already written lines for an
6280 entire SMTP session, which doesn't seem right. The memory is now reset if a
6281 new message is started.
6283 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6284 debugging output unless log_rewrite was also set.
6286 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6287 of (int)(handle) when we know that handle contains (void *)(-1).
6289 61. The Exim daemon panic-logs an error return when it closes the incoming
6290 connection. However "connection reset by peer" seems to be common, and
6291 isn't really an error worthy of noting specially, so that particular error
6294 62. When Exim is trying to find all the local interfaces, it used to panic and
6295 die if the ioctl to get the interface flags failed. However, it seems that
6296 on at least one OS (Solaris 9) it is possible to have an interface that is
6297 included in the list of interfaces, but for which you get a failure error
6298 for this call. This happens when the interface is not "plumbed" into a
6299 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6300 failure of the "get flags" call assumes that the interface is down.
6302 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6303 makes life easier for people who are doing arithmetic on fields extracted
6304 from dates, where you often get leading zeros that should not be
6305 interpreted as octal.
6307 64. Added qualify_domain to the redirect router, to override the global
6310 65. If a pathologically long header line contained very many addresses (the
6311 report of this problem mentioned 10 000) and each of them was rewritten,
6312 Exim could use up a very large amount of memory. (It kept on making new
6313 copies of the header line as it rewrote, and never released the old ones.)
6314 At the expense of a bit more processing, the header rewriting function has
6315 been changed so that it no longer eats memory in this way.
6317 66. The generation of the Received: header has been moved from the time that a
6318 message starts to be received, to the time that it finishes. The timestamp
6319 in the Received: header should now be very close to that of the <= log
6320 line. There are two side-effects of this change:
6322 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6323 logged header lines no longer include the local Received: line, because
6324 it has not yet been created. The same applies to a copy of the message
6325 that is returned to a non-SMTP sender when a message is rejected.
6327 (b) When a filter file is tested using -bf, no additional Received: header
6328 is added to the test message. After some thought, I decided that this
6331 This change does not affect the value of $received_for. It is still set
6332 after address rewriting, but before local_scan() is called.
6334 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6336 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6337 gave an unhelpful panic error message, and a defer error. I have managed to
6338 change this behaviour so that it now rejects any supplied certificate,
6339 which seems right, as the list of acceptable certificates is empty.
6341 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6342 gave an unhelpful defer error. I have not managed to make this reject any
6343 supplied certificates, but the error message it gives is "no certificate
6344 supplied", which is not helpful.
6346 70. exigrep's output now also includes lines that are not associated with any
6347 message, but which match the given pattern. Implemented by a patch from
6348 Martin Sluka, which also tidied up the Perl a bit.
6350 71. Recipient callout verification, like sender verification, was using <> in
6351 the MAIL FROM command. This isn't really the right thing, since the actual
6352 sender may affect whether the remote host accepts the recipient or not. I
6353 have changed it to use the actual sender in the callout; this means that
6354 the cache record is now keyed on a recipient/sender pair, not just the
6355 recipient address. There doesn't seem to be a real danger of callout loops,
6356 since a callout by the remote host to check the sender would use <>.
6357 [SEE ABOVE: changed after hitting problems.]
6359 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6360 temporary errors. However, in the case of such a code being given after
6361 the end of a data transmission (i.e. after ".") Exim was failing to write
6362 a retry record for the message. (Yes, there was some broken host that was
6363 actually sending 8xx at this point.)
6365 73. An unknown lookup type in a host list could cause Exim to panic-die when
6366 the list was checked. (An example that provoked this was putting <; in the
6367 middle of a list instead of at the start.) If this happened during a DATA
6368 ACL check, a -D file could be left lying around. This kind of configuration
6369 error no longer causes Exim to die; instead it causes a defer error. The
6370 incident is still logged to the main and panic logs.
6372 74. Buglet left over from Exim 3 conversion. The message "too many messages
6373 in one connection" was written to the rejectlog but not the mainlog, except
6374 when address rewriting (yes!) was being logged.
6376 75. Added write_rejectlog option.
6378 76. When a system filter was run not as root (that is, when system_filter_user
6379 was set), the values of the $n variables were not being returned to the
6380 main process; thus, they were not subsequently available in the $sn
6383 77. Added +return_path_on_delivery log selector.
6385 78. A connection timeout was being treated differently from recipients deferred
6386 when testing hosts_max_try with a message that was older than the host's
6387 retry timeout. (The host should not be counted, thus allowing all hosts to
6388 be tried at least once before bouncing.) This may have been the cause of an
6389 occasionally reported bug whereby a message would remain on the queue
6390 longer than the retry timeout, but would be bounced if a delivery was
6391 forced. I say "may" because I never totally pinned down the problem;
6392 setting up timeout/retry tests is difficult. See also the next item.
6394 79. The ultimate address timeout was not being applied to errors that involved
6395 a combination of host plus message (for example, a timeout on a MAIL
6396 command). When an address resolved to a number of possible hosts, and they
6397 were not all tried for each delivery (e.g. because of hosts_max_try), a
6398 message could remain on the queue longer than the retry timeout.
6400 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6403 81. Fixed an obscure SMTP outgoing bug which required at least the following
6404 conditions: (a) there was another message waiting for the same server;
6405 (b) the server returned 5xx to all RCPT commands in the first message so
6406 that the message was not completed; (c) the server dropped the connection
6407 or gave a negative response to the RSET that Exim sends to abort the
6408 transaction. The observed case was a dropped connection after DATA that had
6409 been sent in pipelining mode. That is, the server had advertised PIPELINING
6410 but was not implementing it correctly. The effect of the bug was incorrect
6411 behaviour, such as trying another host, and this could lead to a crash.
6417 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6418 and daemon.c were passed as pointers to ints; they should have been
6419 pointers to socklen_t variables (which are typically unsigned ints).
6421 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6424 3. Fixed a really odd bug that affected only the testing scheme; patching a
6425 certain fixed string in the binary changed the value of another string that
6426 happened to be identical to the end of the original first string.
6428 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6429 name", it returns that address as the IP address. On some operating
6430 systems (e.g. Solaris), it also passes back the IP address string as the
6431 "host name". However, on others (e.g. Linux), it passes back an empty
6432 string. Exim wasn't checking for this, and was changing the host name to an
6433 empty string, assuming it had been canonicalized.
6435 5. Although rare, it is permitted to have more than one PTR record for a given
6436 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6437 all the names associated with an address, because they do in Solaris.
6438 However, it seems that they do not in Linux for data that comes from the
6439 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6440 I found this out when I moved to a new Linux workstation and tried to run
6441 the Exim test suite.
6443 To get round this problem I have changed the code so that it now does its
6444 own call to the DNS to look up PTR records when searching for a host name.
6445 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6446 addresses that are only in /etc/hosts are still found.
6448 This behaviour is, however, controlled by an option called host_lookup_
6449 order, which defaults to "bydns:byaddr". If people want to use the other
6450 order, or indeed, just use one or the other means of lookup, they can
6451 specify it in this variable.
6453 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6454 some operating systems, this comes back from gethostbyaddr() as an empty
6455 string, and this is what Exim used to test for. However, it seems that in
6456 other systems, "." is yielded. Exim now tests for this case too.
6458 7. The values of check_spool_space and check_log_space are now held internally
6459 as a number of kilobytes instead of an absolute number of bytes. If a
6460 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6461 kilobyte. This means that much larger values can be stored.
6463 8. Exim monitor: an attempt to get the action menu when not actually pointing
6464 at a message produces an empty menu entitled "No message selected". This
6465 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6466 no entries in it ("Shell widget menu has zero width and/or height"). So I
6467 have added a single, blank menu entry in this case.
6469 9. Added ${quote_local_part.
6471 10. MIME decoding is now applied to the contents of Subject: header lines when
6474 11. Now that a reference to $sender_host_address automatically causes a reverse
6475 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6476 host lookup before query-style lookups in lists that might use this
6477 variable. This has therefore been abolished, and the "net-" prefix is no
6478 longer necessary for query-style lookups.
6480 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6481 have been a typo for LFLAGS, so it has been changed.
6483 13. The install script calls Exim with "-C /dev/null" in order to find the
6484 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6485 to be output. However, since Exim outputs its version number before the
6486 error, it didn't break the script. It just looked ugly. I fixed this by
6487 always allowing "-C /dev/null" if the caller is root.
6489 14. Ignore overlarge ACL variable number when reading spool file - insurance
6490 against a later release with more variables having written the file.
6492 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6493 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6494 this, as well as the form without the "IPv6" on the front.
6496 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6497 OS/Makefile-Darwin file.
6499 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6500 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6502 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6503 variable. However, some OS use ints for inodes. I've added cast to long int
6504 to get rid of the compiler warning.
6506 19. I had forgotten to lock out "/../" in configuration file names when
6507 ALT_CONFIG_PREFIX was set.
6509 20. Routers used for verification do not need to specify transports. However,
6510 if such a router generated a host list, and callout was configured, Exim
6511 crashed, because it could not find a port number from the (non-existent)
6512 transport. It now assumes port 25 in this circumstance.
6514 21. Added the -t option to exigrep.
6516 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6517 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6518 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6519 them is set, LOOKUP_LSEARCH is forced.
6521 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6523 24. Added sender and host information to the "rejected by local_scan()" log
6524 line; previously there was no indication of these.
6526 25. Added .include_if_exists.
6528 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6529 Linux. It turns out that not all file systems support this. Apparently some
6530 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6531 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6532 not supported on the file descriptor, is now ignored when Exim is trying to
6533 sync a directory. This applies only to Linux.
6535 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6537 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6538 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6541 29. Added support for maildirsize files from supplied patch (modified a bit).
6543 30. The use of :fail: followed by an empty string could lead Exim to respond to
6544 sender verification failures with (e.g.):
6546 550 Verification failed for <xxx>
6547 550 Sender verify failed
6549 where the first response line was missing the '-' that indicates it is not
6550 the final line of the response.
6552 31. The loop for finding the name of the user that called Exim had a hardwired
6553 limit of 10; it now uses the value of finduser_retries, which is used for
6554 all other user lookups.
6556 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6559 33. Exim was neglecting to zero errno before one call of strtol() when
6560 expanding a string and expecting an integer value. On some systems this
6561 resulted in spurious "integer overflow" errors. Also, it was casting the
6562 result into an int without checking.
6564 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6565 did not work. The code looks as if it has *never* worked, though it appears
6566 to have been documented since at least release 1.62. I have made it work.
6568 35. The "timeout_DNS" error in retry rules, also documented since at least
6569 1.62, also never worked. As it isn't clear exactly what this means, and
6570 clearly it isn't a major issue, I have abolished the feature by treating it
6571 as "timeout", and writing a warning to the main and panic logs.
6573 36. The display of retry rules for -brt wasn't always showing the error code
6576 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6577 timeout_connect_A, timeout_connect_MX.
6579 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6580 to the empty sender.
6582 39. The daemon was not analysing the content of -oX till after it had closed
6583 stderr and disconnected from the controlling terminal. This meant that any
6584 syntax errors were only noted on the panic log, and the return code from
6585 the command was 0. By re-arranging the code a little, I've made the
6586 decoding happen first, so such errors now appear on stderr, and the return
6587 code is 1. However, the actual setting up of the sockets still happens in
6588 the disconnected process, so errors there are still only recorded on the
6591 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6592 connections (as happens on some IP stacks) was logged at start up time as
6593 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6594 it from "IPv6 and IPv4", which means that two separate sockets are being
6597 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6598 says whether AF_INET or AF_INET6 was passed as an argument.
6600 42. Exiwhat output was messed up when time zones were included in log
6603 43. Exiwhat now gives more information about the daemon's listening ports,
6604 and whether -tls-on-connect was used.
6606 44. The "port" option of the smtp transport is now expanded.
6608 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6609 silently ignored. Now an error message is written to the main and panic
6612 46. There's a new ACL modifier called "logwrite" which writes to a log file
6613 as soon as it is encountered.
6615 47. Added $local_user_uid and $local_user_gid at routing time.
6617 48. Exim crashed when trying to verify a sender address that was being
6620 49. Exim was recognizing only a space character after ".include". It now also
6621 recognizes a tab character.
6623 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6624 extracting the relevant information from the specification. The man page no
6625 longer contains scrambled data for the -d option, and I've added a section
6626 at the front about calling Exim under different names.
6628 51. Added "extra_headers" argument to the "mail" command in filter files.
6630 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6633 53. Installed eximstats 1.29.
6635 54. Added transport_filter_timeout as a generic transport option.
6637 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6638 Cc: header lines. This was required by RFC 822, but it not required by RFC
6641 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6642 incoming messages that did not have them. Now it does so only if the
6643 message originates locally, that is, if there is no associated remote host
6644 address. When Resent- header lines are present, this applies to the Resent-
6645 lines rather than the non-Resent- lines.
6647 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6648 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6650 58. Messages for configuration errors now include the name of the main
6651 configuration file - useful now that there may be more than one file in a
6652 list (.included file names were always shown).
6654 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6655 for those rare installations that do not start the daemon as root or run it
6656 setuid root. I've cut out the call to initgroups() if the daemon is not
6659 60. The Exim user and group can now be bound into the binary as text strings
6660 that are looked up at the start of Exim's processing.
6662 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6664 62. Added $mailstore_basename variable.
6666 63. Installed patch to sieve.c from Michael Haardt.
6668 64. When Exim failed to open the panic log after failing to open the main log,
6669 the original message it was trying to log was written to stderr and debug
6670 output, but if they were not available (the usual case in production), it
6671 was lost. Now it is written to syslog before the two lines that record the
6672 failures to open the logs.
6674 65. Users' Exim filters run in subprocesses under the user's uid. It is
6675 possible for a "deliver" command or an alias in a "personal" command to
6676 provoke an address rewrite. If logging of address rewriting is configured,
6677 this fails because the process is not running as root or exim. There may be
6678 a better way of dealing with this, but for the moment (because 4.30 needs
6679 to be released), I have disabled address rewrite logging when running a
6680 filter in a non-root, non-exim process.
6686 1. The buildconfig auxiliary program wasn't quoting the value set for
6687 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6688 not defined. This bug was masked in 4.22 by the effect that was fixed in
6691 2. Some messages that were rejected after a message id was allocated were
6692 shown as "incomplete" by exigrep. It no longer does this for messages that
6693 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6695 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6696 have allow_domain_literals set, the ID did not get logged in the <= line.
6697 Domain literals are now always recognized in Message-ID: header lines.
6699 4. The first argument for a ${extract expansion item is the key name or field
6700 number. Leading and trailing spaces in this item were not being ignored,
6701 causing some misleading effects.
6703 5. When deliver_drop_privilege was set, single queue runner processes started
6704 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6705 same command in the process it spins off) were not dropping privilege.
6707 6. When the daemon running as "exim" started a queue runner, it always
6708 re-executed Exim in the spun-off process. This is a waste of effort when
6709 deliver_drop_privilege is set. The new process now just calls the
6710 queue-runner function directly.
6716 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6719 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6720 ignored. Though the use of -oP was forcing the writing of a pid file, it
6721 was always written to the default place.
6723 3. If the message "no IP address found for host xxxx" is generated during
6724 incoming verification, it is now followed by identification of the incoming
6725 connection (so you can more easily find what provoked it).
6727 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6729 5. Added some features to "harden" Exim a bit more against certain attacks:
6731 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6732 be put in Local/Makefile. This is like the never_users runtime option,
6733 but it cannot be overridden. The default setting is "root".
6735 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6736 prefix string with which any file named in a -C command line option
6739 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6740 is retained for -C and -D only if the caller of Exim is root. Without
6741 it, the exim user may also use -C and -D and retain privilege.
6743 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6744 command line option is disabled.
6746 6. Macro names set by the -D option must start with an upper case letter, just
6747 like macro names defined in the configuration file.
6749 7. Added "dereference=" facility to LDAP.
6751 8. Two instances of the typo "uknown" in the source files are fixed.
6753 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6754 the Configure-Makefile script screwed up while processing it.
6756 10. Incorporated PCRE 4.4.
6758 11. The SMTP synchronization check was not operating right at the start of an
6759 SMTP session. For example, it could not catch a HELO sent before the client
6760 waited for the greeting. There is now a check for outstanding input at the
6761 point when the greeting is written. Because of the duplex, asynchronous
6762 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6763 way, but not yet received, when the check is performed.
6765 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6766 on TCP/IP sockets, because this apparently causes some broken clients to
6769 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6770 unchanged) from the Cygwin maintainer.
6772 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6773 was supported instead of testing for mbx. Effectively a typo.
6775 15. The spa authenticator server code was not checking that the input it
6776 received was valid base64.
6778 16. The debug output line for the "set" modifier in ACLs was not showing the
6779 name of the variable that was being set.
6781 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6783 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6784 lookup on its own. Something else was needed to trigger the lookup. For
6785 example, a match in host_lookup or the need for a host name in a host list.
6786 Now, if $sender_host_name is referenced and the host name has not yet been
6787 looked up, a lookup is performed. If the lookup fails, the variable remains
6788 empty, and $host_lookup_failed is set to "1".
6790 19. Added "eqi" as a case-independent comparison operator.
6792 20. The saslauthd authentication condition could segfault if neither service
6793 nor realm was specified.
6795 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6796 error message that was logged was misleading, and incoming SMTP
6797 connections were dropped. The message is now more accurate, and temporary
6798 errors are given to SMTP connections.
6800 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6801 (except RSET) with a 421 error, until QUIT is received. However, it was
6802 failing to send a response to QUIT.
6804 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6805 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6806 ACL is now run after the helo_try_verify_hosts code.
6808 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6809 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6810 case-independent, so other case variants are also recognized). Apparently
6811 some systems use these upper case variants.
6813 25. If more than two messages were waiting for the same host, and a transport
6814 filter was specified for the transport, Exim sent two messages over the
6815 same TCP/IP connection, and then failed with "socket operation on non-
6816 socket" when it tried to send the third.
6818 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6820 27. The extern definition of crypt16() in expand.c was not being excluded when
6821 the OS had its own crypt16() function.
6823 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6824 as a preferred synonym for return_size_limit, both as an option and as an
6827 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6829 30. Changed the default configuration ACL to relax the local part checking rule
6830 for addresses that are not in any local domains. For these addresses,
6831 slashes and pipe symbols are allowed within local parts, but the sequence
6832 /../ is explicitly forbidden.
6834 31. SPA server authentication was not clearing the challenge buffer before
6837 32. log_message in a "warn" ACL statement was writing to the reject log as
6838 well as to the main log, which contradicts the documentation and doesn't
6839 seem right (because no rejection is happening). So I have stopped it.
6841 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6842 However, I am unable to do any testing of this.
6844 34. Fixed an infelicity in the appendfile transport. When checking directories
6845 for a mailbox, to see if any needed to be created, it was accidentally
6846 using path names with one or more superfluous leading slashes; tracing
6847 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6849 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6850 contents are added to the log line that is written for every discarded
6851 recipient. (Previously a log_message setting was ignored.)
6853 36. The ${quote: operator now quotes the string if it is empty.
6855 37. The install script runs exim in order to find its version number. If for
6856 some reason other than non-existence or emptiness, which it checks, it
6857 could not run './exim', it was installing it with an empty version number,
6858 i.e. as "exim-". This error state is now caught, and the installation is
6861 38. An argument was missing from the function that creates an error message
6862 when Exim fails to connect to the socket for saslauthd authentication.
6863 This could cause Exim to crash, or give a corrupted message.
6865 39. Added isip, isip4, and isip6 to ${if conditions.
6867 40. The ACL variables $acl_xx are now saved with the message, and can be
6868 accessed later in routers, transports, and filters.
6870 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6871 strings in the file are not string-expanded.
6873 42. If a MAIL command specified a SIZE value that was too large to fit into an
6874 int variable, the check against message_size_limit failed. Such values are
6875 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6876 day this will have to be increased, but I don't think I want to be around
6877 when emails are that large.
6884 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6885 iconv() is not standard in FreeBSD.
6887 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6888 IPv6 enabled. The observed symptom was a segmentation fault on return from
6889 the function os_common_find_running_interfaces() in src/os.c.
6891 3. In the check_special_case() function in daemon.c I had used "errno" as an
6892 argument name, which causes warnings on some systems. This was basically a
6893 typo, since it was named "eno" in the comments!
6895 4. The code that waits for the clock to tick (at a resolution of some fraction
6896 of a second) so as to ensure message-id uniqueness was always waiting for
6897 at least one whole tick, when it could have waited for less. [This is
6898 almost certainly not relevant at current processor speeds, where it is
6899 unlikely to ever wait at all. But we try to future-proof.]
6901 5. The function that sleeps for a time interval that includes fractions of a
6902 second contained a race. It did not block SIGALRM between setting the
6903 timer, and suspending (a couple of lines later). If the interval was short
6904 and the sigsuspend() was delayed until after it had expired, the suspension
6905 never ended. On busy systems this could lead to processes getting stuck for
6908 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6909 process, before it forks any delivery processes. The open lookup caching
6910 mechanism meant that the open file or database connection was passed into
6911 the delivery process. The problem was that delivery processes always tidy
6912 up cached lookup data. This could cause a problem for the next delivery
6913 process started by the queue runner, because the external queue runner
6914 process does not know about the closure. So the next delivery process
6915 still has data in the lookup cache. In the case of a file lookup, there was
6916 no problem because closing a file descriptor in a subprocess doesn't affect
6917 the parent. However, if the lookup was caching a connection to a database,
6918 the connection was closed, and the second delivery process was likely to
6919 see errors such as "PGSQL: query failed: server closed the connection
6920 unexpectedly". The problem has been fixed by closing all cached lookups
6921 in a queue runner before running a delivery process.
6923 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6924 seem to have the "const" qualifier which it has on other OS. I've
6927 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6928 *of the same type* (client or server) with the same public name that an
6929 error should be diagnosed.
6931 9. When Exim looked up a host name for an IP address, but failed to find the
6932 original IP address when looking up the host name (a safety check), it
6933 output the message "<ip address> does not match any IP for NULL", which was
6934 confusing, to say the least. The bug was that the host name should have
6935 appeared instead of "NULL".
6937 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6938 user that is built into the binary, and the -C or -D options is used, root
6939 privilege is dropped before the configuration file is read. In addition,
6940 logging is switched to stderr instead of the normal log files. If the
6941 configuration then re-defines the Exim user, the unprivileged environment
6942 is probably not what is expected, so Exim logs a panic warning message (but
6945 However, if deliver_drop_privilege is set, the unprivileged state may well
6946 be exactly what is intended, so the warning has been cut out in that case,
6947 and Exim is allowed to try to write to its normal log files.
6953 1. smtp_return_error_details was not giving details for temporary sender
6954 or receiver verification errors.
6956 2. Diagnose a configuration error if two authenticators have the same public
6959 3. Exim used not to create the message log file for a message until the first
6960 delivery attempt. This could be confusing when incoming messages were held
6961 for policy or load reasons. The message log file is now created at the time
6962 the message is received, and an initial "Received" line is written to it.
6964 4. The automatically generated man page for command line options had a minor
6965 bug that caused no ill effects; however, a more serious problem was that
6966 the procedure for building the man page automatically didn't always
6967 operate. Consequently, release 4.20 contains an out-of-date version. This
6968 shouldn't happen again.
6970 5. When building Exim with embedded Perl support, the script that builds the
6971 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6972 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6974 6. The freeze_tell option was not being used for messages that were frozen on
6975 arrival, either by an ACL or by local_scan().
6977 7. Added the smtp_incomplete_transaction log selector.
6979 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6980 was accepting AUTH without a new EHLO.
6982 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6983 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6986 10. Logging of TCP/IP connections (when configured) now happens in the main
6987 daemon process instead of the child process, so that the TCP/IP connection
6988 count is more accurate (but it can never be perfect).
6990 11. The use of "drop" in a nested ACL was not being handled correctly in the
6991 outer ACL. Now, if condition failure induced by the nested "drop" causes
6992 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6993 or "require"), the connection is dropped.
6995 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6996 that yield "discard" can now be used with an "accept" or a "discard" verb,
6997 but an error is generated for any others (because I can't see a useful way
6998 to define what should happen).
7000 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7001 are now processed in the same way as lines in the Exim configuration file.
7002 In particular, continuation lines are supported.
7004 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7006 15. Added -ti meaning -t -i.
7008 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7009 domains, and warn by logging if others are found.
7011 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7012 returned from GIFCONF when Exim is trying to find the list of interfaces on
7013 a host. The code in os.c has been modified to copy each ifreq to an aligned
7014 structure in all cases.
7016 Also, in some cases, the returned ifreq's were being copied to a 'struct
7017 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7018 means the last couple of bytes of an IPv6 address could be chopped if the
7019 ifreq contained only a normal sockaddr (14 bytes storage).
7021 18. Named domain lists were not supported in the hosts_treat_as_local option.
7022 An entry such as +xxxx was not recognized, and was treated as a literal
7025 19. Ensure that header lines added by a DATA ACL are included in the reject log
7026 if the ACL subsequently rejects the message.
7028 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7029 MD5 (which is deprecated).
7031 21. When testing a filter file using -bf, Exim was writing a message when it
7032 took the sender from a "From " line in the message, but it was not doing so
7033 when it took $return_path from a Return-Path: header line. It now does.
7035 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7036 with a valid header line field name (a series of printing characters
7037 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7039 23. Changed "disc" in the source to "disk" to conform to the documentation and
7040 the book and for uniformity.
7042 24. Ignore Sendmail's -Ooption=value command line item.
7044 25. When execve() failed while trying to run a command in a pipe transport,
7045 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7046 could be confused with a return value of 69 from the command itself. This
7047 has been changed to 127, the value the shell returns if it is asked to run
7048 a non-existent command. The wording for the related log line suggests a
7049 non-existent command as the problem.
7051 26. If received_header_text expands to an empty string, do not add a Received:
7052 header line to the message. (Well, it adds a token one on the spool, but
7053 marks it "old" so that it doesn't get used or transmitted.)
7055 27. Installed eximstats 1.28 (addition of -nt option).
7057 28. There was no check for failure on the call to getsockname() in the daemon
7058 code. This can fail if there is a shortage of resources on the system, with
7059 ENOMEM, for example. A temporary error is now given on failure.
7061 29. Contrary to the C standard, it seems that in some environments, the
7062 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7063 program. Exim now does this explicitly; it affects the formatting of
7064 timestamps using strftime().
7066 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7067 complaints. I've now initialized all the variables, to avoid this.
7069 32. Header lines added by a system filter were not being "seen" during
7070 transport-time rewrites.
7072 33. The info_callback() function passed to OpenSSL is set up with type void
7073 (*)(SSL *, int, int), as described somewhere. However, when calling the
7074 function (actually a macro) that sets it up, the type void(*)() is
7075 expected. I've put in a cast to prevent warnings from picky compilers.
7077 34. If a DNS black list lookup found a CNAME record, but there were no A
7078 records associated with the domain it pointed at, Exim crashed.
7080 35. If a DNS black list lookup returned more than one A record, Exim ignored
7081 all but the first. It now scans all returned addresses if a particular IP
7082 value is being sought. In this situation, the contents of the
7083 $dnslist_value variable are a list of all the addresses, separated by a
7086 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7087 to accept a host name and all its aliases if the forward lookup for any of
7088 them yielded the IP address of the incoming connection. Now it accepts only
7089 those names whose forward lookup yields the correct IP address. Any other
7090 names are discarded. This closes a loophole whereby a rogue DNS
7091 administrator could create reverse DNS records to break through a
7092 wildcarded host restriction in an ACL.
7094 37. If a user filter or a system filter that ran in a subprocess used any of
7095 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7096 the wrong values were passed to the pipe command ($thisaddress had the
7097 value of $0, $0 had the value of $1, etc). This bug was introduced by
7098 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7100 38. Improved the line breaking for long SMTP error messages from ACLs.
7101 Previously, if there was no break point between 40 and 75 characters, Exim
7102 left the rest of the message alone. Two changes have been made: (a) I've
7103 reduced the minimum length to 35 characters; (b) if it can't find a break
7104 point between 35 and 75 characters, it looks ahead and uses the first one
7105 that it finds. This may give the occasional overlong line, but at least the
7106 remaining text gets split now.
7108 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7109 file descriptors might be low, and that setting 1000 would always raise it.
7110 It turns out that in some environments, the limit is already over 1000 and
7111 that lowering it causes trouble. So now Exim takes care not to decrease it.
7113 40. When delivering a message, the value of $return_path is set to $sender_
7114 address at the start of routing (routers may change the value). By an
7115 oversight, this default was not being set up when an address was tested by
7116 -bt or -bv, which affected the outcome if any router or filter referred to
7119 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7120 log only when "message" or "log_message" are set. However, if one of the
7121 conditions was an address verification, or a call to a nested ACL, the
7122 messages generated by the underlying test were being passed through. This
7123 no longer happens. The underlying message is available in $acl_verify_
7124 message for both "message" and "log_message" expansions, so it can be
7125 passed through if needed.
7127 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7128 new expansion $bh_ to give the encoded byte string without charset
7129 translation. Translation happens only if iconv() is available; HAVE_ICONV
7130 indicates this at build time. HEADERS_CHARSET gives the charset to
7131 translate to; headers_charset can change it in the configuration, and
7132 "headers charset" can change it in an individual filter file.
7134 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7135 that creates RFC 2047 encoded "words" labels them as that charset instead
7136 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7137 expansion operator; (ii) when Exim creates a From: line for a local
7138 message; (iii) when a header line is rewritten to include a "phrase" part.
7140 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7141 buggy, causing it to skip the first lines of messages whose message ID
7142 ended in 'D'. This would not have bitten before Exim release 4.14, because
7143 message IDs were unlikely to end in 'D' before then. The effect was to have
7144 incorrect size information for certain domains.
7146 45. #include "config.h" was missing at the start of the crypt16.c module. This
7147 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7150 46. If there was a timeout during a "random" callout check, Exim treated it as
7151 a failure of the random address, and carried on sending RSET and the real
7152 address. If the delay was just some slowness somewhere, the response to the
7153 original RCPT would be taken as a response to RSET and so on, causing
7154 mayhem of various kinds.
7156 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7157 when I implemented it. It didn't allow for the fact that some option values
7158 may legitimately be negative (e.g. size_addition), and it didn't even do
7159 the right test for positive values.
7161 48. Domain names in DNS records are case-independent. Exim always looks them up
7162 in lower case. Some resolvers return domain names in exactly the case they
7163 appear in the zone file, that is, they may contain uppercase letters. Not
7164 all resolvers do this - some return always lower case. Exim was treating a
7165 change of case by a resolver as a change of domain, similar to a widening
7166 of a domain abbreviation. This triggered its re-routing code and so it was
7167 trying to route what was effectively the same domain again. This normally
7168 caused routing to fail (because the router wouldn't handle the domain
7169 twice). Now Exim checks for this case specially, and just changes the
7170 casing of the domain that it ultimately uses when it transmits the message
7173 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7176 50. If a filter generated a file delivery with a non-absolute name (possible if
7177 no home directory exists for the router), the forbid_file option was not
7180 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7181 the existing equality matching.
7183 52. Exim was using ints instead of ino_t variables in some places where it was
7184 dealing with inode numbers.
7186 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7187 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7188 called TMPDIR, and if it finds it is different, it changes its value.
7190 54. The smtp_printf() function is now made available to local_scan() so
7191 additional output lines can be written before returning. There is also an
7192 smtp_fflush() function to enable the detection of a dropped connection.
7193 The variables smtp_input and smtp_batched_input are exported to
7196 55. Changed the default runtime configuration: the message "Unknown user"
7197 has been removed from the ACL, and instead placed on the localuser router,
7198 using the cannot_route_message feature. This means that any verification
7199 failures that generate their own messages won't get overridden. Similarly,
7200 the "Unrouteable address" message that was in the ACL for unverifiable
7201 relay addresses has also been removed.
7203 56. Added hosts_avoid_esmtp to the smtp transport.
7205 57. The exicyclog script was not checking for the esoteric option
7206 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7207 will work only if exicyclog is run under the appropriate euid.
7209 58. Following a discussion on the list, the rules by which Exim recognises line
7210 endings on incoming messages have been changed. The -dropcr and drop_cr
7211 options are now no-ops, retained only for backwards compatibility. The
7212 following line terminators are recognized: LF CRLF CR. However, special
7213 processing applies to CR:
7215 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7216 nor a local message in the state where . is a terminator.
7218 (ii) If a bare CR is encountered in a header line, an extra space is added
7219 after the line terminator so as not to end the header. The reasoning
7220 behind this is that bare CRs in header lines are most likely either
7221 to be mistakes, or people trying to play silly games.
7223 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7224 was being incorrectly given as 18 bytes larger than it should have been.
7225 This is a VOB (very old bug).
7227 60. This may never have affected anything current, but just in case it has:
7228 When the local host is found other than at the start of a list of hosts,
7229 the local host, those with the same MX, and any that follow, are discarded.
7230 When the list in question was part of a longer list of hosts, the following
7231 hosts (not currently being processed) were also being discarded. This no
7232 longer happens. I'm not sure if this situation could ever has previously
7235 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7238 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7239 header. This is something that is recommended in a new Internet Draft, and
7240 is something that is documented as being done by Sendmail. There are two
7241 possible values. For messages generated by the autoreply transport, Exim
7244 Auto-Submitted: auto-replied
7246 whereas for all other generated messages (e.g. bounces) it adds
7248 Auto-Submitted: auto-generated
7250 63. The "personal" condition in filters now includes a test for the
7251 Auto-Submitted: header. If it contains the string "auto-" the message it
7252 not considered personal.
7254 64. Added rcpt_include_affixes as a generic transport option.
7256 65. Added queue_only_override (default true).
7258 66. Added the syslog_duplication option.
7260 67. If what should have been the first header line of a message consisted of
7261 a space followed by a colon, Exim was mis-interpreting it as a header line.
7262 It isn't of course - it is syntactically invalid and should therefore be
7263 treated as the start of the message body. The misbehaviour could have
7264 caused a number of strange effects, including loss of data in subsequent
7265 header lines, and spool format errors.
7267 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7268 client host had authenticated. This control can now be exercised by an ACL
7269 for more flexibility.
7271 69. By default, callouts do not happen when testing with -bh. There is now a
7272 variant, -bhc, which does actually run the callout code, including
7273 consulting and updating the callout cache.
7275 70. Added support for saslauthd authentication, courtesy of Alexander
7278 71. If statvfs() failed on the spool or log directories while checking their
7279 size for availability, Exim confusingly gave the error "space shortage".
7280 Furthermore, in debugging mode it crashed with a floating point exception.
7281 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7282 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7283 serious problem, Exim now writes to the main and panic logs when this
7284 happens, with details of the failure. It then refuses to accept the
7285 incoming message, giving the message "spool directory problem" or "log
7286 directory problem" with a 421 code for SMTP messages.
7288 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7289 0, 1, and 2 exist, because some OS complain for execs without them (see
7290 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7291 descriptors. However, the code omitted to check that the open succeeded,
7292 causing mysterious errors if for some reason the permissions on /dev/null
7293 got screwed. Now Exim writes a message to the main and panic logs, and
7294 bombs out if it can't open /dev/null.
7296 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7297 interact so that it is all more flexible. It is supposed to remain
7298 backwards compatible. Also added extra_local_interfaces.
7300 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7301 to bomb out with an assertion failure - to the client this appears as a
7302 connection drop. This problem occurs in the part of the code that was taken
7303 from the Samba project. Fortunately, the assertion is in a very simple
7304 function, so I have fixed this by reproducing the function inline in the
7305 one place where it is called, and arranging for authentication to fail
7306 instead of killing the process with assert().
7308 75. The SPA client code was not working when the server requested OEM rather
7309 than Unicode encoding.
7311 76. Added code to make require_files with a specific uid setting more usable in
7312 the case where statting the file as root fails - usually a non-root-mounted
7313 NFS file system. When this happens and the failure is EACCES, Exim now
7314 forks a subprocess and does the per-uid checking as the relevant uid.
7316 77. Added process_log_path.
7318 78. If log_file_path was not explicitly set, a setting of check_log_space or
7319 check_log_inodes was ignored.
7321 79. If a space check for the spool or log partitions fails, the incident is now
7322 logged. Of course, in the latter case the data may get lost...
7324 80. Added the %p formatting code to string_format() so that it can be used to
7325 print addresses in debug_print(). Adjusted all the address printing in the
7326 debugging in store.c to use %p rather than %d.
7328 81. There was a concern that a line of code in smtp_in.c could overflow a
7329 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7330 initially expressed, the concern was not well-founded, because trailing
7331 spaces are removed early. However, if the trailing spaces were followed by
7332 a NULL, they did not get removed, so the overflow was possible. Two fixes
7335 (a) I re-wrote the offending code in a cleaner fashion.
7336 (b) If an incoming SMTP command contains a NULL character, it is rejected
7339 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7340 runs initgroups(), so that if the Exim user is in any additional groups,
7341 they will be used during message reception.
7347 The change log for 4.20 and earlier releases has been archived.