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/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917, OVE-20190718-0006)
17 JH/01 Remove code calling the customisable local_scan function, unless a new
18 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
20 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
21 non-signal-safe functions being used.
23 JH/03 Bug 2269: When presented with a received message having a stupidly large
24 number of DKIM-Signature headers, disable DKIM verification to avoid
25 a resource-consumption attack. The limit is set at twenty.
27 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
28 report of oldest_pass in ${authres } in consequence, and separate out
29 some descriptions of reasons for verification fail.
31 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
32 files in the spool were present and unlocked. A queue-runner could spot
33 them, resulting in a duplicate delivery. Fix that by doing the unlock
34 after the unlink. Investigation by Tim Stewart. Take the opportunity to
35 add more error-checking on spoolfile handling while that code is being
38 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
39 No known attacks, no CVE, this is defensive hardening.
41 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
42 a queue-runner could start a delivery while other operations were ongoing.
43 Cutthrough delivery was a common victim, resulting in duplicate delivery.
44 Found and investigated by Tim Stewart. Fix by using the open message data
45 file handle rather than opening another, and not locally closing it (which
46 releases a lock) for that case, while creating the temporary .eml format
47 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
49 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
50 $sender_verify_failure/$recipient_verify_failure to "random".
52 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
55 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
56 Previously this would segfault.
58 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
61 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
62 like zero, since the resolver should be doing this for us, But we need one
63 as a CNAME but no MX presence gets the CNAME returned; we need to check
64 that doesn't point to an MX to declare it "no MX returned" rather than
65 "error, loop". A new main option is added so the older capability of
66 following some limited number of chain links is maintained.
68 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
70 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
71 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
72 now-deprecated earlier definitions used only specified the range up to TLS
73 1.2 (in the older-version library docs).
75 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
77 JH/15 Rework TLS client-side context management. Stop using a global, and
78 explicitly pass a context around. This enables future use of TLS for
79 connections to service-daemons (eg. malware scanning) while a client smtp
80 connection is using TLS; with cutthrough connections this is quite likely.
82 JH/16 Fix ARC verification to do AS checks in reverse order.
84 JH/17 Support a "tls" option on the ${readsocket } expansion item.
86 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
87 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
88 Previously the "utf8" would be re-prepended for every additional message.
90 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
91 Previously thery were accepted, resulting in issues when attempting to
92 forward messages to a non-supporting MTA.
94 PP/02 Let -n work with printing macros too, not just options.
96 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
97 one parent address was copied, and bogus data was used at delivery-logging
98 time. Either a crash (after delivery) or bogus log data could result.
99 Discovery and analysis by Tim Stewart.
101 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
102 Previously if the string ended mid-character, we did not insert the
103 promised '?' replacement.
105 PP/04 Documentation: current string operators work on bytes, not codepoints.
107 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
108 should pack well giving a smaller memory footprint so better caching and
109 therefore performance. Group the declarations where this can't be done so
110 that the byte-sized flag variables are not interspersed among pointer
111 variables, giving a better chance of good packing by the compiler.
113 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
114 non-null, to avoid issues with sites running BATV. Previously reports were
115 sent with an empty envelope sender so looked like bounces.
117 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
118 The ignore_error flag wasn't being returned from the filter subprocess so
119 was not set for later routers. Investigation and fix by Matthias Kurz.
121 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
122 and a msg:complete for the whole, when a message is manually removed using
123 -Mrm. Developement by Matthias Kurz, hacked on by JH.
125 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
126 a "Gnu special" function, asprintf() in the DB utility binary builds; I
127 hope that is portable enough.
129 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
130 requiring a known-CA anchor certificate; make it now rely entirely on the
131 TLSA as an anchor. Checking the name on the leaf cert against the name
132 on the A-record for the host is still done for TA (but not for EE mode).
134 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
135 included in delivery lines for non-proxied connections, when compiled with
136 SUPPORT_SOCKS and running with proxy logging enabled.
138 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
139 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
140 move the existing event to fire before the normal logging of message
141 failure so that custom logging is bracketed by normal logging.
143 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
144 msg:fail:internal event. Developement by Matthias Kurz.
146 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
147 far too small for todays use of crypto signatures stored there. Go all
148 the way to the max DNS message size of 64kB, even though this might be
149 overmuch for IOT constrained device use.
151 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
152 copy a string over itself. The library routine is documented as not
153 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
155 JH/32 For main options check_spool_space and check_inode_space, where the
156 platform supports 64b integers, support more than the previous 2^31 kB
157 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
158 the previous G, M, k.
160 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
161 $authenticated_fail_id variable on authentication failure. Previously
164 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
165 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
166 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
167 GNUTLS_SEC_PARAM_MEDIUM.
169 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
170 side. Previously we would continue as if no SNI had been received.
172 JH/36 Harden the handling of string-lists. When a list consisted of a sole
173 "<" character, which should be a list-separator specification, we walked
174 off past the nul-terimation.
176 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
177 causes) even when the retry time is not yet met. Previously they were
178 not, meaning that when (say) an account was over-quota and temp-rejecting,
179 and multiple senders' messages were queued, only one sender would get
180 notified on each configured delay_warning cycle.
182 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
184 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
185 cipher-suites, an error can be left on the stack even for a succeeding
186 accept; this results in impossible error messages when a later operation
189 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
190 return error codes indicating retry. Under TLS1.3 this becomes required.
192 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
193 it only wrote the new authenticators, resulting in a lack of tracking of
194 peer changes of ESMTP extensions until the next cache flush.
196 JH/41 Fix the loop reading a message header line to check for integer overflow,
197 and more-often against header_maxsize. Previously a crafted message could
198 induce a crash of the recive process; now the message is cleanly rejected.
200 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
201 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
207 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
208 When redis_servers is set to a list of > 1 element, and the Redis servers
209 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
210 case of MOVED into a DEFER case instead, thus moving the query onto the
211 next server in the list. For a cluster of N elements, all N servers must
212 be defined in redis_servers.
214 GF/02 Catch and remove uninitialized value warning in exiqsumm
215 Check for existence of @ARGV before looking at $ARGV[0]
217 JH/01 Replace the store_release() internal interface with store_newblock(),
218 which internalises the check required to safely use the old one, plus
219 the allocate and data copy operations duplicated in both (!) of the
220 extant use locations.
222 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
223 modifier. This matches the restriction on the commandline.
225 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
226 Previously only the last row was returned.
228 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
229 we assumed that tags in the header were well-formed, and parsed the
230 element content after inspecting only the first char of the tag.
231 Assumptions at that stage could crash the receive process on malformed
234 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
235 While running the DKIM ACL we operate on the Permanent memory pool so that
236 variables created with "set" persist to the DATA ACL. Also (at any time)
237 DNS lookups that fail create cache records using the Permanent pool. But
238 expansions release any allocations made on the current pool - so a dnsdb
239 lookup expansion done in the DKIM ACL releases the memory used for the
240 DNS negative-cache, and bad things result. Solution is to switch to the
241 Main pool for expansions.
242 While we're in that code, add checks on the DNS cache during store_reset,
243 active in the testsuite.
244 Problem spotted, and debugging aided, by Wolfgang Breyha.
246 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
247 When none of the hosts presented to a transport match an already-open
248 connection, close it and proceed with the list. Previously we would
249 queue the message. Spotted by Lena with Yahoo, probably involving
252 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
253 Previously a spurious "250 OK id=" response was appended to the proper
256 JH/08 The "support for" informational output now, which built with Content
257 Scanning support, has a line for the malware scanner interfaces compiled
258 in. Interface can be individually included or not at build time.
260 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
261 by the template makefile "src/EDITME". The "STREAM" support for an older
262 ClamAV interface method is removed.
264 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
265 rows affected is given instead).
267 JH/11 The runtime Berkeley DB library version is now additionally output by
268 "exim -d -bV". Previously only the compile-time version was shown.
270 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
271 SMTP connection. Previously, when one had more receipients than the
272 first, an abortive onward connection was made. Move to full support for
273 multiple onward connections in sequence, handling cutthrough connection
274 for all multi-message initiating connections.
276 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
277 routers. Previously, a multi-recipient message would fail to match the
278 onward-connection opened for the first recipient, and cause its closure.
280 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
281 a timeout on read on a GnuTLS initiating connection, resulting in the
282 initiating connection being dropped. This mattered most when the callout
283 was marked defer_ok. Fix to keep the two timeout-detection methods
286 JH/15 Relax results from ACL control request to enable cutthrough, in
287 unsupported situations, from error to silently (except under debug)
288 ignoring. This covers use with PRDR, frozen messages, queue-only and
291 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
293 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
294 metadata, resulting in a crash in free().
296 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
297 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
298 Broken also in d185889f4, with init system revamp.
300 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
301 Previously we abruptly closed the connection after reading a malware-
302 found indication; now we go on to read the "scan ok" response line,
305 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
306 ACL. Previously, a crash would result.
308 JH/19 Speed up macro lookups during configuration file read, by skipping non-
309 macro text after a replacement (previously it was only once per line) and
310 by skipping builtin macros when searching for an uppercase lead character.
312 JH/20 DANE support moved from Experimental to mainline. The Makefile control
313 for the build is renamed.
315 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
316 was allocated for every new TLS startup, meaning one per message. Fix
317 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
319 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
320 reported the original. Fix to report (as far as possible) the ACL
321 result replacing the original.
323 JH/23 Fix memory leak during multi-message connections using STARTTLS under
324 OpenSSL. Certificate information is loaded for every new TLS startup,
325 and the resources needed to be freed.
327 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
329 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
332 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
333 DATA response info to the (existing) per-recipient response info for
334 the "C=" log element. It can have useful tracking info from the
335 destination system. Patch from Simon Arlott.
337 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
338 length value. Previously this would segfault.
340 HS/02 Support Avast multiline protoocol, this allows passing flags to
341 newer versions of the scanner.
343 JH/28 Ensure that variables possibly set during message acceptance are marked
344 dead before release of memory in the daemon loop. This stops complaints
345 about them when the debug_store option is enabled. Discovered specifically
346 for sender_rate_period, but applies to a whole set of variables.
347 Do the same for the queue-runner and queue-list loops, for variables set
348 from spool message files. Do the same for the SMTP per-message loop, for
349 certain variables indirectly set in ACL operations.
351 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
352 as a multi-recipient message from a mailinglist manager). The coding had
353 an arbitrary cutoff number of characters while checking for more input;
354 enforced by writing a NUL into the buffer. This corrupted long / fast
355 input. The problem was exposed more widely when more pipelineing of SMTP
356 responses was introduced, and one Exim system was feeding another.
357 The symptom is log complaints of SMTP syntax error (NUL chars) on the
358 receiving system, and refused recipients seen by the sending system
359 (propating to people being dropped from mailing lists).
360 Discovered and pinpointed by David Carter.
362 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
363 replaced by the ${authresults } expansion.
365 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
367 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
368 allows proper process termination in container environments.
370 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
371 Previously the "final dot" had a newline after it; ensure it is CR,LF.
373 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
374 and "err_perm", deprecated since 4.83 when the RFC-defined words
375 "temperror" and "permerror" were introduced.
377 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
378 transport-filters or DKIM-signing. The restriction was lost in the
379 consolidation of verify-callout and delivery SMTP handling.
380 Extend the restriction to also cover ARC-signing.
382 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
383 in defer=pass mode supply a 450 to the initiator. Previously the message
386 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
387 tls_require_ciphers is used as before.
389 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
390 "pass_unscanned". Only tmpfails from the scanner are written to
391 the paniclog, as they may require admin intervention (permission
392 denied, license issues). Other scanner errors (like decompression
393 bombs) do not cause a paniclog entry.
395 JH/36 Fix reinitialisation of DKIM logging variable between messages.
396 Previously it was possible to log spurious information in receive log
399 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
400 triggered odd behaviour from Outlook Express clients.
402 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
405 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
406 since the IETF WG has not yet settled on that versus the original
407 "bare" representation.
409 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
410 Previously the millisecond value corrupted the output.
411 Fix also for syslog_pid=no and log_selector +pid, for which the pid
412 corrupted the output.
418 JH/01 Rework error string handling in TLS interface so that the caller in
419 more cases is responsible for logging. This permits library-sourced
420 string to be attached to addresses during delivery, and collapses
421 pairs of long lines into single ones.
423 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
424 during configuration. Wildcards are allowed and expanded.
426 JH/02 Rework error string handling in DKIM to pass more info back to callers.
427 This permits better logging.
429 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
430 do not close it down and have the child transport start it up again on
431 the passed-on TCP connection. Instead, proxy the child (and any
432 subsequent ones) for TLS via a unix-domain socket channel. Logging is
433 affected: the continued delivery log lines do not have any DNSSEC, TLS
434 Certificate or OCSP information. TLS cipher information is still logged.
436 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
437 identical IP addresses on different listening ports. Will also affect
440 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
441 add noisy ifdef guards to special-case this sillyness.
442 Patch from Bernd Kuhls.
444 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
445 than 255 are no longer allowed.
447 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
448 Disable the session-cache too, which might reduce our load. Since we
449 currrectly use a new context for every connection, both as server and
450 client, there is no benefit for these.
451 GnuTLS appears to not support tickets server-side by default (we don't
452 call gnutls_session_ticket_enable_server()) but client side is enabled
453 by default on recent versions (3.1.3 +) unless the PFS priority string
456 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
457 <https://reproducible-builds.org/specs/source-date-epoch/>.
459 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
460 the check for any unsuccessful recipients did not notice the limit, and
461 erroneously found still-pending ones.
463 JH/08 Pipeline CHUNKING command and data together, on kernels that support
464 MSG_MORE. Only in-clear (not on TLS connections).
466 JH/09 Avoid using a temporary file during transport using dkim. Unless a
467 transport-filter is involved we can buffer the headers in memory for
468 creating the signature, and read the spool data file once for the
469 signature and again for transmission.
471 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
472 4.77 as the kernel support then wasn't solid, having issues in 64bit
473 mode. Now, it's been long enough. Add support for FreeBSD also.
475 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
476 case where the routing stage had gathered several addresses to send to
477 a host before calling the transport for the first, we previously failed
478 to close down TLS in the old transport process before passing the TCP
479 connection to the new process. The new one sent a STARTTLS command
480 which naturally failed, giving a failed delivery and bloating the retry
481 database. Investigation and fix prototype from Wolfgang Breyha.
483 JH/12 Fix check on SMTP command input synchronisation. Previously there were
484 false-negatives in the check that the sender had not preempted a response
485 or prompt from Exim (running as a server), due to that code's lack of
486 awareness of the SMTP input buffering.
488 PP/04 Add commandline_checks_require_admin option.
489 Exim drops privileges sanely, various checks such as -be aren't a
490 security problem, as long as you trust local users with access to their
491 own account. When invoked by services which pass untrusted data to
492 Exim, this might be an issue. Set this option in main configuration
493 AND make fixes to the calling application, such as using `--` to stop
496 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
497 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
498 responses to those, into a single TLS record each way (this usually means
499 a single packet). As a side issue, smtp_enforce_sync now works on TLS
502 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
503 affects you only if you're dancing at the edge of the param size limits.
504 If you are, and this message makes sense to you, then: raise the
505 configured limit or use OpenSSL 1.1. Nothing we can do for older
508 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
509 cmdline element to get the documented default one. Previously it was
512 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
513 are unsuable for non-message-specific reasons - eg. connection timeout,
514 banner-time rejection.
516 JH/16 Fix logging of delivery remote port, when specified by router, under
519 PP/06 Repair manualroute's ability to take options in any order, even if one
520 is the name of a transport.
523 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
525 JH/17 Change the list-building routines interface to use the expanding-string
526 triplet model, for better allocation and copying behaviour.
528 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
529 Previously it was constructed the first time a possibly-matching string
530 was met in the configuration file input during startup; now it is done
533 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
534 compatible one, to avoid the (poorly documented) possibility of a config
535 file in the working directory redirecting the DB files, possibly correpting
536 some existing file. CVE-2017-10140 assigned for BDB.
538 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
539 cache-hot. Previously, although the result was properly cached, the
540 initial verify call returned a defer.
542 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
543 the main verify for receipient in uncached-mode.
545 JH/22 Retire historical build files to an "unsupported" subdir. These are
546 defined as "ones for which we have no current evidence of testing".
548 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
549 if present. Previously it was ignored.
551 JH/24 Start using specified-initialisers in C structure init coding. This is
552 a C99 feature (it's 2017, so now considered safe).
554 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
555 if was a fixed-sized field and bitmask ops via macros; it is now more
558 PP/07 GitHub PR 56: Apply MariaDB build fix.
559 Patch provided by Jaroslav Škarvada.
561 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
562 during Coverity cleanups [4.87 JH/47]
563 Diagnosis and fix provided by Michael Fischer v. Mollard.
565 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
566 the right size to place the terminating semicolon on its own folded
567 line, the header hash was calculated to an incorrect value thanks to
568 the (relaxed) space the fold became.
570 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
571 and confused the parent.
573 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
574 which could crash as a result. This could lead to undeliverable messages.
576 JH/28 Logging: "next input sent too soon" now shows where input was truncated
579 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
580 matters on fast-turnover and PID-randomising systems, which were getting
581 out-of-order delivery.
583 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
584 a possibly-overlapping copy. The symptom was that "Remote host closed
585 connection in response to HELO" was logged instead of the actual 4xx
588 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
589 Previously only that bufferd was discarded, resulting in SYMTP command
592 JH/32 DKIM: when a message has multiple signatures matching an identity given
593 in dkim_verify_signers, run the dkim acl once for each. Previously only
594 one run was done. Bug 2189.
596 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
597 "panic the current process" to "deliberately defer". The panic log is
598 still written with the problem list name; the mail and reject logs now
599 get a temp-reject line for the message that was being handled, saying
600 something like "domains check lookup or other defer". The SMTP 451
601 message is still "Temporary local problem".
603 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
604 A crafted sequence of BDAT commands could result in in-use memory beeing
605 freed. CVE-2017-16943.
607 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
608 from SMTP input. Previously it was always done; now only done for DATA
609 and not BDAT commands. CVE-2017-16944.
611 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
612 to the message (such as an overlong header line). Previously this was
613 not done and we did not exit BDAT mode. Followon from the previous item
614 though a different problem.
620 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
621 than -2003 did; needs libidn2 in addition to libidn.
623 JH/02 The path option on a pipe transport is now expanded before use.
625 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
626 Patch provided by "Björn", documentation fix added too.
628 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
629 missing a wire-to-host endian conversion.
631 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
632 close after a BDAT command line could be taken as a following command,
633 giving a synch failure. Fix by only checking for synch immediately
634 before acknowledging the chunk.
636 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
637 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
638 macro. Patches provided by Josh Soref.
640 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
641 Previously we did not; the RFC seems ambiguous and VRFY is not listed
642 by IANA as a service extension. However, John Klensin suggests that we
645 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
646 the dkim code may be unix-mode line endings rather than smtp wire-format
647 CRLF, so prepend a CR to any bare LF.
649 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
650 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
652 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
653 the first were themselves being wrongly included in the feed into dkim
654 processing; with most chunk sizes in use this resulted in an incorrect
655 body hash calculated value.
657 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
658 DKIM signature block, for verification. Although advised against by
659 standards it is specifically not ruled illegal.
661 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
663 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
664 missing a body hash (the bh= tag).
666 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
667 It seems that HAProxy sends the Proxy Protocol information in clear and
668 only then does a TLS startup, so do the same.
670 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
671 TCP connections (such as for Spamd) unless the daemon successfully set
672 Fast Open mode on its listening sockets. This fixes breakage seen on
673 too-old kernels or those not configured for Fast Open, at the cost of
674 requiring both directions being enabled for TFO, and TFO never being used
675 by non-daemon-related Exim processes.
677 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
678 endings, at least on the first header line. Try to canonify any that get
679 past that check, despite the cost.
681 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
682 now limited to an arbitrary five deep, while parsing addresses with the
683 strip_excess_angle_brackets option enabled.
685 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
686 instead leave the unprompted TLS handshake in socket buffer for the
687 TLS library to consume.
689 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
691 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
693 JH/16 Drop variables when they go out of scope. Memory management drops a whole
694 region in one operation, for speed, and this leaves assigned pointers
695 dangling. Add checks run only under the testsuite which checks all
696 variables at a store-reset and panics on a dangling pointer; add code
697 explicitly nulling out all the variables discovered. Fixes one known
698 bug: a transport crash, where a dangling pointer for $sending_ip_address
699 originally assigned in a verify callout, is re-used.
701 PP/06 Drop '.' from @INC in various Perl scripts.
703 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
705 PP/08 Reduce a number of compilation warnings under clang; building with
706 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
707 should be warning-free.
709 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
711 HS/01 Fix portability problems introduced by PP/08 for platforms where
712 realloc(NULL) is not equivalent to malloc() [SunOS et al].
714 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
715 chunk. This allows us to accept broken chunked messages. We need a more
716 general solution here.
718 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
719 already-broken messages in the queue.
721 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
723 JH/19 Fix reference counting bug in routing-generated-address tracking.
729 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
730 supports it and a size is available (ie. the sending peer gave us one).
732 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
733 years of being deprecated). The replacements are the ACLs
734 acl_smtp_mime and acl_not_smtp_mime.
736 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
737 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
738 or even an in-clear connection were permitted. Now, if the host lookup
739 was dnssec and dane was requested then the host is only used if the
740 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
741 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
742 if one fails this test.
743 This means that a poorly-configured remote DNS will make it incommunicado;
744 but it protects against a DNS-interception attack on it.
746 JH/04 Bug 1810: make continued-use of an open smtp transport connection
747 non-noisy when a race steals the message being considered.
749 JH/05 If main configuration option tls_certificate is unset, generate a
750 self-signed certificate for inbound TLS connections.
752 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
753 in rewrites and routers.
755 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
756 and logged a warning sing 4.83; now they are a configuration file error.
758 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
759 (lacking @domain). Apply the same qualification processing as RCPT.
761 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
763 JH/10 Support ${sha256:} applied to a string (as well as the previous
766 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
767 a cutthrough deliver is pending, as we always want to make a connection.
768 This also avoids re-routing the message when later placing the cutthrough
769 connection after a verify cache hit.
770 Do not update it with the verify result either.
772 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
773 when routing results in more than one destination address.
775 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
776 signing (which inhibits the cutthrough capability). Previously only
777 the presence of an option was tested; now an expansion evaluating as
778 empty is permissible (obviously it should depend only on data available
779 when the cutthrough connection is made).
781 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
782 the relevant preceding SMTP command did not note the pipelining mode.
784 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
785 Previously they were not counted.
787 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
788 as one having no matching records. Previously we deferred the message
789 that needed the lookup.
791 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
792 distinguished as "(=".
794 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
795 for missing MX records. Previously it only worked for missing A records.
797 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
799 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
800 after the data-go-ahead and data-ack. Patch from Jason Betts.
802 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
803 even for a "none" policy. Patch from Tony Meyer.
805 JH/22 Fix continued use of a connection for further deliveries. If a port was
806 specified by a router, it must also match for the delivery to be
809 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
810 When one of the recipients of a message was unsuitable for the connection
811 (has no matching addresses), we lost track of needing to mark it
812 deferred. As a result mail would be lost.
814 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
816 JH/25 Decoding ACL controls is now done using a binary search; the source code
817 takes up less space and should be simpler to maintain. Merge the ACL
818 condition decode tables also, with similar effect.
820 JH/26 Fix problem with one_time used on a redirect router which returned the
821 parent address unchanged. A retry would see the parent address marked as
822 delivered, so not attempt the (identical) child. As a result mail would
825 JH/27 Fix a possible security hole, wherein a process operating with the Exim
826 UID can gain a root shell. Credit to http://www.halfdog.net/ for
827 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
830 JH/28 Enable {spool,log} filesystem space and inode checks as default.
831 Main config options check_{log,spool}_{inodes,space} are now
832 100 inodes, 10MB unless set otherwise in the configuration.
834 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
835 Previously it only applied to the main-section connection policy
838 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
840 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
841 by me. Added RFC7919 DH primes as an alternative.
843 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
844 are not in the system include path.
846 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
847 GnuTLS, when a session startup failed (eg because the client disconnected)
848 Exim did stdio operations after fclose. This was exposed by a recent
849 change which nulled out the file handle after the fclose.
851 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
852 signed directly by the cert-signing cert, rather than an intermediate
853 OCSP-signing cert. This is the model used by LetsEncrypt.
855 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
857 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
858 an incoming connection.
860 HS/02 Bug 1802: Do not half-close the connection after sending a request
863 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
864 fallback to "prime256v1".
866 JH/34 SECURITY: Use proper copy of DATA command in error message.
867 Could leak key material. Remotely exploitable. CVE-2016-9963.
873 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
874 and 3.4.4 - once the server is enabled to respond to an OCSP request
875 it does even when not requested, resulting in a stapling non-aware
876 client dropping the TLS connection.
878 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
879 support variable-length bit vectors. No functional change.
881 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
882 The I= interface field on outgoing lines is now after the H= remote
883 host field, same as incoming lines. There is a separate
884 outgoing_interface log selector which allows you to disable the
887 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
888 If not running log_selector +smtp_connection the mainlog would be held
889 open indefinitely after a "too many connections" event, including to a
890 deleted file after a log rotate. Leave the per net connection logging
891 leaving it open for efficiency as that will be quickly detected by the
892 check on the next write.
894 HS/01 Bug 1671: Fix post transport crash.
895 Processing the wait-<transport> messages could crash the delivery
896 process if the message IDs didn't exist for some reason. When
897 using 'split_spool_directory=yes' the construction of the spool
898 file name failed already, exposing the same netto behaviour.
900 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
901 mime_regex ACL conditions.
903 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
904 to DSN fail messages (bounces): remote IP, remote greeting, remote response
905 to HELO, local diagnostic string.
907 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
908 log line to debug. Even when configured with a tls authenticator many
909 client connections are expected to not authenticate in this way, so
910 an authenticate fail is not an error.
912 HS/02 Add the Exim version string to the process info. This way exiwhat
913 gives some more detail about the running daemon.
915 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
916 matter for fast-change records such as DNSBLs.
918 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
919 retry record, even if constant. There may be multiple transports with
920 different interface settings and the retry behaviour needs to be kept
923 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
925 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
927 JH/10 Bug 840: fix log_defer_output option of pipe transport
929 JH/11 Bug 830: use same host for all RCPTS of a message, even under
930 hosts_randomize. This matters a lot when combined with mua_wrapper.
932 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
933 ${quote_pgsql:<string>} operator.
935 JH/13 Bug 1708: avoid misaligned access in cached lookup.
937 JH/14 Change header file name for freeradius-client. Relevant if compiling
938 with Radius support; from the Gentoo tree and checked under Fedora.
940 JH/15 Bug 1712: Introduce $prdr_requested flag variable
942 JH/16 Bug 1714: Permit an empty string as expansion result for transport
943 option transport_filter, meaning no filtering.
945 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
947 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
948 defaults to "*" (all hosts). The variable is now available when not built
949 with TLS, default unset, mainly to enable keeping the testsuite sane.
950 If a server certificate is not supplied (via tls_certificate) an error is
951 logged, and clients will find TLS connections fail on startup. Presumably
952 they will retry in-clear.
953 Packagers of Exim are strongly encouraged to create a server certificate
954 at installation time.
956 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
957 with the $config_file variable.
959 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
960 in transport context, after the attempt, and per-recipient. The latter type
961 is per host attempted. The event data is the error message, and the errno
962 information encodes the lookup type (A vs. MX) used for the (first) host,
963 and the trailing two digits of the smtp 4xx response.
965 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
966 to write to mainlog (or rejectlog, paniclog) in the window between file
967 creation and permissions/ownership being changed. Particularly affects
968 installations where exicyclog is run as root, rather than exim user;
969 result is that the running daemon panics and dies.
971 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
973 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
974 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
975 "pri" and "weight". Note that the previous implicit priority given by the
976 list order is no longer honoured.
978 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
981 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
982 by defining SUPPORT_SOCKS.
984 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
985 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
986 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
987 variables are renamed to proxy_{local,external}_{address,port}.
989 JH/27 Move Internationalisation support from Experimental to mainline, enabled
990 for a build by defining SUPPORT_I18N
992 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
993 of the query string, and make ${quote_redis:} do that quoting.
995 JH/29 Move Events support from Experimental to mainline, enabled by default
996 and removable for a build by defining DISABLE_EVENT.
998 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1000 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1001 cached by the daemon.
1003 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1004 by defining LOOKUP_REDIS. The libhiredis library is required.
1006 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1007 keys are given for lookup.
1009 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1010 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1011 only supported when built with TLS support. The PolarSSL SHA routines
1012 are still used when the TLS library is too old for convenient support.
1014 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1015 openssl_options), for security. OpenSSL forces this from version 1.1.0
1016 server-side so match that on older versions.
1018 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1019 allocation for $value could be released as the expansion processing
1020 concluded, but leaving the global pointer active for it.
1022 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1023 and to use the domains and local_parts ACL conditions.
1025 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1026 incorrectly not doubled on cutthrough transmission, hence seen as a
1027 body-termination at the receiving system - resulting in truncated mails.
1028 Commonly the sender saw a TCP-level error, and retransmitted the message
1029 via the normal store-and-forward channel. This could result in duplicates
1030 received - but deduplicating mailstores were liable to retain only the
1031 initial truncated version.
1033 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1035 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1037 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1038 we're in there, support oversigning also; bug 1309.
1040 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1042 HS/04 Add support for keep_environment and add_environment options.
1044 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1045 either intentional arithmetic overflow during PRNG, or testing config-
1048 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1049 delivery resulted in actual delivery. Cancel cutthrough before DATA
1052 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1053 rejects a recipient: pass the reject to the originator.
1055 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1056 Many were false-positives and ignorable, but it's worth fixing the
1059 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1060 for the new environment-manipulation done at startup. Move the routines
1061 from being local to tls.c to being global via the os.c file.
1063 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1064 an extract embedded as result-arg for a map, the first arg for extract
1065 is unavailable so we cannot tell if this is a numbered or keyed
1066 extraction. Accept either.
1072 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1075 JH/02 The smtp transport option "multi_domain" is now expanded.
1077 JH/03 The smtp transport now requests PRDR by default, if the server offers
1080 JH/04 Certificate name checking on server certificates, when exim is a client,
1081 is now done by default. The transport option tls_verify_cert_hostnames
1082 can be used to disable this per-host. The build option
1083 EXPERIMENTAL_CERTNAMES is withdrawn.
1085 JH/05 The value of the tls_verify_certificates smtp transport and main options
1086 default to the word "system" to access the system default CA bundle.
1087 For GnuTLS, only version 3.0.20 or later.
1089 JH/06 Verification of the server certificate for a TLS connection is now tried
1090 (but not required) by default. The verification status is now logged by
1091 default, for both outbound TLS and client-certificate supplying inbound
1094 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1097 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1098 Status Notification (bounce) messages are now MIME format per RFC 3464.
1099 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1100 under the control of the dsn_advertise_hosts option, and routers may
1101 have a dsn_lasthop option.
1103 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1104 default, modifiable by a malware= option. The list separator for
1105 the options can now be changed in the usual way. Bug 68.
1107 JH/10 The smtp_receive_timeout main option is now expanded before use.
1109 JH/11 The incoming_interface log option now also enables logging of the
1110 local interface on delivery outgoing connections.
1112 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1113 if the interface and destination host and port all match.
1115 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1118 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1119 Patch from Andrew Lewis.
1121 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1122 now supports optional time-restrictions, weighting, and priority
1123 modifiers per server. Patch originally by <rommer@active.by>.
1125 JH/16 The spamd_address main option now supports a mixed list of local
1126 and remote servers. Remote servers can be IPv6 addresses, and
1127 specify a port-range.
1129 JH/17 Bug 68: The spamd_address main option now supports an optional
1130 timeout value per server.
1132 JH/18 Bug 1581: Router and transport options headers_add/remove can
1133 now have the list separator specified.
1135 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1138 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1141 JH/21 Support for the A6 type of dns record is withdrawn.
1143 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1144 rather than the verbs used.
1146 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1147 from 255 to 1024 chars.
1149 JH/24 Verification callouts now attempt to use TLS by default.
1151 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1152 are generic router options now. The defaults didn't change.
1154 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1155 Original patch from Alexander Shikoff, worked over by JH.
1157 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1158 files if ZCAT_COMMAND is not executable.
1160 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1162 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1164 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1165 Normally benign, it bites when the pair was led to by a CNAME;
1166 modern usage is to not canonicalize the domain to a CNAME target
1167 (and we were inconsistent anyway for A-only vs AAAA+A).
1169 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1171 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1172 when evaluating $sender_host_dnssec.
1174 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1175 $sender_helo_dnssec variable.
1177 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1179 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1181 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1183 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1184 documented as working, but never had. Support all but $spam_report.
1186 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1187 added for tls authenticator.
1189 HS/03 Add perl_taintmode main config option
1195 TL/01 When running the test suite, the README says that variables such as
1196 no_msglog_check are global and can be placed anywhere in a specific
1197 test's script, however it was observed that placement needed to be near
1198 the beginning for it to behave that way. Changed the runtest perl
1199 script to read through the entire script once to detect and set these
1200 variables, reset to the beginning of the script, and then run through
1201 the script parsing/test process like normal.
1203 TL/02 The BSD's have an arc4random API. One of the functions to induce
1204 adding randomness was arc4random_stir(), but it has been removed in
1205 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1206 function when detected.
1208 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1209 cause callback expansion.
1211 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1212 syntax errors in an expansion can be treated as a string instead of
1213 logging or causing an error, due to the internal use of bool_lax
1214 instead of bool when processing it.
1216 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1217 server certificates when making smtp deliveries.
1219 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1221 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1223 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1225 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1226 Merged patch from Sebastian Wiedenroth.
1228 JH/05 Fix results-pipe from transport process. Several recipients, combined
1229 with certificate use, exposed issues where response data items split
1230 over buffer boundaries were not parsed properly. This eventually
1231 resulted in duplicates being sent. This issue only became common enough
1232 to notice due to the introduction of connection certificate information,
1233 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1235 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1236 size buffer was used, resulting in syntax errors when an expansion
1239 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1240 version 3.3.6 or later.
1242 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1243 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1244 both become "event_action", the variables become $event_name, $event_data
1245 and $event_defer_errno. There is a new variable $verify_mode, usable in
1246 routers, transports and related events. The tls:cert event is now also
1247 raised for inbound connections, if the main configuration event_action
1250 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1251 early OCSP support, but no stapling (appears to be less than 1.0.0).
1253 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1254 server certificate names available under the smtp transport option
1255 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1258 JH/10 Time-related extraction expansions from certificates now use the main
1259 option "timezone" setting for output formatting, and are consistent
1260 between OpenSSL and GnuTLS compilations. Bug 1541.
1262 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1263 encoded parameter in the incoming message. Bug 1558.
1265 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1266 include certificate info, eximon was claiming there were spoolfile
1269 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1271 JH/14 Log delivery-related information more consistently, using the sequence
1272 "H=<name> [<ip>]" wherever possible.
1274 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1275 are problematic for Debian distribution, omit them from the release
1278 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1280 JH/16 Fix string representation of time values on 64bit time_t architectures.
1283 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1284 output list separator was used.
1289 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1290 checkers that were complaining about end of non-void function with no
1293 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1294 This was a regression introduced in 4.83 by another bugfix.
1296 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1298 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1299 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1305 TF/01 Correctly close the server side of TLS when forking for delivery.
1307 When a message was received over SMTP with TLS, Exim failed to clear up
1308 the incoming connection properly after forking off the child process to
1309 deliver the message. In some situations the subsequent outgoing
1310 delivery connection happened to have the same fd number as the incoming
1311 connection previously had. Exim would try to use TLS and fail, logging
1312 a "Bad file descriptor" error.
1314 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1315 utilities have not been installed.
1317 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1318 temporary space as the ACL may create new global variables.
1320 TL/01 LDAP support uses per connection or global context settings, depending
1321 upon the detected version of the libraries at build time.
1323 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1324 to extract and use the src ip:port in logging and expansions as if it
1325 were a direct connection from the outside internet. PPv2 support was
1326 updated based on HAProxy spec change in May 2014.
1328 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1330 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1331 Properly escape header and check for NULL return.
1333 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1336 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1338 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1339 characters in header names, implemented as a verify condition.
1340 Contributed by Michael Fischer v. Mollard.
1342 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1343 results permerror and temperror. Previous values are deprecated but
1344 still accepted. In a future release, err_perm and err_temp will be
1345 completely removed, which will be a backward incompatibility if the
1346 ACL tests for either of these two old results. Patch contributed by
1347 user bes-internal on the mailing list.
1349 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1351 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1352 selectors, in both main and reject logs.
1354 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1357 JH/07 Add malware type "sock" for talking to simple daemon.
1359 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1361 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1362 routers/transports under cutthrough routing.
1364 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1365 numbers. Touch up "bool" conditional to keep the same definition.
1367 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1369 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1371 JH/12 Expand items in router/transport headers_add or headers_remove lists
1372 individually rather than the list as a whole. Bug 1452.
1374 Required for reasonable handling of multiple headers_ options when
1375 they may be empty; requires that headers_remove items with embedded
1376 colons must have them doubled (or the list-separator changed).
1378 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1379 view the policy declared in the DMARC record. Currently, $dmarc_status
1380 is a combined value of both the record presence and the result of the
1383 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1385 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1386 dnslookup router and the smtp transport (applying to the forward
1389 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1390 of ldap servers used for a specific lookup. Patch provided by Heiko
1393 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1394 New variable $lookup_dnssec_authenticated for observability.
1396 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1397 Patch submitted by Lars Timman.
1399 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1401 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1402 Requires trusted mode and valid format message id, aborts otherwise.
1403 Patch contributed by Heiko Schlichting.
1405 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1406 certextract with support for various fields. Bug 1358.
1408 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1409 is requested by default, modifiable by smtp transport option
1412 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1413 operate on certificate variables to give certificate fingerprints
1414 Also new ${sha256:cert_variable}.
1416 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1418 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1419 Christian Aistleitner.
1421 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1423 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1424 file. Patch from Wolfgang Breyha.
1426 JH/25 Expand the coverage of the delivery $host and $host_address to
1427 client authenticators run in verify callout. Bug 1476.
1429 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1430 align with daemon_smtp_ports. Bug 72.
1432 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1433 support and error reporting did not work properly.
1435 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1436 and is readable. Patch from Andrew Colin Kissa.
1438 TL/14 Enhance documentation of ${run expansion and how it parses the
1439 commandline after expansion, particularly in the case when an
1440 unquoted variable expansion results in an empty value.
1442 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1444 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1445 by renaming away from it. Addresses GH issue 15, reported by
1448 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1450 TL/15 SECURITY: prevent double expansion in math comparison functions
1451 (can expand unsanitized data). Not remotely exploitable.
1458 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1460 PP/02 Make -n do something, by making it not do something.
1461 When combined with -bP, the name of an option is not output.
1463 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1466 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1467 $sender_host_name and config options to manage this, and basic check
1470 PP/05 DSCP support for outbound connections and control modifier for inbound.
1472 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1473 (Only plugin which currently uses this is kerberos4, which nobody should
1474 be using, but we should make it available and other future plugins might
1475 conceivably use it, even though it would break NAT; stuff *should* be
1476 using channel bindings instead).
1478 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1479 name; added for Sendmail compatibility; requires admin caller.
1480 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1481 just ignore it); requires trusted caller.
1482 Also parse but ignore: -Ac -Am -X<logfile>
1485 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1487 TL/02 Add +smtp_confirmation as a default logging option.
1489 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1490 Patch by Magnus Holmgren from 2007-02-20.
1492 TL/04 Bugzilla 1281 - Spec typo.
1493 Bugzilla 1283 - Spec typo.
1494 Bugzilla 1290 - Spec grammar fixes.
1496 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1498 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1500 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1501 mailing list by by Dmitry Isaikin.
1503 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1505 JH/02 Support "G" suffix to numbers in ${if comparisons.
1507 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1509 NM/01 Bugzilla 1197 - Spec typo
1510 Bugzilla 1196 - Spec examples corrections
1512 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1514 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1515 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1518 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1519 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1521 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1522 "acl {{name}{arg}...}", and optional args on acl condition
1525 JH/05 Permit multiple router/transport headers_add/remove lines.
1527 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1529 JH/07 Avoid using a waiting database for a single-message-only transport.
1530 Performance patch from Paul Fisher. Bugzilla 1262.
1532 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1535 JH/09 Add $headers_added variable, with content from use of ACL modifier
1536 add_header (but not yet added to the message). Bugzilla 199.
1538 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1539 Pulled from Bugzilla 817 by Wolfgang Breyha.
1541 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1543 (nb: this is the same fix as in Exim 4.80.1)
1545 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1548 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1551 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1553 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1554 advertise SMTP AUTH mechanism to us, instead of a generic
1555 protocol violation error. Also, make Exim more robust to bad
1556 data from the Dovecot auth socket.
1558 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1560 When a queue runner is handling a message, Exim first routes the
1561 recipient addresses, during which it prunes them based on the retry
1562 hints database. After that it attempts to deliver the message to
1563 any remaining recipients. It then updates the hints database using
1566 So if a recipient address works intermittently, it can get repeatedly
1567 deferred at routing time. The retry hints record remains fresh so the
1568 address never reaches the final cutoff time.
1570 This is a fairly common occurrence when a user is bumping up against
1571 their storage quota. Exim had some logic in its local delivery code
1572 to deal with this. However it did not apply to per-recipient defers
1573 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1575 This change adds a proper retry rule check during routing so that the
1576 final cutoff time is checked against the message's age. We only do
1577 this check if there is an address retry record and there is not a
1578 domain retry record; this implies that previous attempts to handle
1579 the address had the retry_use_local_parts option turned on. We use
1580 this as an approximation for the destination being like a local
1581 delivery, as in LMTP.
1583 I suspect this new check makes the old local delivery cutoff check
1584 redundant, but I have not verified this so I left the code in place.
1586 TF/02 Correct gecos expansion when From: is a prefix of the username.
1588 Test 0254 submits a message to Exim with the header
1592 When I ran the test suite under the user fanf2, Exim expanded
1593 the header to contain my full name, whereas it should have added
1594 a Resent-Sender: header. It erroneously treats any prefix of the
1595 username as equal to the username.
1597 This change corrects that bug.
1599 GF/01 DCC debug and logging tidyup
1600 Error conditions log to paniclog rather than rejectlog.
1601 Debug lines prefixed by "DCC: " to remove any ambiguity.
1603 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1605 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1606 Bug spotted by Jeremy Harris; was flawed since initial commit.
1607 Would have resulted in OCSP responses post-SNI triggering an Exim
1608 NULL dereference and crash.
1610 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1612 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1613 Bug detection, analysis and fix by Samuel Thibault.
1614 Bugzilla 1331, Debian bug #698092.
1616 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1618 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1619 Server implementation by Todd Lyons, client by JH.
1620 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1621 config variable "prdr_enable" controls whether the server
1622 advertises the facility. If the client requests PRDR a new
1623 acl_data_smtp_prdr ACL is called once for each recipient, after
1624 the body content is received and before the acl_smtp_data ACL.
1625 The client is controlled by both of: a hosts_try_prdr option
1626 on the smtp transport, and the server advertisement.
1627 Default client logging of deliveries and rejections involving
1628 PRDR are flagged with the string "PRDR".
1630 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1631 fclose(). Diagnosis by Todd Lyons.
1633 PP/17 Update configure.default to handle IPv6 localhost better.
1634 Patch by Alain Williams (plus minor tweaks).
1637 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1638 This is now consistent with GnuTLS, and is now documented: the
1639 previous undocumented portable approach to treating the option as
1640 unset was to force an expansion failure. That still works, and
1641 an empty string is now equivalent.
1643 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1644 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1645 not performing validation itself.
1647 PP/20 Added force_command boolean option to pipe transport.
1648 Patch from Nick Koston, of cPanel Inc.
1650 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1653 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1655 PP/21 Fix eximon continuous updating with timestamped log-files.
1656 Broken in a format-string cleanup in 4.80, missed when I repaired the
1657 other false fix of the same issue.
1658 Report and fix from Heiko Schlichting.
1661 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1662 Report from Prashanth Katuri.
1664 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1665 It's SecureTransport, so affects any MacOS clients which use the
1666 system-integrated TLS libraries, including email clients.
1668 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1669 using a MIME ACL for non-SMTP local injection.
1670 Report and assistance in diagnosis by Warren Baker.
1672 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1674 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1676 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1677 last id that failed so it may be referenced in subsequent ACL's.
1679 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1682 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1683 ldap library initialization, allowing self-signed CA's to be
1684 used. Also properly sets require_cert option later in code by
1685 using NULL (global ldap config) instead of ldap handle (per
1686 session). Bug diagnosis and testing by alxgomz.
1688 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1689 the src/util/ subdirectory.
1691 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1692 renamed to Transport Post Delivery Action by Jeremy Harris, as
1695 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1696 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1697 redis_servers = needs to be configured which will be used by the redis
1698 lookup. Patch from Warren Baker, of The Packet Hub.
1700 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1702 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1703 hostname or reverse DNS when processing a host list. Used suggestions
1704 from multiple comments on this bug.
1706 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1708 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1709 Missed a few lines, added it to make the runtest require no keyboard
1712 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1713 contains upper case chars. Make router use caseful_local_part.
1715 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1716 support when GnuTLS has been built with p11-kit.
1722 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1724 This, or similar/improved, will also be change PP/11 of 4.82.
1730 PP/01 Handle short writes when writing local log-files.
1731 In practice, only affects FreeBSD (8 onwards).
1732 Bugzilla 1053, with thanks to Dmitry Isaikin.
1734 NM/01 Bugzilla 949 - Documentation tweak
1736 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1739 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1741 PP/02 Implemented gsasl authenticator.
1743 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1745 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1746 `pkg-config foo` for cflags/libs.
1748 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1749 with rest of GSASL and with heimdal_gssapi.
1751 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1752 `pkg-config foo` for cflags/libs for the TLS implementation.
1754 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1755 properties get this fed in as external SSF. A number of robustness
1756 and debugging improvements to the cyrus_sasl authenticator.
1758 PP/08 cyrus_sasl server now expands the server_realm option.
1760 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1761 Patch by Jeremy Harris.
1763 PP/10 Added dbmjz lookup type.
1765 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1767 PP/12 MAIL args handles TAB as well as SP, for better interop with
1768 non-compliant senders.
1769 Analysis and variant patch by Todd Lyons.
1771 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1772 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1773 Patch from Dirk Mueller <dmueller@suse.com>
1775 PP/13 tls_peerdn now print-escaped for spool files.
1776 Observed some $tls_peerdn in wild which contained \n, which resulted
1777 in spool file corruption.
1779 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1780 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1781 or write after TLS renegotiation, which otherwise led to messages
1784 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1785 as a tracking header (ie: a signed header comes before the signature).
1786 Patch from Wolfgang Breyha.
1788 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1789 comma-sep list; embedded commas doubled.
1791 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1793 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1795 Report and patch from Dmitry Banschikov.
1797 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1798 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1799 failures. We appear to now support TLS1.1+ with Exim.
1801 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1802 lets Exim select keys and certificates based upon TLS SNI from client.
1803 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1804 before an outbound SMTP session. New log_selector, +tls_sni.
1806 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1807 NULL dereference. Report and patch from Alun Jones.
1809 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1810 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1811 Not seeing resolver debug output on NetBSD, but suspect this is a
1812 resolver implementation change.
1814 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1815 Left warnings. Added "eximon gdb" invocation mode.
1817 PP/21 Defaulting "accept_8bitmime" to true, not false.
1819 PP/22 Added -bw for inetd wait mode support.
1821 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1822 locate the relevant includes and libraries. Made this the default.
1824 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1825 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1827 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1828 This may cause build issues on older platforms.
1830 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1831 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1832 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1833 Added SNI support via GnuTLS too.
1834 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1836 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1838 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1839 Applied second patch from Janne, implementing suggestion to default
1840 multiple-strings-in-record handling to match SPF spec.
1842 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1844 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1845 read-only, out of scope).
1846 Patch from Wolfgang Breyha, report from Stuart Northfield.
1848 PP/29 Fix three issues highlighted by clang analyser static analysis.
1849 Only crash-plausible issue would require the Cambridge-specific
1850 iplookup router and a misconfiguration.
1851 Report from Marcin Mirosław.
1853 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1855 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1856 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1857 As part of this, removing so much warning spew let me fix some minor
1858 real issues in debug logging.
1860 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1861 assignment on my part. Fixed.
1863 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1864 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1865 Janne Snabb (who went above and beyond: thank you).
1867 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1868 string otherwise requires a connection and a bunch more work and it's
1869 relatively easy to get wrong. Should also expose TLS library linkage
1872 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1873 64-bit ${eval} (JH/03).
1875 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1876 GNU libc to support some of the 64-bit stuff, should not lead to
1877 conflicts. Defined before os.h is pulled in, so if a given platform
1878 needs to override this, it can.
1880 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1881 protection layer was required, which is not implemented.
1882 Bugzilla 1254, patch from Wolfgang Breyha.
1884 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1885 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1886 tls_dhparam take prime identifiers. Also unbreak combination of
1887 OpenSSL+DH_params+TLSSNI.
1889 PP/39 Disable SSLv2 by default in OpenSSL support.
1895 PP/01 Solaris build fix for Oracle's LDAP libraries.
1896 Bugzilla 1109, patch from Stephen Usher.
1898 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1900 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1903 TF/02 Fix a couple more cases where we did not log the error message
1904 when unlink() failed. See also change 4.74-TF/03.
1906 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1907 lock up or crash if it happened to be inside a call to libc when it
1908 got a SIGUSR1 from exiwhat.
1910 The SIGUSR1 handler appends the current process status to the process
1911 log which is later printed by exiwhat. It used to use the general
1912 purpose logging code to do this, but several functions it calls are
1913 not safe for signals.
1915 The new output code in the SIGUSR1 handler is specific to the process
1916 log, and simple enough that it's easy to inspect for signal safety.
1917 Removing some special cases also simplifies the general logging code.
1918 Removing the spurious timestamps from the process log simplifies
1921 TF/04 Improved ratelimit ACL condition.
1923 The /noupdate option has been deprecated in favour of /readonly which
1924 has clearer semantics. The /leaky, /strict, and /readonly update modes
1925 are mutually exclusive. The update mode is no longer included in the
1926 database key; it just determines when the database is updated. (This
1927 means that when you upgrade Exim will forget old rate measurements.)
1929 Exim now checks that the per_* options are used with an update mode that
1930 makes sense for the current ACL. For example, when Exim is processing a
1931 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1932 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1933 must specify per_mail/readonly. If you omit the update mode it defaults to
1934 /leaky where that makes sense (as before) or /readonly where required.
1936 The /noupdate option is now undocumented but still supported for
1937 backwards compatibility. It is equivalent to /readonly except that in
1938 ACLs where /readonly is required you may specify /leaky/noupdate or
1939 /strict/noupdate which are treated the same as /readonly.
1941 A useful new feature is the /count= option. This is a generalization
1942 of the per_byte option, so that you can measure the throughput of other
1943 aggregate values. For example, the per_byte option is now equivalent
1944 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1946 The per_rcpt option has been generalized using the /count= mechanism
1947 (though it's more complicated than the per_byte equivalence). When it is
1948 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1949 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1950 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1951 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1952 Note that using per_rcpt with a non-readonly update mode in more than
1953 one ACL will cause the recipients to be double-counted. (The per_mail
1954 and per_byte options don't have this problem.)
1956 The handling of very low rates has changed slightly. If the computed rate
1957 is less than the event's count (usually one) then this event is the first
1958 after a long gap. In this case the rate is set to the same as this event's
1959 count, so that the first message of a spam run is counted properly.
1961 The major new feature is a mechanism for counting the rate of unique
1962 events. The new per_addr option counts the number of different
1963 recipients that someone has sent messages to in the last time period. It
1964 behaves like per_rcpt if all the recipient addresses are different, but
1965 duplicate recipient addresses do not increase the measured rate. Like
1966 the /count= option this is a general mechanism, so the per_addr option
1967 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1968 example, measure the rate that a client uses different sender addresses
1969 with the options per_mail/unique=$sender_address. There are further
1970 details in the main documentation.
1972 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1974 TF/06 Removed a few PCRE remnants.
1976 TF/07 Automatically extract Exim's version number from tags in the git
1977 repository when doing development or release builds.
1979 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1980 Bugzilla 879. Patch from Paul Fisher.
1982 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1983 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1986 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1988 PP/05 Variable $av_failed, true if the AV scanner deferred.
1989 Bugzilla 1078. Patch from John Horne.
1991 PP/06 Stop make process more reliably on build failure.
1992 Bugzilla 1087. Patch from Heiko Schlittermann.
1994 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1995 Bugzilla 1089. Patch from Heiko Schlittermann.
1997 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1998 Bugzilla 1131. Patch from Holger Weiß.
2000 PP/09 Handle IPv6 addresses with SPF.
2001 Bugzilla 860. Patch from Wolfgang Breyha.
2003 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2005 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2008 PP/11 match_* no longer expand right-hand-side by default.
2009 New compile-time build option, EXPAND_LISTMATCH_RHS.
2010 New expansion conditions, "inlist", "inlisti".
2012 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2014 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2016 PP/14 fix log_write() format string regression from TF/03.
2017 Bugzilla 1152. Patch from Dmitry Isaikin.
2023 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2025 PP/02 Harmonised TLS library version reporting; only show if debugging.
2026 Layout now matches that introduced for other libraries in 4.74 PP/03.
2028 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2030 PP/04 New "dns_use_edns0" global option.
2032 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2035 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2036 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2038 TK/01 Updated PolarSSL code to 0.14.2.
2039 Bugzilla 1097. Patch from Andreas Metzler.
2041 PP/07 Catch divide-by-zero in ${eval:...}.
2042 Fixes bugzilla 1102.
2044 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2047 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2048 format-string attack -- SECURITY: remote arbitrary code execution.
2050 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2051 time unintentionally subject to list matching rules, letting the header
2052 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2053 arbitrary string expansion). This allowed for information disclosure.
2055 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2056 INT_MIN/-1 -- value coerced to INT_MAX.
2062 NM/01 Workaround for PCRE version dependency in version reporting
2065 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2066 This fixes portability to compilers other than gcc, notably
2067 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2069 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2070 makefiles for portability to HP-UX and POSIX correctness.
2072 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2073 Also via indented variable definition in the Makefile.
2074 (Debugging by Oliver Heesakkers).
2076 PP/02 Restore caching of spamd results with expanded spamd_address.
2077 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2079 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2080 Improves build reliability. Fix from: Frank Elsner
2082 NM/02 Fix wide character breakage in the rfc2047 coding
2083 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2085 NM/03 Allow underscore in dnslist lookups
2086 Fixes bug 1026. Patch from Graeme Fowler
2088 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2089 Code patches from Adam Ciarcinski of NetBSD.
2091 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2094 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2095 is logged, to avoid truncation. Patch from John Horne.
2097 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2098 Patch from Jakob Hirsch.
2100 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2101 SQL string expansion failure details.
2102 Patch from Andrey Oktyabrski.
2104 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2105 Patch from Simon Arlott.
2107 PP/09 New lookups functionality failed to compile on old gcc which rejects
2108 extern declarations in function scope.
2109 Patch from Oliver Fleischmann
2111 PP/10 Use sig_atomic_t for flags set from signal handlers.
2112 Check getgroups() return and improve debugging.
2113 Fixed developed for diagnosis in bug 927 (which turned out to be
2116 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2117 Patch from Mark Zealey.
2119 PP/12 Bugzilla 1056: Improved spamd server selection.
2120 Patch from Mark Zealey.
2122 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2123 Based on patch from Heiko Schlittermann.
2125 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2126 Patch from Uwe Doering, sign-off by Michael Haardt.
2128 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2129 variable declaration deep within a block. Bug and patch from
2132 PP/15 lookups-Makefile IRIX compatibility coercion.
2134 PP/16 Make DISABLE_DKIM build knob functional.
2136 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2137 Patch by Simon Arlott
2139 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2140 variable argument macros. Our copy now differs from upstream.
2146 TF/01 Failure to get a lock on a hints database can have serious
2147 consequences so log it to the panic log.
2149 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2150 controlled using the smtp_confirmation log selector.
2152 TF/03 Include the error message when we fail to unlink a spool file.
2154 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2155 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2156 for maintaining out-of-tree patches for some time.
2158 PP/01 Bugzilla 139: Documentation and portability issues.
2159 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2160 Handle per-OS dynamic-module compilation flags.
2162 PP/02 Let /dev/null have normal permissions.
2163 The 4.73 fixes were a little too stringent and complained about the
2164 permissions on /dev/null. Exempt it from some checks.
2165 Reported by Andreas M. Kirchwitz.
2167 PP/03 Report version information for many libraries, including
2168 Exim version information for dynamically loaded libraries. Created
2169 version.h, now support a version extension string for distributors
2170 who patch heavily. Dynamic module ABI change.
2172 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2173 privilege escalation vulnerability whereby the Exim run-time user
2174 can cause root to append content of the attacker's choosing to
2177 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2180 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2181 If dropping privileges for untrusted macros, we disabled normal logging
2182 on the basis that it would fail; for the Exim run-time user, this is not
2183 the case, and it resulted in successful deliveries going unlogged.
2184 Fixed. Reported by Andreas Metzler.
2190 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2191 only prepend for the Resent-* case. Fixes regression introduced in
2192 Exim 4.70 by NM/22 for Bugzilla 607.
2194 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2195 increasing numbers of administrators be bitten by this.
2197 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2199 PP/03 Bugzilla 994: added openssl_options main configuration option.
2201 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2203 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2205 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2207 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2208 without a peer certificate, leading to a segfault because of an
2209 assumption that peers always have certificates. Be a little more
2210 paranoid. Problem reported by Martin Tscholak.
2212 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2213 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2214 NB: ClamAV planning to remove STREAM in "middle of 2010".
2215 CL also introduces -bmalware, various -d+acl logging additions and
2216 more caution in buffer sizes.
2218 PP/09 Implemented reverse_ip expansion operator.
2220 PP/10 Bugzilla 937: provide a "debug" ACL control.
2222 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2224 PP/12 Bugzilla 973: Implement --version.
2226 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2228 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2230 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2232 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2233 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2234 ignore trailing whitespace.
2236 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2238 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2241 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2242 Notification from Dr Andrew Aitchison.
2244 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2245 ExtendedDetectionInfo response format.
2246 Notification from John Horne.
2248 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2251 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2252 XSL and documented dependency on system catalogs, with examples of how
2255 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2258 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2259 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2260 configuration file which is writeable by the Exim user or group.
2262 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2263 of configuration files to cover files specified with the -C option if
2264 they are going to be used with root privileges, not just the default
2267 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2268 option (effectively making it always true).
2270 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2271 files to be used while preserving root privileges.
2273 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2274 that rogue child processes cannot use them.
2276 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2277 run-time user, instead of root.
2279 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2280 Exim run-time user without dropping privileges.
2282 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2283 result string, instead of calling string_vformat() twice with the same
2286 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2287 for other users. Others should always drop root privileges if they use
2288 -C on the command line, even for a whitelisted configure file.
2290 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2292 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2298 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2299 $data_path, and $header_path variables; fixed documentation bugs and
2302 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2303 exipick to access non-standard spools, including the "frozen" queue
2306 NM/01 Bugzilla 965: Support mysql stored procedures.
2307 Patch from Alain Williams
2309 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2311 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2312 Patch from Andreas Metzler
2314 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2315 Patch from Kirill Miazine
2317 NM/05 Bugzilla 671: Added umask to procmail example.
2319 JJ/03 installed exipick 20100323.0, fixing doc bug
2321 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2322 directory. Notification and patch from Dan Rosenberg.
2324 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2326 TK/02 Improve log output when DKIM signing operation fails.
2328 MH/01 Treat the transport option dkim_domain as a colon separated
2329 list, not as a single string, and sign the message with each element,
2330 omitting multiple occurences of the same signer.
2332 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2333 Bugzilla 985, 986. Patch by Simon Arlott
2335 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2336 Patch by Simon Arlott
2338 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2339 MBX locking. Notification from Dan Rosenberg.
2345 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2347 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2349 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2351 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2353 NM/04 Bugzilla 915: Flag broken perl installation during build.
2359 TK/01 Added patch by Johannes Berg that expands the main option
2360 "spamd_address" if it starts with a dollar sign.
2362 TK/02 Write list of recipients to X-Envelope-Sender header when building
2363 the mbox-format spool file for content scanning (suggested by Jakob
2366 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2367 (http://www.dcc-servers.net/) support via dccifd. Activated by
2368 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2370 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2371 by Mark Daniel Reidel <mr@df.eu>.
2373 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2374 When building exim an external PCRE library is now needed -
2375 PCRE is a system library on the majority of modern systems.
2376 See entry on PCRE_LIBS in EDITME file.
2378 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2379 conversation. Added nologin parameter to request.
2380 Patch contributed by Kirill Miazine.
2382 TF/01 Do not log submission mode rewrites if they do not change the address.
2384 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2386 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2387 log files in place. Contributed by Roberto Lima.
2389 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2391 TF/03 Bugzilla 615: When checking the local_parts router precondition
2392 after a local_part_suffix or local_part_prefix option, Exim now
2393 does not use the address's named list lookup cache, since this
2394 contains cached lookups for the whole local part.
2396 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2397 Robert Millan. Documentation is in experimental-spec.txt.
2399 TF/04 Bugzilla 668: Fix parallel build (make -j).
2401 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2403 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2404 Patch provided by Jan Srzednicki.
2406 TF/05 Leading white space used to be stripped from $spam_report which
2407 wrecked the formatting. Now it is preserved.
2409 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2410 that they are available at delivery time.
2412 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2414 TF/08 TLS error reporting now respects the incoming_interface and
2415 incoming_port log selectors.
2417 TF/09 Produce a more useful error message if an SMTP transport's hosts
2418 setting expands to an empty string.
2420 NM/06 Bugzilla 744: EXPN did not work under TLS.
2421 Patch provided by Phil Pennock.
2423 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2424 Patch provided by Richard Godbee.
2426 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2427 acl_smtp_notquit, added index entry.
2429 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2430 Patch provided by Eugene Bujak.
2432 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2433 accept(). Patch provided by Maxim Dounin.
2435 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2436 Patch provided by Phil Pennock.
2438 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2440 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2441 Patch provided by Brad "anomie" Jorsch.
2443 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2444 Patch provided by Dean Brooks.
2446 TK/05 Add native DKIM support (does not depend on external libraries).
2448 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2449 Patch provided by Graeme Fowler.
2451 NM/16 Bugzilla 851: Documentation example syntax fix.
2453 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2455 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2458 NM/19 Bugzilla 745: TLS version reporting.
2459 Patch provided by Phil Pennock.
2461 NM/20 Bugzilla 167: bool: condition support.
2462 Patch provided by Phil Pennock.
2464 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2465 clients. Patch provided by Phil Pennock.
2467 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2468 Patch provided by Brad "anomie" Jorsch.
2470 NM/23 Bugzilla 687: Fix misparses in eximstats.
2471 Patch provided by Heiko Schlittermann.
2473 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2474 Patch provided by Heiko Schlittermann.
2476 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2477 plus update to original patch.
2479 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2481 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2482 Patch provided by David Brownlee.
2484 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2486 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2488 NM/30 Bugzilla 888: TLS documentation bugfixes.
2490 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2492 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2493 Unlike the original bugzilla I have changed all shell scripts in src tree.
2495 NM/33 Bugzilla 898: Transport filter timeout fix.
2496 Patch by Todd Rinaldo.
2498 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2499 Patch by Serge Demonchaux.
2501 NM/35 Bugzilla 39: Base64 decode bug fixes.
2502 Patch by Jakob Hirsch.
2504 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2506 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2508 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2510 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2516 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2517 ALT-N's libdkim that I have put here:
2518 http://duncanthrax.net/exim-experimental/
2520 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2521 were called 'true' and it seems that C99 defines that as a reserved
2522 keyword to be used with 'bool' variable types. That means you could
2523 not include C99-style headers which use bools without triggering
2524 build errors in sieve.c.
2526 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2527 as mailq or other aliases. Changed the --help handling significantly
2528 to do whats expected. exim_usage() emits usage/help information.
2530 SC/01 Added the -bylocaldomain option to eximstats.
2532 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2534 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2536 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2542 PH/01 Another patch from the Sieve maintainer.
2544 PH/02 When an IPv6 address is converted to a string for single-key lookup
2545 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2546 dots are used instead of colons so that keys in lsearch files need not
2547 contain colons. This was done some time before quoting was made available
2548 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2549 (notated using the quote facility) so as to distinguish them from IPv4
2550 keys. This meant that lookups for IP addresses in host lists did not work
2551 for iplsearch lookups.
2553 This has been fixed by arranging for IPv6 addresses to be expressed with
2554 colons if the lookup type is iplsearch. This is not incompatible, because
2555 previously such lookups could never work.
2557 The situation is now rather anomalous, since one *can* have colons in
2558 ordinary lsearch keys. However, making the change in all cases is
2559 incompatible and would probably break a number of configurations.
2561 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2564 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2565 conversion specification without a maximum field width, thereby enabling
2566 a rogue spamd server to cause a buffer overflow. While nobody in their
2567 right mind would setup Exim to query an untrusted spamd server, an
2568 attacker that gains access to a server running spamd could potentially
2569 exploit this vulnerability to run arbitrary code as the Exim user.
2571 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2572 $primary_hostname instead of what libspf2 thinks the hosts name is.
2574 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2575 a directory entry by the name of the lookup key. Previously, if a
2576 symlink pointed to a non-existing file or a file in a directory that
2577 Exim lacked permissions to read, a lookup for a key matching that
2578 symlink would fail. Now it is enough that a matching directory entry
2579 exists, symlink or not. (Bugzilla 503.)
2581 PH/03 The body_linecount and body_zerocount variables are now exported in the
2584 PH/04 Added the $dnslist_matched variable.
2586 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2587 This means they are set thereafter only if the connection becomes
2590 PH/06 Added the client_condition to authenticators so that some can be skipped
2591 by clients under certain conditions.
2593 PH/07 The error message for a badly-placed control=no_multiline_responses left
2594 "_responses" off the end of the name.
2596 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2598 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2599 (without spaces) instead of just copying the configuration text.
2601 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2603 PH/11 Added $max_received_linelength.
2605 PH/12 Added +ignore_defer and +include_defer to host lists.
2607 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2608 way in which PCRE > 7.0 is built.
2610 PH/14 Implemented queue_only_load_latch.
2612 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2613 MAIL command. The effect was to mangle the value on 64-bit systems.
2615 PH/16 Another patch from the Sieve maintainer.
2617 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2619 PH/18 If a system quota error occurred while trying to create the file for
2620 a maildir delivery, the message "Mailbox is full" was not appended to the
2621 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2622 only to a quota excession during the actual writing of the file.
2624 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2625 characters?) which causes problems in log lines. The DN values are now
2626 passed through string_printing() before being added to log lines.
2628 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2629 and InterBase are left for another time.)
2631 PH/21 Added message_body_newlines option.
2633 PH/22 Guard against possible overflow in moan_check_errorcopy().
2635 PH/23 POSIX allows open() to be a macro; guard against that.
2637 PH/24 If the recipient of an error message contained an @ in the local part
2638 (suitably quoted, of course), incorrect values were put in $domain and
2639 $local_part during the evaluation of errors_copy.
2645 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2646 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2649 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2650 issue a MAIL command.
2652 PH/02 In an ACL statement such as
2654 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2656 if a client was not listed at all, or was listed with a value other than
2657 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2658 the condition was not true (as it should be), so access was not denied.
2659 The bug was that the ! inversion was incorrectly passed on to the second
2660 item. This has been fixed.
2662 PH/03 Added additional dnslists conditions == and =& which are different from
2663 = and & when the dns lookup returns more than one IP address.
2665 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2666 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2668 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2669 FSYNC, which compiles an option called disable_fsync that allows for
2670 bypassing fsync(). The documentation is heavily laced with warnings.
2672 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2674 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2675 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2676 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2677 including adding "make clean"; (3) Added -fPIC when compiling the test
2678 dynamically loaded module, to get rid of a warning.
2680 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2681 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2682 = 0s. The bug is otherwise harmless.
2684 PH/07 There was a bug in the dovecot authenticator such that the value of
2685 $auth1 could be overwritten, and so not correctly preserved, after a
2686 successful authentication. This usually meant that the value preserved by
2687 the server_setid option was incorrect.
2689 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2691 PH/09 Installed PCRE release 7.0.
2693 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2694 run for batched SMTP input. It is now run at the start of every message
2695 in the batch. While fixing this I discovered that the process information
2696 (output by running exiwhat) was not always getting set for -bs and -bS
2697 input. This is fixed, and it now also says "batched" for BSMTP.
2699 PH/11 Added control=no_pipelining.
2701 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2702 patch, slightly modified), and move the expansion of helo_data till after
2703 the connection is made in the smtp transport (so it can use these
2706 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2708 PH/14 Added log_selector = +pid.
2710 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2712 PH/16 Add ${if forany and ${if forall.
2714 PH/17 Added dsn_from option to vary the From: line in DSNs.
2716 PH/18 Flush SMTP output before performing a callout, unless control =
2717 no_callout_flush is set.
2719 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2720 was true (the default) a successful delivery failed to delete the retry
2721 item, thus causing premature timeout of the address. The bug is now
2724 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2726 PH/21 Long custom messages for fakedefer and fakereject are now split up
2727 into multiline responses in the same way that messages for "deny" and
2728 other ACL rejections are.
2730 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2731 with slight modification.
2733 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2734 draft, changing the syntax and factoring some duplicate code".
2736 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2737 for deliveries of the second and subsequent messages over the same SMTP
2740 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2741 ${reduce, with only minor "tidies".
2743 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2745 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2746 expansion side effects.
2748 PH/27 When a message times out after an over-quota error from an Exim-imposed
2749 quota, the bounce message says "mailbox is full". This message was not
2750 being given when it was a system quota that was exceeded. It now should
2753 MH/03 Made $recipients available in local_scan(). local_scan() already has
2754 better access to the recipient list through recipients_list[], but
2755 $recipients can be useful in postmaster-provided expansion strings.
2757 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2758 in the case of a MAIL command with additional options following the
2759 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2760 were accidentally chopped off.
2762 PH/29 SMTP synchronization checks are implemented when a command is read -
2763 there is a check that no more input is waiting when there shouldn't be
2764 any. However, for some commands, a delay in an ACL can mean that it is
2765 some time before the response is written. In this time, more input might
2766 arrive, invalidly. So now there are extra checks after an ACL has run for
2767 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2768 pipelining has not been advertised.
2770 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2772 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2773 returned in the SMTP response when smtp_return_error_details was false.
2774 This has been fixed.
2776 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2777 instead of the C I/O that was originally supplied, because problems were
2778 reported on Solaris.
2780 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2781 Exim which did not show up earlier: it was assuming that a call to
2782 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2783 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2784 was a macro that became an assignment, so it seemed to work. This has
2785 changed to a proper function call with a void return, hence the compile
2786 error. Exim's code has been fixed.
2788 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2791 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2793 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2795 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2796 because it turns out to be too big or there is a timeout) there is no
2797 "Completed" line in the log. When some messages of this type were
2798 selected by exigrep, they were listed as "not completed". Others were
2799 picked up by some special patterns. I have improved the selection
2800 criteria to be more general.
2802 PH/38 The host_find_failed option in the manualroute router can now be set
2803 to "ignore", to completely ignore a host whose IP address cannot be
2804 found. If all hosts are ignored, the behaviour is controlled by the new
2805 host_all_ignored option.
2807 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2808 homing or because of multiple MX records with /mx) generated more than
2809 one IP address, and the following item turned out to be the local host,
2810 all the secondary addresses of the first item were incorrectly removed
2811 from the list, along with the local host and any following hosts (which
2812 is what is supposed to happen).
2814 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2815 whoever called Exim into the -H file. In the case of the daemon it was
2816 behaving confusingly. When first started, it used values for whoever
2817 started the daemon, but after a SIGHUP it used the Exim user (because it
2818 calls itself on a restart). I have changed the code so that it now always
2821 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2822 message are rejected with the same error (e.g. no authentication or bad
2823 sender address), and a DATA command is nevertheless sent (as can happen
2824 with PIPELINING or a stupid MUA), the error message that was given to the
2825 RCPT commands is included in the rejection of the DATA command. This is
2826 intended to be helpful for MUAs that show only the final error to their
2829 PH/42 Another patch from the Sieve maintainer.
2831 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2832 Eximstats - Fixed some broken HTML links and added missing column headers
2834 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2835 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2837 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2839 PH/43 Yet another patch from the Sieve maintainer.
2841 PH/44 I found a way to check for a TCP/IP connection going away before sending
2842 the response to the final '.' that terminates a message, but only in the
2843 case where the client has not sent further data following the '.'
2844 (unfortunately, this is allowed). However, in many cases there won't be
2845 any further data because there won't be any more messages to send. A call
2846 to select() can be used: if it shows that the input is "ready", there is
2847 either input waiting, or the socket has been closed. An attempt to read
2848 the next input character can distinguish the two cases. Previously, Exim
2849 would have sent an OK response which the client would never have see.
2850 This could lead to message repetition. This fix should cure that, at
2851 least in a lot of common cases.
2853 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2854 advertised in response to EHLO.
2860 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2861 fixed by 4.65/MH/01 (is this a record?) are fixed:
2863 (i) An empty string was always treated as zero by the numeric comparison
2864 operators. This behaviour has been restored.
2866 (ii) It is documented that the numeric comparison operators always treat
2867 their arguments as decimal numbers. This was broken in that numbers
2868 starting with 0 were being interpreted as octal.
2870 While fixing these problems I realized that there was another issue that
2871 hadn't been noticed. Values of message_size_limit (both the global option
2872 and the transport option) were treated as octal if they started with 0.
2873 The documentation was vague. These values are now always treated as
2874 decimal, and I will make that clear in the documentation.
2880 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2881 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2884 MH/01 Don't check that the operands of numeric comparison operators are
2885 integers when their expansion is in "skipping" mode (fixes bug
2886 introduced by 4.64-PH/07).
2888 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2889 child addresses, Exim now panics and dies. Previously, because the count
2890 is held in a short int, deliveries were likely to be lost. As such a
2891 large number of recipients for a single message is ridiculous
2892 (performance will be very, very poor), I have chosen to impose a limit
2893 rather than extend the field.
2899 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2900 leftover -K file (the existence of which was triggered by #402).
2901 While we were at it, introduced process PID as part of the -K
2902 filename. This should rule out race conditions when creating
2905 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2906 processing considerably. Previous code took too long for large mails,
2907 triggering a timeout which in turn triggers #401.
2909 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2910 in the DK code in transports.c. sendfile() is not really portable,
2911 hence the _LINUX specificness.
2913 TF/01 In the add_headers option to the mail command in an Exim filter,
2914 there was a bug that Exim would claim a syntax error in any
2915 header after the first one which had an odd number of characters
2918 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2919 callout verification, Exim cached a "reject" for the entire domain. This
2920 is correct for most verifications, but it is not correct for a recipient
2921 verification with use_sender or use_postmaster set, because in that case
2922 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2923 case of MAIL FROM:<> rejection from other early rejections (e.g.
2924 rejection of HELO). When verifying a recipient using a non-null MAIL
2925 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2926 Whatever the result of the callout, the value of the domain cache is
2927 left unchanged (for any other kind of callout, getting as far as trying
2928 RCPT means that the domain itself is ok).
2930 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2933 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2934 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2935 session, the daemon ignores ECONNECT errors and logs others; it now
2936 ignores EPIPE as well.
2938 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2939 (quoted-printable decoding).
2941 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2942 later the small subsequent patch to fix an introduced bug.
2944 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2946 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2948 PH/08 An error is now given if message_size_limit is specified negative.
2950 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2951 to be given (somewhat) arbitrary names.
2953 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2956 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2957 miscellaneous code fixes
2959 PH/10 Added the log_reject_target ACL modifier to specify where to log
2962 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2963 hostname. This is wrong, because it relates to the incoming message (and
2964 probably the interface on which it is arriving) and not to the outgoing
2965 callout (which could be using a different interface). This has been
2966 changed to use the value of the helo_data option from the smtp transport
2967 instead - this is what is used when a message is actually being sent. If
2968 there is no remote transport (possible with a router that sets up host
2969 addresses), $smtp_active_hostname is used.
2971 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2972 tweaks were necessary in order to get it to work (see also 21 below):
2973 (a) The code assumed that strncpy() returns a negative number on buffer
2974 overflow, which isn't the case. Replaced with Exim's string_format()
2976 (b) There were several signed/unsigned issues. I just did the minimum
2977 hacking in of casts. There is scope for a larger refactoring.
2978 (c) The code used strcasecmp() which is not a standard C function.
2979 Replaced with Exim's strcmpic() function.
2980 (d) The code set only $1; it now sets $auth1 as well.
2981 (e) A simple test gave the error "authentication client didn't specify
2982 service in request". It would seem that Dovecot has changed its
2983 interface. Fortunately there's a specification; I followed it and
2984 changed what the client sends and it appears to be working now.
2986 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2989 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2990 address A is aliased to B and C, where B exists and C does not. Without
2991 -v the output is "A verified" because verification stops after a
2992 successful redirection if more than one address is generated. However,
2993 with -v the child addresses are also verified. Exim was outputting "A
2994 failed to verify" and then showing the successful verification for C,
2995 with its parentage. It now outputs "B failed to verify", showing B's
2996 parentage before showing the successful verification of C.
2998 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2999 look up a TXT record in a specific list after matching in a combined
3002 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3003 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3004 they consult the DNS. I had assumed they would set it the way they
3005 wanted; and indeed my experiments on Linux seem to show that in some
3006 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3007 To be on the safe side, however, I have now made the interface to
3008 host_find_byname() similar to host_find_bydns(), with an argument
3009 containing the DNS resolver options. The host_find_byname() function now
3010 sets these options at its start, just as host_find_bydns() does. The smtp
3011 transport options dns_qualify_single and dns_search_parents are passed to
3012 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3013 of host_find_byname() use the default settings of RES_DEFNAMES
3014 (qualify_single) but not RES_DNSRCH (search_parents).
3016 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3017 spool_read_header() do less string testing, by means of a preliminary
3018 switch on the second character of optional "-foo" lines. (This is
3019 overdue, caused by the large number of possibilities that now exist.
3020 Originally there were few.) While I was there, I also converted the
3021 str(n)cmp tests so they don't re-test the leading "-" and the first
3022 character, in the hope this might squeeze out yet more improvement.
3024 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3025 flag allowing group syntax was set by the header_syntax check but not
3026 turned off, possible causing trouble later; (2) The flag was not being
3027 set at all for the header_verify test, causing "group"-style headers to
3028 be rejected. I have now set it in this case, and also caused header_
3029 verify to ignore an empty address taken from a group. While doing this, I
3030 came across some other cases where the code for allowing group syntax
3031 while scanning a header line wasn't quite right (mostly, not resetting
3032 the flag correctly in the right place). These bugs could have caused
3033 trouble for malformed header lines. I hope it is now all correct.
3035 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3036 with the "reply" argument non-NULL. The code, however (which originally
3037 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3038 but it didn't always do it. This confused somebody who was copying the
3039 code for some other use. I have removed all the tests.
3041 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3042 feature that was used to support insecure browsers during the U.S. crypto
3043 embargo. It requires special client support, and Exim is probably the
3044 only MTA that supported it -- and would never use it because real RSA is
3045 always available. This code has been removed, because it had the bad
3046 effect of slowing Exim down by computing (never used) parameters for the
3047 RSA_EXPORT functionality.
3049 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3050 authenticator to fail if there's a tab character in the incoming data
3051 (there should never be unless someone is messing about, as it's supposed
3052 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3053 if the connection is using TLS or if the remote IP is the same as the
3054 local IP, and the "valid-client-cert option" if a client certificate has
3057 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3058 authenticators. This can be used for authorization after authentication
3059 succeeds. (In the case of plaintext, it servers for both authentication
3062 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3063 if any retry times were supplied.
3065 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3066 connection, where there is no client IP address to check. In this
3067 situation, the verify now always succeeds.
3069 PH/25 Applied John Jetmore's -Mset patch.
3071 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3073 PH/27 In a string expansion for a processed (not raw) header when multiple
3074 headers of the same name were present, leading whitespace was being
3075 removed from all of them, but trailing whitespace was being removed only
3076 from the last one. Now trailing whitespace is removed from each header
3077 before concatenation. Completely empty headers in a concatenation (as
3078 before) are ignored.
3080 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3081 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3083 PH/29 [Removed. This was a change that I later backed out, and forgot to
3084 correct the ChangeLog entry (that I had efficiently created) before
3085 committing the later change.]
3087 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3088 address errors (4xx response to RCPT) over the same connection as other
3089 messages routed to the same hosts. Such deliveries are always "forced",
3090 so retry times are not inspected. This resulted in far too many retries
3091 for the affected addresses. The effect occurred only when there were more
3092 hosts than the hosts_max_try setting in the smtp transport when it had
3093 the 4xx errors. Those hosts that it had tried were not added to the list
3094 of hosts for which the message was waiting, so if all were tried, there
3095 was no problem. Two fixes have been applied:
3097 (i) If there are any address or message errors in an SMTP delivery, none
3098 of the hosts (tried or untried) are now added to the list of hosts
3099 for which the message is waiting, so the message should not be a
3100 candidate for sending over the same connection that was used for a
3101 successful delivery of some other message. This seems entirely
3102 reasonable: after all the message is NOT "waiting for some host".
3103 This is so "obvious" that I'm not sure why it wasn't done
3104 previously. Hope I haven't missed anything, but it can't do any
3105 harm, as the worst effect is to miss an optimization.
3107 (ii) If, despite (i), such a delivery is accidentally attempted, the
3108 routing retry time is respected, so at least it doesn't keep
3109 hammering the server.
3111 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3112 in ${readsocket because some servers need this prod.
3114 PH/32 Added some extra debug output when updating a wait-xxx database.
3116 PH/33 The hint "could be header name not terminated by colon", which has been
3117 given for certain expansion errors for a long time, was not being given
3118 for the ${if def:h_colon_omitted{... case.
3120 PH/34 The spec says: "With one important exception, whenever a domain list is
3121 being scanned, $domain contains the subject domain." There was at least
3122 one case where this was not true.
3124 PH/35 The error "getsockname() failed: connection reset by peer" was being
3125 written to the panic log as well as the main log, but it isn't really
3126 panic-worthy as it just means the connection died rather early on. I have
3127 removed the panic log writing for the ECONNRESET error when getsockname()
3130 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3131 runs only) independently of the message's sender address. This meant
3132 that, if the 4xx error was in fact related to the sender, a different
3133 message to the same recipient with a different sender could confuse
3134 things. In particular, this can happen when sending to a greylisting
3135 server, but other circumstances could also provoke similar problems.
3136 I have changed the default so that the retry time for these errors is now
3137 based a combination of the sender and recipient addresses. This change
3138 can be overridden by setting address_retry_include_sender=false in the
3141 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3142 remote server are returned as part of bounce messages. This was not
3143 happening for LMTP over a pipe (the lmtp transport), but now it is the
3144 same for both kinds of LMTP.
3146 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3147 in header lines that were in fact CNAMEs. This is no longer the case.
3149 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3150 and queue runs started by the daemon processed all messages. This has
3151 been fixed so that -R and -S can now usefully be given with -q<time>.
3153 PH/40 Import PCRE release 6.7 (fixes some bugs).
3155 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3157 PH/42 Give an error if -q is specified more than once.
3159 PH/43 Renamed the variables $interface_address and $interface_port as
3160 $received_ip_address and $received_port, to make it clear that these
3161 values apply to message reception, and not to the outgoing interface when
3162 a message is delivered. (The old names remain recognized, of course.)
3164 PH/44 There was no timeout on the connect() call when using a Unix domain
3165 socket in the ${readsocket expansion. There now is.
3167 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3168 be meaningful with "accept".
3170 SC/01 Eximstats V1.43
3171 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3173 SC/02 Eximstats V1.44
3174 Use a glob alias rather than an array ref in the generated
3175 parser. This improves both readability and performance.
3177 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3178 Collect SpamAssassin and rejection statistics.
3179 Don't display local sender or destination tables unless
3180 there is data to show.
3181 Added average volumes into the top table text output.
3183 SC/04 Eximstats V1.46
3184 Collect data on the number of addresses (recipients)
3185 as well as the number of messages.
3187 SC/05 Eximstats V1.47
3188 Added 'Message too big' to the list of mail rejection
3189 reasons (thanks to Marco Gaiarin).
3191 SC/06 Eximstats V1.48
3192 Mainlog lines which have GMT offsets and are too short to
3193 have a flag are now skipped.
3195 SC/07 Eximstats V1.49 (Alain Williams)
3196 Added the -emptyok flag.
3198 SC/08 Eximstats V1.50
3199 Fixes for obtaining the IP address from reject messages.
3201 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3202 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3203 whitespace changes from 4.64-PH/27
3205 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3208 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3209 are found, allow negative numbers in numeric criteria)
3211 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3213 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3214 to match changes made in 4.64-PH/43
3216 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3218 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3219 "domain" is actually an IP address, and force a failure. This locks out
3220 those revolvers/nameservers that support "A-for-A" lookups, in
3221 contravention of the specifications.
3223 PH/48 When a host name was looked up from an IP address, and the subsequent
3224 forward lookup of the name timed out, the host name was left in
3225 $sender_host_name, contrary to the specification.
3227 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3228 restricted to single-key lookups, Exim was not diagnosing an error if
3229 * or *@ was used with a query-style lookup.
3231 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3233 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3234 long ago, but noone interested enough thought of it. Let's just say that
3235 the "1.1" means that there are some new functions that weren't there at
3236 some point in the past.
3238 PH/51 Error processing for expansion failure of helo_data from an smtp
3239 transport during callout processing was broken.
3241 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3242 tested/used via the -bh/-bhc/-bs options.
3244 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3245 bug, fixed in subsequent PCRE releases).
3247 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3248 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3250 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3256 SC/01 Use a glob alias rather than an array ref in eximstats generated
3257 parser. This improves both readability and performance.
3259 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3260 Don't display local sender or destination tables in eximstats unless
3261 there is data to show.
3262 Added average volumes into the eximstats top table text output.
3264 SC/03 Collect data on the number of addresses (recipients) as well
3265 as the number of messages in eximstats.
3267 TF/01 Correct an error in the documentation for the redirect router. Exim
3268 does (usually) call initgroups() when daemonizing.
3270 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3271 with consistent privilege compared to when running as a daemon.
3273 TF/03 Note in the spec that $authenticated_id is not set for local
3274 submissions from trusted users.
3276 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3277 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3279 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3280 by adding some example configuration directives to the default
3281 configuration file. A little bit of work is required to uncomment the
3282 directives and define how usernames and passwords are checked, but
3283 there is now a framework to start from.
3285 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3286 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3287 without this. I don't know how relevant this is to other LDAP libraries.
3289 PH/02 Add the verb name to the "unknown ACL verb" error.
3291 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3293 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3295 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3296 directory not expanded when it should be if an expanded home directory
3297 was set for the address (which is overridden by the transport).
3299 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3302 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3303 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3304 because it is too late at that time, and has no effect.
3306 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3307 security issue with \' (bugzilla #107). I could not use the
3308 PQescapeStringConn() function, because it needs a PGconn value as one of
3311 PH/08 When testing addresses using -bt, indicate those final addresses that
3312 are duplicates that would not cause an additional delivery. At least one
3313 person was confused, thinking that -bt output corresponded to deliveries.
3314 (Suppressing duplicates isn't a good idea as you lose the information
3315 about possibly different redirections that led to the duplicates.)
3317 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3318 systems where poll() doesn't work, in particular OS X.
3320 PH/10 Added more information to debugging output for retry time not reached.
3322 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3323 operations in malware.c.
3325 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3328 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3329 syslog_duplication set false, log lines that would normally be written
3330 both the the main log and to the reject log were not written to syslog at
3333 PH/14 In the default configuration, change the use of "message" in ACL warn
3334 statements to "add_header".
3336 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3337 not followed by a command (e.g. "seen endif").
3339 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3340 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3343 PH/17 Added extra conditions to the default value of delay_warning_condition
3347 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3348 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3349 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3352 The Auto-Submitted: and various List- headers are standardised, whereas I
3353 don't think Precedence: ever was.
3355 PH/18 Refactored debugging code in route_finduser() to show more information,
3356 in particular, the error code if getpwnam() issues one.
3358 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3359 This is apparently needed in addition to the PH/07 change above to avoid
3360 any possible encoding problems.
3362 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3363 but not after initializing Perl.
3365 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3366 output them only if debugging. By default they are written stderr,
3367 apparently, which is not desirable.
3369 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3372 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3375 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3377 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3378 authenticated or an ident call has been made. Suppress the default
3379 values for $authenticated_id and $authenticated_sender (but permit -oMai
3380 and -oMas) when testing with -bh.
3382 PH/24 Re-jigged the order of the tests in the default configuration so that the
3383 tests for valid domains and recipients precede the DNS black list and CSA
3384 tests, on the grounds that those ones are more expensive.
3386 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3387 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3388 This bug exists in every version of Exim that I still have, right back to
3391 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3392 However, an attempt to turn on case-sensitivity in a regex key by
3393 including (?-i) didn't work because the subject string was already
3394 lowercased, and the effects were non-intuitive. It turns out that a
3395 one-line patch can be used to allow (?-i) to work as expected.
3401 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3402 other effects) broke the use of negated acl sub-conditions.
3404 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3407 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3408 "Deny" causes Exim to reject the incoming connection with a 554 error.
3409 Unfortunately, if there is a major crisis, such as a disk failure,
3410 tcp-wrappers gives "deny", whereas what one would like would be some
3411 kind of temporary error. A kludge has been added to help with this.
3412 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3413 554 error is used if errno is still zero or contains ENOENT (which occurs
3414 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3417 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3419 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3420 errors. Otherwise a message that provokes a temporary error (when other
3421 messages do not) can cause a whole host to time out.
3423 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3424 addresses were routed directly to files or pipes from a redirect router.
3425 File deliveries just didn't batch; pipe deliveries might have suffered
3428 PH/06 A failure to get a lock for a hints database would erroneously always say
3429 "Failed to get write lock", even when it was really a read lock.
3431 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3432 of 0600. This has been changed to use the value of the lockfile_mode
3433 option (which defaults to 0600).
3435 PH/08 Applied small patch from the Sieve maintainer.
3437 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3438 folder from quota calculations, a direct delivery into this folder messed
3439 up the contents of the maildirsize file. This was because the regex was
3440 used only to exclude .Trash (or whatever) when the size of the mailbox
3441 was calculated. There was no check that a delivery was happening into an
3442 excluded directory. This bug has been fixed by ignoring all quota
3443 processing for deliveries into excluded directories.
3445 PH/10 Added the maildirfolder_create_regex option to appendfile.
3451 PH/01 The code for finding all the local interface addresses on a FreeBSD
3452 system running IPv6 was broken. This may well have applied to all BSD
3453 systems, as well as to others that have similar system calls. The broken
3454 code found IPv4 interfaces correctly, but gave incorrect values for the
3455 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3456 that it would not match correctly against @[] and not recognize the IPv6
3459 PH/02 The ipliteral router was not recognizing addresses of the form user@
3460 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3462 PH/03 Added disable_ipv6.
3464 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3465 decoded form, because it is most often used to construct To: headers
3466 lines in autoreplies, and the decoded form may well be syntactically
3467 invalid. However, $reply_address has leading white space removed, and all
3468 newlines turned into spaces so that the autoreply transport does not
3471 PH/05 If group was specified without a user on a router, and no group or user
3472 was specified on a transport, the group from the router was ignored.
3474 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3475 for visible compile-time settings that can be used to change these
3476 numbers, for those that want even more. Backwards compatibility with old
3477 spool files has been maintained. However, going back to a previous Exim
3478 release will lost any variables that are in spool files.
3480 PH/07 Two small changes when running in the test harness: increase delay when
3481 passing a TCP/IP connection to a new process, in case the original
3482 process has to generate a bounce, and remove special handling of
3483 127.0.0.2 (sic), which is no longer necessary.
3485 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3486 be the same on different OS.
3488 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3491 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3492 whether --show-vars was specified or not
3494 JJ/02 exipick: Added support for new ACL variable spool format introduced
3497 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3498 syntactically invalid From: or Reply-to: line, and a filter used this to
3499 generate an autoreply, and therefore failed to obtain an address for the
3500 autoreply, Exim could try to deliver to a non-existent relative file
3501 name, causing unrelated and misleading errors. What now happens is that
3502 it logs this as a hard delivery error, but does not attempt to create a
3505 PH/11 The exinext utility has a -C option for testing purposes, but although
3506 the given file was scanned by exinext itself; it wasn't being passed on
3507 when Exim was called.
3509 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3510 an end-of-file indication when reading a command response.
3512 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3513 compiled. In many other places in Exim, IPv6 addresses are always
3514 recognized, so I have changed this. It also means that IPv4 domain
3515 literals of the form [IPV4:n.n.n.n] are now always recognized.
3517 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3518 used if the router is not running as root, for example, when verifying at
3519 ACL time, or when using -bh. The debugging output from this situation was
3520 non-existent - all you got was a failure to exec. I have made two
3523 (a) Failures to set uid/gid, the current directory, or a process leader
3524 in a subprocess such as that created by queryprogram now generate
3525 suitable debugging output when -d is set.
3527 (b) The queryprogram router detects when it is not running as root,
3528 outputs suitable debugging information if -d is set, and then runs
3529 the subprocess without attempting to change uid/gid.
3531 PH/15 Minor change to Makefile for building test_host (undocumented testing
3534 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3535 additional section of a DNS packet that returns MX or SRV records.
3536 Instead, it always explicitly searches for A/AAAA records. This avoids
3537 major problems that occur when a DNS server includes only records of one
3538 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3539 fixed another bug: if SRV records were looked up and the corresponding
3540 address records were *not* found in the additional section, the port
3541 values from the SRV records were lost.
3543 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3544 using the correct key (the original address) when searching the retry
3545 rules in order to find which one to use for generating the retry hint.
3547 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3548 adding the default one. Similarly, if it contains a Reply-To: header, the
3549 errors_reply_to option, if set, is not used.
3551 PH/19 When calculating a retry time, Exim used to measure the "time since
3552 failure" by looking at the "first failed" field in the retry record. Now
3553 it does not use this if it is later than than the arrival time of the
3554 message. Instead it uses the arrival time. This makes for better
3555 behaviour in cases where some deliveries succeed, thus re-setting the
3556 "first failed" field. An example is a quota failure for a huge message
3557 when small messages continue to be delivered. Without this change, the
3558 "time since failure" will always be short, possible causing more frequent
3559 delivery attempts for the huge message than are intended.
3560 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3562 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3563 $1, $2, $3) because the numerical variables can be reset during some
3564 expansion items (e.g. "match"), thereby losing the authentication data.
3566 PH/21 Make -bV show the size of off_t variables so that the test suite can
3567 decide whether to run tests for quotas > 2G.
3569 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3570 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3571 filecount value is greater than 2G or if a quota value is greater than 2G
3572 on a system where the size of off_t is not greater than 4, a panic error
3575 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3576 never match. The debug and -bh output now contains an explicit error
3577 message indicating a malformed IPv4 address or mask.
3579 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3580 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3581 PH/23 above applies.
3583 PH/25 Do not write to syslog when running in the test harness. The only
3584 occasion when this arises is a failure to open the main or panic logs
3585 (for which there is an explicit test).
3587 PH/26 Added the /no_tell option to "control=freeze".
3589 PH/27 If a host name lookup failed very early in a connection, for example, if
3590 the IP address matched host_lookup and the reverse lookup yielded a name
3591 that did not have a forward lookup, an error message of the form "no IP
3592 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3593 could be logged. Now it outputs the IP address instead of "NULL".
3595 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3596 allows the sender and the authenticated sender to be set when
3597 submitting a message from within Exim. Since child_open_exim() is
3598 documented for local_scan(), the new function should be too.
3600 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3601 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3602 results in an empty string is now treated as unset.
3604 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3606 PH/31 Added sender_verify_fail logging option.
3608 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3609 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3610 tidied the source and removed it altogether.
3612 PH/33 When a queue run was abandoned because the load average was too high, a
3613 log line was always written; now it is written only if the queue_run log
3614 selector is set. In addition, the log line for abandonment now contains
3615 information about the queue run such as the pid. This is always present
3616 in "start" and "stop" lines but was omitted from the "abandon" line.
3618 PH/34 Omit spaces between a header name and the colon in the error message that
3619 is given when verify = headers_syntax fails (if there are lots of them,
3620 the message gets confusing).
3622 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3623 names, as there are now some PTR records that contain slashes. This check
3624 is only to protect against broken name servers that fall over on strange
3625 characters, so the fact that it applies to all lookups doesn't matter.
3627 PH/36 Now that the new test suite is complete, we can remove some of the
3628 special code in Exim that was needed for the old test suite. For example,
3629 sorting DNS records because real resolvers return them in an arbitrary
3630 order. The new test suite's fake resolver always returns records in the
3633 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3634 bounces) except when queue_only is set, to avoid logging races between
3635 the different processes.
3637 PH/38 Panic-die if .include specifies a non-absolute path.
3639 PH/39 A tweak to the "H" retry rule from its user.
3641 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3642 a label. They prevented compilation on older perls.
3644 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3645 a warning to be raised on newish perls.
3647 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3648 on queue. Changes to match documented behaviour of showing count of
3649 messages matching specified criteria.
3651 PH/40 Changed the default ident timeout from 30s to 5s.
3653 PH/41 Added support for the use of login_cap features, on those BSD systems
3654 that have them, for controlling the resources used by pipe deliveries.
3656 PH/42 The content-scanning code uses fopen() to create files in which to put
3657 message data. Previously it was not paying any attention to the mode of
3658 the files. Exim runs with umask(0) because the rest of the code creates
3659 files with open(), and sets the required mode explicitly. Thus, these
3660 files were ending up world-writeable. This was not a big issue, because,
3661 being within the spool directory, they were not world-accessible. I have
3662 created a function called modefopen, which takes an additional mode
3663 argument. It sets umask(777), creates the file, chmods it to the required
3664 mode, then resets the umask. All the relevant calls to fopen() in the
3665 content scanning code have been changed to use this function.
3667 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3668 to 24 hours. This avoids potential overflow problems when processing G
3669 and H retry rules. I suspect nobody ever tinkers with this value.
3671 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3673 PH/45 When the plaintext authenticator is running as a client, the server's
3674 challenges are checked to ensure they are valid base64 strings. By
3675 default, the authentication attempt is cancelled if an invalid string is
3676 received. Setting client_ignore_invalid_base64 true ignores these errors.
3677 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3678 they are received. Thus, the responses can be made to depend on the
3679 challenges. If an invalid string is ignored, an empty string is placed in
3682 PH/46 Messages that are created by the autoreply transport now contains a
3683 References: header, in accordance with RFCs 2822 and 3834.
3685 PH/47 Added authenticated_sender_force to the smtp transport.
3687 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3689 PH/49 Installed latest patch from the Sieve maintainer.
3691 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3692 was also set, the appendfile transport was unnecessarily scanning a
3693 directory of message files (e.g. for maildir delivery) to find the count
3694 of files (along with the size), even though it did not need this
3695 information. It now does the scan only if it needs to find either the
3696 size of the count of files.
3698 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3700 PH/52 Two bugs concerned with error handling when the smtp transport is
3703 (i) Exim was not creating retry information for temporary errors given
3704 for individual recipients after the DATA command when the smtp transport
3705 was used in LMTP mode. This meant that they could be retried too
3706 frequently, and not timed out correctly.
3708 (ii) Exim was setting the flag that allows error details to be returned
3709 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3710 recipients that were returned after the DATA command.
3712 PH/53 This is related to PH/52, but is more general: for any failing address,
3713 when detailed error information was permitted to be returned to the
3714 sender, but the error was temporary, then after the final timeout, only
3715 "retry timeout exceeded" was returned. Now it returns the full error as
3716 well as "retry timeout exceeded".
3718 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3719 do this, and (what is worse) MTAs that accept it.
3721 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3722 will now be deprecated.
3724 PH/56 New os.c-cygwin from the Cygwin maintainer.
3726 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3727 formats (previously only available in exim formats via -bpr, -bpru,
3728 and -bpra. Now also available in native and exiqgrep formats)
3730 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3731 with very large, slow to parse queues
3733 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3735 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3737 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3738 responses to authentication challenges, though it was showing the
3739 challenges; (ii) I've removed the CR characters from the debug output for
3742 PH/58 Allow for the insertion of a newline as well as a space when a string
3743 is turned into more than one encoded-word during RFC 2047 encoding. The
3744 Sieve code now uses this.
3746 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3747 data_4xx, lost_connection, tls_required.
3749 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3750 message was being sent as an SMTP response.
3752 PH/61 Add -l and -k options to exicyclog.
3754 PH/62 When verifying, if an address was redirected to one new address, so that
3755 verification continued, and the new address failed or deferred after
3756 having set something in $address_data, the value of $address_data was not
3757 passed back to the ACL. This was different to the case when no
3758 redirection occurred. The value is now passed back in both cases.
3760 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3761 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3762 use login_cap.h, so on its own it isn't the distinguishing feature. The
3763 new name refers directly to the setclassresources() function.
3765 PH/65 Added configuration files for NetBSD3.
3767 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3769 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3770 is preferred over IPv4.
3772 PH/68 The bounce_return_message and bounce_return_body options were not being
3773 honoured for bounces generated during the reception of non-SMTP messages.
3774 In particular, this applied to messages rejected by the ACL. This bug has
3775 been fixed. However, if bounce_return_message is true and bounce_return_
3776 body is false, the headers that are returned for a non-SMTP message
3777 include only those that have been read before the error was detected.
3778 (In the case of an ACL rejection, they have all been read.)
3780 PH/69 The HTML version of the specification is now built in a directory called
3781 spec_html instead of spec.html, because the latter looks like a path with
3782 a MIME-type, and this confuses some software.
3784 PH/70 Catch two compiler warnings in sieve.c.
3786 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3787 function verify_get_ident() calls ip_connect() to connect a socket, but
3788 if the "connect()" function timed out, ip_connect() used to close the
3789 socket. However, verify_get_ident() also closes the socket later, and in
3790 between Exim writes to the log, which may get opened at this point. When
3791 the socket was closed in ip_connect(), the log could get the same file
3792 descriptor number as the socket. This naturally causes chaos. The fix is
3793 not to close the socket in ip_connect(); the socket should be closed by
3794 the function that creates it. There was only one place in the code where
3795 this was missing, in the iplookup router, which I don't think anybody now
3796 uses, but I've fixed it anyway.
3798 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3799 well as to direct DNS lookups. Otherwise the handling of names in host
3800 lists is inconsistent and therefore confusing.
3806 PH/01 Two changes to the default runtime configuration:
3808 (1) Move the checks for relay_from_hosts and authenticated clients from
3809 after to before the (commented out) DNS black list checks.
3811 (2) Add control=submission to the relay_from_hosts and authenticated
3812 clients checks, on the grounds that messages accepted by these
3813 statements are most likely to be submissions.
3815 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3817 (1) Generate an error if the third argument for the ${prvs expansion is
3820 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3823 (3) Reset the variables that are obtained from the first argument of
3824 ${prvscheck and used in the second argument before leaving the code,
3825 because their memory is reclaimed, so using them afterwards may do
3828 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3829 one (it's much easier than Tom thought :-).
3831 (5) Because of (4), we can now allow for the use of $prvscheck_result
3832 inside the third argument.
3834 PH/03 For some reason, the default setting of PATH when running a command from
3835 a pipe transport was just "/usr/bin". I have changed it to
3838 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3839 anything to be listed in the output from -bV.
3841 PH/05 When a filter generated an autoreply, the entire To: header line was
3842 quoted in the delivery log line, like this:
3844 => >A.N.Other <ano@some.domain> <original@ddress> ...
3846 This has been changed so that it extracts the operative address. There
3847 may be more than one such address. If so, they are comma-separated, like
3850 => >ano@some.domain,ona@other.domain <original@ddress> ...
3852 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3853 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3854 not being looked up in the rDNS to get a host name, Exim was showing the
3855 IP address twice in Received: lines, even though the IP addresses were
3856 identical. For example:
3858 Received: from [1.2.3.4] (helo=[1.2.3.4])
3860 However, if the real host name was known, it was omitting the HELO data
3861 if it matched the actual IP address. This has been tidied up so that it
3862 doesn't show the same IP address twice.
3864 PH/07 When both +timestamp and +memory debugging was on, the value given by
3865 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3866 called by the debug printing, thereby overwriting the timestamp buffer.
3867 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3869 PH/08 When the original message was included in an autoreply transport, it
3870 always said "this is a copy of the message, including all the headers",
3871 even if body_only or headers_only was set. It now gives an appropriate
3874 PH/09 Applied a patch from the Sieve maintainer which:
3876 o fixes some comments
3877 o adds the (disabled) notify extension core
3878 o adds some debug output for the result of if/elsif tests
3879 o points to the current vacation draft in the documentation
3880 and documents the missing references header update
3884 o fixes a bug in processing the envelope test (when testing
3885 multiple envelope elements, the last element determined the
3888 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3889 Electronic Mail") by including:
3891 Auto-submitted: auto-generated
3893 in the messages that it generates (bounce messages and others, such as
3894 warnings). In the case of bounce messages for non-SMTP messages, there was
3895 also a typo: it was using "Auto_submitted" (underscore instead of
3896 hyphen). Since every message generated by Exim is necessarily in response
3897 to another message, thes have all been changed to:
3899 Auto-Submitted: auto-replied
3901 in accordance with these statements in the RFC:
3903 The auto-replied keyword:
3905 - SHOULD be used on messages sent in direct response to another
3906 message by an automatic process,
3908 - MUST NOT be used on manually-generated messages,
3910 - MAY be used on Delivery Status Notifications (DSNs) and Message
3911 Disposition Notifications (MDNs),
3913 - MUST NOT be used on messages generated by automatic or periodic
3914 processes, except for messages which are automatic responses to
3917 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3918 to the default Received: header definition.
3920 PH/12 Added log selector acl_warn_skipped (default on).
3922 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3923 variables because (a) they are in the wrong storage pool and (b) even if
3924 they were copied, it wouldn't work properly because of the caching.
3926 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3927 checking when decoding. Apparently there are clients that generate
3928 overlong encoded strings. Why am I not surprised?
3930 PH/15 If the first argument of "${if match_address" was not empty, but did not
3931 contain an "@" character, Exim crashed. Now it writes a panic log message
3932 and treats the condition as false.
3934 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3936 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3937 extension "envelope-auth". The code is finished and in agreement with
3938 other implementations, but there is no documentation so far and in fact,
3939 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3940 not changing the active code.
3942 Print executed "if" and "elsif" statements when debugging is used. This
3943 helps a great deal to understand what a filter does.
3945 Document more things not specified clearly in RFC3028. I had all this
3946 sorted out, when out of a sudden new issues came to my mind. Oops."
3948 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3951 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3952 canonical form (as documented). However, after a host name lookup from
3953 the IP address, check_host() was doing a simple string comparison with
3954 addresses acquired from the DNS when checking that the found name did
3955 have the original IP as one of its addresses. Since any found IPv6
3956 addresses are likely to be in abbreviated form, the comparison could
3957 fail. Luckily, there already exists a function for doing the comparison
3958 by converting both addresses to binary, so now that is used instead of
3959 the text comparison.
3961 PH/20 There was another similar case to PH/19, when a complete host name was
3962 given in a host list; looking up its IP address could give an abbreviated
3963 form, whereas the current host's name might or might not be abbreviated.
3964 The same fix has been applied.
3970 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3971 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3974 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3975 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3977 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3979 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3980 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3981 Also, for "Auto-Submitted", treat anything other than "no" as
3982 non-personal, in accordance with RFC 3834. (Previously it treated
3983 anything starting "auto-" as non-personal.)
3985 TF/01 The control=submission/name=... option had a problem with syntax
3986 errors if the name included a slash character. The /name= option
3987 now slurps the rest of the string, so it can include any characters
3988 but it must come last in the list of options (after /sender_retain
3991 PH/05 Some modifications to the interface to the fake nameserver for the new
3999 TK/01 Added the "success_on_redirect" address verification option. See
4000 NewStuff for rationale and an example.
4002 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4004 PH/02 Patch to exigrep to allow it to work on syslog lines.
4006 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4007 fread() to skip over the body file's header line, because in Cygwin the
4008 header line is locked and is inaccessible.
4010 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4011 co-exist for some time) to make it clear that it is the Exim ID that is
4012 referenced, not the Message-ID: header line.
4014 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4015 string_format() function, because snprintf() does not exist on all
4018 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4019 ${stat: expansion item.
4021 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4022 protocol synchronization error", to keep the pedants happy.
4024 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4025 well as for IRIX systems, when gcc is being used. See the host.c source
4028 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4030 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4033 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4034 transport and to the smtp transport in LMTP mode.
4036 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4038 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4039 run a filter in a subprocess. This could lead to confusion in subsequent
4040 lookups in the parent process. There should also be a search_tidyup() at
4041 the end of the subprocess.
4043 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4044 only if the host matched helo_try_verify_hosts, which caused the
4045 verification to occur when the EHLO/HELO command was issued. The ACL just
4046 tested the remembered result. Now, if a previous verification attempt has
4047 not happened, "verify = helo" does it there and then.
4049 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4051 TK/03 Fix log output including CR from clamd.
4053 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4054 exist provoked a memory error which could cause a segfault.
4056 PH/15 Installed PCRE 6.2
4058 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4060 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4061 of the problem. Specifically, suggested +O2 rather than +O1 for the
4064 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4066 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4067 record, the port was not correctly logged when the outgoing_port log
4068 selector was set (it logged the transort's default port).
4070 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4071 fallback_hosts, and "hosts" in the smtp transport.
4073 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4074 host errors such as "Connection refused".
4076 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4077 authentication with radiusclient 0.4.9:
4079 - Error returned from rc_read_config was caught wrongly
4080 - Username/password not passed on to radius server due to wrong length.
4082 The presumption is that some radiusclient API changes for 4.51/PH/17
4083 were not taken care of correctly. The code is still untested by me (my
4084 Linux distribution still has 0.3.2 of radiusclient), but it was
4085 contributed by a Radius user.
4087 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4088 expanding the "port" option of the smtp transport.
4090 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4091 while reading a MIME header. Thanks to Tom Hughes for a patch.
4093 PH/24 Include config.h inside local_scan.h so that configuration settings are
4096 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4097 that in an ACL for RCPT (for example), you can examine exactly what was
4100 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4101 commands, but it was not correctly comparing the address with the actual
4102 client host address. Thus, it would show the EHLO address in Received:
4103 header lines when this was not necessary.
4105 PH/27 Added the % operator to ${eval:}.
4107 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4108 it should be ignoring failures (because with -C, for example, it has lost
4109 privilege). It wasn't ignoring creation failures other than "already
4112 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4115 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4116 because an input line was too long, either on its own, or by virtue of
4117 too many continuations, the temporary file was not being removed, and the
4118 return code was incorrect.
4120 PH/31 Missing "BOOL" in function definition in filtertest.c.
4122 PH/32 Applied Sieve patches from the maintainer.
4124 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4126 PH/33 Added "verify = not_blind".
4128 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4129 Local/Makefile (with some defaults set). These are used in built scripts
4130 such as exicyclog, but they have never been used in the exim_install
4131 script (though there are many overriding facilities there). I have
4132 arranged that the exim_install script now takes note of these two
4135 PH/35 Installed configuration files for Dragonfly.
4137 PH/36 When a locally submitted message by a trusted user did not contain a
4138 From: header, and the sender address was obtained from -f or from an SMTP
4139 MAIL command, and the trusted user did not use -F to supply a sender
4140 name, $originator_name was incorrectly used when constructing a From:
4141 header. Furthermore, $originator_name was used for submission mode
4142 messages from external hosts without From: headers in a similar way,
4143 which is clearly wrong.
4145 PH/37 Added control=suppress_local_fixups.
4147 PH/38 When log_selector = +received_sender was set, and the addition of the
4148 sender made the log line's construction buffer exactly full, or one byte
4149 less than full, an overflow happened when the terminating "\n" was
4152 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4153 when the result of a list match is failure because a DNS lookup failed.
4155 PH/40 RM_COMMAND is now used in the building process.
4157 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4158 the "build-* directories that it finds.
4160 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4161 address in a domain literal was a prefix of an interface address.
4163 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4164 when verifying a sender address, unless rewrite_headers is false.
4166 PH/44 Wrote a long comment about why errors_to addresses are verified as
4167 recipients, not senders.
4169 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4170 the ratelimit ACL was added.
4172 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4174 PH/46 Added extra information about PostgreSQL errors to the error string.
4176 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4177 avoiding the need to install special zones in a real server. This is
4178 backwards compatible; if it can't find the fake resolver, it drops back.
4179 Thus, both old and new test suites can be run.
4181 TF/02 Added util/ratelimit.pl
4183 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4184 clock is set back in time.
4186 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4187 Candler <B.Candler@pobox.com>.
4189 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4190 OK for addresses that are the result of redirections.
4192 PH/48 A number of further additions for the benefit of the new test suite,
4193 including a fake gethostbyname() that interfaces to the fake DNS resolver
4196 TF/06 The fix for widen_domains has also been applied to qualify_single and
4197 search_parents which are the other dnslookup options that can cause
4200 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4203 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4205 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4206 service" if it encountered an MX record with an empty target hostname.
4207 The message is now "an MX or SRV record indicated no SMTP service".
4209 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4210 if the DNS of the sending site is misconfigured. This is quite a
4211 common situation. This change restores the behaviour of treating a
4212 helo verification defer as a failure.
4214 PH/51 If self=fail was set on a router, the bounce message did not include the
4215 actual error message.
4221 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4223 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4224 command itself ended in error, the underlying message about the transport
4225 filter timeout was being overwritten with the pipe command error. Now the
4226 underlying error message should be appended to the second error message.
4228 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4230 PH/02 Reduce the amount of output that "make" produces by default. Full output
4231 can still be requested.
4233 PH/03 The warning log line about a condition test deferring for a "warn" verb
4234 was being output only once per connection, rather than after each
4235 occurrence (because it was using the same function as for successful
4236 "warn" verbs). This seems wrong, so I have changed it.
4238 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4239 it should not have, which might have caused a crash in the right
4240 circumstances, but probably never did.
4242 PH/04 Installed a modified version of Tony Finch's patch to make submission
4243 mode fix the return path as well as the Sender: header line, and to
4244 add a /name= option so that you can make the user's friendly name appear
4247 TF/03 Added the control = fakedefer ACL modifier.
4249 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4250 Mark Lowes for thorough testing.
4252 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4254 TK/03 Merged latest SRS patch from Miles Wilton.
4256 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4257 with the definition in sysexits.h (which is #included earlier).
4258 Fortunately, Exim does not actually use EX_OK. The code used to try to
4259 preserve the sysexits.h value, by assuming that macro definitions were
4260 scanned for macro replacements. I have been disabused of this notion,
4261 so now the code just undefines EX_OK before #including unistd.h.
4263 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4264 in the smtp transport. When a block could not be written in a single
4265 write() function, the timeout was being re-applied to each part-write.
4266 This seems wrong - if the receiver was accepting one byte at a time it
4267 would take for ever. The timeout is now adjusted when this happens. It
4268 doesn't have to be particularly precise.
4270 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4271 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4273 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4274 without a domain if the check to <postmaster@domain> fails.
4276 SC/01 Eximstats: added -xls and the ability to specify output files
4277 (patch written by Frank Heydlauf).
4279 SC/02 Eximstats: use FileHandles for outputting results.
4281 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4283 SC/04 Eximstats: fixed display of large numbers with -nvr option
4285 SC/05 Eximstats: fixed merging of reports with empty tables.
4287 SC/06 Eximstats: added the -include_original_destination flag
4289 SC/07 Eximstats: removed tabs and trailing whitespace.
4291 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4293 TK/06 MBOX spool code: Add real "From " MBOX separator line
4294 so the .eml file is really in mbox format (even though
4295 most programs do not really care). Patch from Alex Miller.
4297 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4298 The latter is generated from $received_to and is only set if the
4299 message has one envelope recipient. SA can use these headers,
4300 obviously out-of-the-box. Patch from Alex Miller.
4302 PH/08 The ${def test on a variable was returning false if the variable's
4303 value was "0", contrary to what the specification has always said!
4304 The result should be true unless the variable is empty.
4306 PH/09 The syntax error of a character other than { following "${if
4307 def:variable_name" (after optional whitespace) was not being diagnosed.
4308 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4309 accidental colon was present, for example, could give incorrect results.
4311 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4312 result is used (not including appendfile, where other changes are about
4315 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4316 This involved changing a lot of size variables from int to off_t. It
4317 should work with maildirs and everything.
4319 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4320 spamd dying while we are connected to it.
4322 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4325 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4326 function for BDB 4.3.
4328 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4330 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4331 the sources for PCRE have been re-arranged and more files are now
4334 PH/15 The code I had for printing potentially long long variables in PH/11
4335 above was not the best (it lost precision). The length of off_t variables
4336 is now inspected at build time, and an appropriate printing format (%ld
4337 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4338 to be "long long int" or "long int". This is needed for the internal
4339 formatting function string_vformat().
4341 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4342 the configuration file to be ":syslog", then the script "guesses" where
4343 the logs files are, rather than using the compiled in default. In our
4344 case the guess is not the same as the compiled default, so the script
4345 suddenly stopped working when I started to use syslog. The patch checks
4346 to see if log_file_path is "". If so, it attempts to read it from exim
4347 with no configuration file to get the compiled in version, before it
4348 falls back to the previous guessing code."
4350 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4351 implementing BATV in an Exim configuration. See NewStuff for the gory
4354 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4355 Makefile that are specific to HP-UX.
4357 PH/18 If the "use_postmaster" option was set for a recipient callout together
4358 with the "random" option, the postmaster address was used as the MAIL
4359 FROM address for the random test, but not for the subsequent recipient
4360 test. It is now used for both.
4362 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4363 patch removes a few documentation additions to RFC 3028, because the
4364 latest draft now contains them. It adds the new en;ascii-case comparator
4365 and a new error check for 8bit text in MIME parts. Comparator and
4366 require names are now matched exactly. I enabled the subaddress
4367 extension, but it is not well tested yet (read: it works for me)."
4369 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4370 rework some of the code of TK/09 above to avoid the hardwired use of
4371 "%lld" and "long long". Replaced the call to snprintf() with a call to
4374 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4375 records point to non-existent hosts", "retry timeout exceeded", and
4376 "retry time not reached for any host after a long failure period".
4378 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4379 experimental DomainKeys support:
4381 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4382 (2) On an error such as an illegally used "control", the wrong name for
4383 the control was given.
4385 These problems did NOT occur unless DomainKeys support was compiled.
4387 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4389 PH/24 Added ${if match_ip condition.
4391 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4392 databases so that it will be absolutely obvious if a crash occurs in the
4393 DB library. This is a regular occurrence (often caused by mis-matched
4396 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4397 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4398 picked up on a user's system that detects such things. There doesn't seem
4399 to be a gcc warning option for this - only an attribute that has to be
4400 put on the function's prototype. It seems that in Fedora Core 4 they have
4401 set this on a number of new functions. No doubt there will be more in due
4404 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4405 specify a transport. However, if an address that was verified by such a
4406 router was the subject of a callout, Exim crashed because it tried to
4407 read the rcpt_include_affixes from the non-existent transport. Now it
4408 just assumes that the setting of that option is false. This bug was
4409 introduced by 4.51/PH/31.
4411 PH/28 Changed -d+all to exclude +memory, because that information is very
4412 rarely of interest, but it makes the output a lot bigger. People tend to
4413 do -d+all out of habit.
4415 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4416 code in os-type was giving problems when libc.so lives in lib64, like on
4419 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4420 aren't the modern standard, and it seems that some systems' include files
4421 don't always have them. Exim was already checking for some of the newer
4422 ones like T_AAAA, and defining it itself. I've added checks for all the
4423 record types that Exim uses.
4425 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4426 not automatically generating a new one, as it is supposed to. This
4427 prevented TLS from working. If the file did exist, but contained invalid
4428 data, a new version was generated, as expected. It was only the case of a
4429 non-existent file that was broken.
4431 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4432 with a change in libdomainkeys > 0.64.
4434 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4435 from DNS. If the selector record carries the flag, it now has
4436 precedence over the domain-wide flag.
4438 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4440 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4441 the use of a port name that isn't defined in /etc/services) occurred, the
4442 message was deferred as in a normal delivery, and thus remained on the
4443 spool, instead of being failed because of the mua_wrapper setting. This
4444 is now fixed, and I tidied up some of the mua_wrapper messages at the
4447 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4448 the messages in a hash of arrays rather than using individual hashes.
4449 This is a bit cleaner and results in dramatic memory savings, albeit
4450 at a slight CPU cost.
4452 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4453 as requested by Marc Sherman.
4455 SC/10 Eximstats: added histograms for user specified patterns as requested
4458 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4460 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4461 fopen() in the content-scanning modules that did not already have it.
4467 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4468 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4470 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4472 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4474 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4475 to test Sieve filters that use "vacation".
4477 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4478 that changes the way the GnuTLS parameters are stored in the cache file.
4479 The new format can be generated externally. For backward compatibility,
4480 if the data in the cache doesn't make sense, Exim assumes it has read an
4481 old-format file, and it generates new data and writes a new file. This
4482 means that you can't go back to an older release without removing the
4485 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4486 work if there are any delivery delays because "one_time" forces the
4487 parent to be marked "delivered", so its unseen clone is never tried
4488 again. For this reason, Exim now forbids the simultaneous setting of
4491 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4492 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4493 entry if you want to know the details. The fix, however, neglected to
4494 consider the case when local delivery batching is involved. The test for
4495 "previously delivered" was not happening when checking to see if an
4496 address could be batched with a previous (undelivered) one; under
4497 certain circumstances this could lead to multiple deliveries to the same
4500 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4501 in its include files, and this causes problems building Exim.
4503 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4504 header_syntax) but Exim was just ignoring anything given after a slash.
4505 In particular, this caused confusion with an attempt to use "verify =
4506 reverse_host_lookup/defer_ok". An error is now given when options are
4507 supplied for verify items that do not have them. (Maybe reverse_host_
4508 lookup should have a defer_ok option, but that's a different point.)
4510 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4511 defined by RFC 821) to 2048, because there were problems with some AUTH
4512 commands, and RFC 1869 says the size should be increased for extended
4513 SMTP commands that take arguments.
4515 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4518 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4519 "unknown" error; now it says that the functionality isn't in the binary.
4521 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4522 an address' error message when a string expansion fails (syntax or
4523 whatever). Otherwise the password may appear in the log. Following change
4524 PH/42 below, there is no longer a chance of it appearing in a bounce
4527 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4529 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4530 of its addresses was ever tried. (Bugzilla bug #2.)
4532 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4533 the result incorrectly in the debug output. (It correctly added a newline
4534 to what was transported.)
4536 TF/01 Added $received_time.
4538 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4539 commented out examples of how to interface to a virus scanner and to
4540 SpamAssassin. Also added commented examples of av_scanner and
4541 spamd_address settings.
4543 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4544 and controls are allowed in which ACLs. There were a couple of minor
4545 errors. Some of the entries in the conditions table (which is a table of
4546 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4547 negation of where the condition IS allowed.
4549 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4551 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4552 header file does not have a version number, so I've had to invent a new
4553 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4554 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4555 radiusclient), but it was contributed by a Radius user.
4557 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4558 files or directories, for OpenSSL.
4560 PH/19 When an Exim process that is running as root has to create an Exim log
4561 file, it does so in a subprocess that runs as exim:exim so as to get the
4562 ownership right at creation (otherwise, other Exim processes might see
4563 the file with the wrong ownership). There was no test for failure of this
4564 fork() call, which would lead to the process getting stuck as it waited
4565 for a non-existent subprocess. Forks do occasionally fail when resources
4566 run out. I reviewed all the other calls to fork(); they all seem to check
4569 PH/20 When checking for unexpected SMTP input at connect time (before writing
4570 the banner), Exim was not dealing correctly with a non-positive return
4571 from the read() function. If the client had disconnected by this time,
4572 the result was a log entry for a synchronization error with an empty
4573 string after "input=" when read() returned zero. If read() returned -1
4574 (an event I could not check), uninitialized data bytes were printed.
4575 There were reports of junk text (parts of files, etc) appearing after
4578 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4580 PH/22 Added support for macro redefinition, and (re)definition in between
4581 driver and ACL definitions.
4583 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4584 forgetting to use the resulting value; it was using the unexpanded value.
4586 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4587 hadn't been configured. The fix is from Juergen Kreileder, who
4588 understands it better than I do:
4590 "Here's what I see happening with three configured cyrus_sasl
4591 authenticators configured (plain, login, cram-md5):
4593 On startup auth_cyrus_sasl_init() gets called for each of these.
4594 This means three calls to sasl_listmech() without a specified mech_list.
4595 => SASL tests which mechs of all available mechs actually work
4596 => three warnings about OTP not working
4597 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4599 With the patch, sasl_listmech() also gets called three times. But now
4600 SASL's mech_list option is set to the server_mech specified in the the
4601 authenticator. Or in other words, the answer from sasl_listmech()
4602 gets limited to just the mech you're testing for (which is different
4604 => the return list contains just 'plain' or 'login', 'cram-md5' or
4605 nothing depending on the value of ob->server_mech.
4607 I've just tested the patch: Authentication still works fine,
4608 unavailable mechs specified in the exim configuration are still
4609 caught, and the auth.log warnings about OTP are gone."
4611 PH/25 When debugging is enabled, the contents of the command line are added
4612 to the debugging output, even when log_selector=+arguments is not
4615 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4616 answer is "GNU", and only if the return is "GNU/something" is the answer
4619 PH/27 $acl_verify_message is now set immediately after the failure of a
4620 verification in an ACL, and so is available in subsequent modifiers. In
4621 particular, the message can be preserved by coding like this:
4623 warn !verify = sender
4624 set acl_m0 = $acl_verify_message
4626 Previously, $acl_verify_message was set only while expanding "message"
4627 and "log_message" when a very denied access.
4629 PH/28 Modified OS/os.c-Linux with
4631 -#ifndef OS_LOAD_AVERAGE
4632 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4634 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4635 nomenclature these days.)
4637 PH/29 Installed patch from the Sieve maintainer that adds the options
4638 sieve_useraddress and sieve_subaddress to the redirect router.
4640 PH/30 In these circumstances:
4641 . Two addresses routed to the same list of hosts;
4642 . First host does not offer TLS;
4643 . First host accepts first address;
4644 . First host gives temporary error to second address;
4645 . Second host offers TLS and a TLS session is established;
4646 . Second host accepts second address.
4647 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4648 and peerdn, if requested) that were in fact used only for the second
4651 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4652 attention to any local part prefix or suffix that was matched by the
4653 router that accepted the address. It now behaves in the same way as it
4654 does for delivery: the affixes are removed from the local part unless
4655 rcpt_include_affixes is set on the transport.
4657 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4658 timeout during the DATA phase of an incoming message.
4660 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4661 applied a patch sanctioned by the Sieve maintainer.
4663 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4664 the uid or gid is negative. A case of a negative gid caused this to be
4665 noticed. The fix allows for either to be negative.
4667 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4668 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4671 PH/36 The change PH/12 above was broken. Fixed it.
4673 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4674 the grounds that routing the same address twice would always produce the
4675 same answer. This might have been true once, but it is certainly no
4676 longer true now. Routing a child address may depend on the previous
4677 routing that produced that child. Some complicated redirection strategies
4678 went wrong when messages had multiple recipients, and made Exim's
4679 behaviour dependent on the order in which the addresses were given.
4681 I have moved the duplicate checking until after the routing is complete.
4682 Exim scans the addresses that are assigned to local and remote
4683 transports, and removes any duplicates. This means that more work will be
4684 done, as duplicates will always all be routed, but duplicates are
4685 presumably rare, so I don't expect this is of any significance.
4687 For deliveries to pipes, files, and autoreplies, the duplicate checking
4688 still happens during the routing process, since they are not going to be
4691 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4692 It corrects a timeout issue with spamd. This is Ian's comment: "The
4693 background is that sometimes spamd either never reads data from a
4694 connection it has accepted, or it never writes response data. The exiscan
4695 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4696 blindly assumes that writes won't block so it may never time out."
4698 PH/39 Allow G after quota size as well as K and M.
4700 PH/40 The value set for $authenticated_id in an authenticator may not contain
4701 binary zeroes or newlines because the value is written to log lines and
4702 to spool files. There was no check on this. Now the value is run through
4703 the string_printing() function so that such characters are converted to
4704 printable escape sequences.
4706 PH/41 $message_linecount is a new variable that contains the total number of
4707 lines in the message. Compare $body_linecount, which is the count for the
4710 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4711 bounce and delay warning messages, except in certain special cases, which
4714 (a) An SMTP error message from a remote host;
4715 (b) A message specified in a :fail: redirection;
4716 (c) A message specified in a "fail" command in a system filter;
4717 (d) A message specified in a FAIL return from the queryprogram router;
4718 (e) A message specified by the cannot_route_message router option.
4720 In these cases only, Exim does include the error details in bounce and
4721 warning messages. There are also a few cases where bland messages such
4722 as "unrouteable address" or "local delivery error" are given.
4724 PH/43 $value is now also set for the "else" part of a ${run expansion.
4726 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4727 being worked on, but at least Exim now implements the latest version to
4730 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4731 process to complete was treated as a delivery failure, a timeout while
4732 writing the message to the pipe was logged, but erroneously treated as a
4733 successful delivery. Such timeouts include transport filter timeouts. For
4734 consistency with the overall process timeout, these timeouts are now
4735 treated as errors, giving rise to delivery failures by default. However,
4736 there is now a new Boolean option for the pipe transport called
4737 timeout_defer, which, if set TRUE, converts the failures into defers for
4738 both kinds of timeout. A transport filter timeout is now identified in
4741 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4742 systems where "make" and "gmake" are different, calling "gmake" at top
4743 level broke things. I've arranged for the value of $(MAKE) to be passed
4744 from the Makefile to this script so that it can call the same version of
4748 A note about Exim versions 4.44 and 4.50
4749 ----------------------------------------
4751 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4752 changes of various kinds. As a consequence, a big documentation update was
4753 needed. This delayed the release for rather longer than seemed good, especially
4754 in the light of a couple of (minor) security issues. Therefore, the changes
4755 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4756 release. So 4.44 and 4.50 are in effect two different branches that both start
4759 I have left the 4.50 change log unchanged; it contains all the changes since
4760 4.43. The change log for 4.44 is below; many of its items are identical to
4761 those for 4.50. This seems to be the most sensible way to preserve the
4762 historical information.
4768 1. Minor wording change to the doc/README.SIEVE file.
4770 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4771 computation of the current number of files was incorrect.
4773 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4774 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4777 4. Give more explanation in the error message when the command for a transport
4778 filter fails to execute.
4780 5. There are several places where Exim runs a non-Exim command in a
4781 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4782 was being done only for the command run by the queryprogram router. It is
4783 now done for all such subprocesses. The other cases are: ${run, transport
4784 filters, and the commands run by the lmtp and pipe transports.
4786 6. Added CONFIGURE_GROUP build-time option.
4788 7. Some older OS have a limit of 256 on the maximum number of file
4789 descriptors. Exim was using setrlimit() to set 1000 as a large value
4790 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4791 systems. I've change it so that if it can't get 1000, it tries for 256.
4793 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4794 was an oversight, and furthermore, ever since the addition of extra
4795 controls (e.g. 4.43/32), the checks on when to allow different forms of
4796 "control" were broken. There should now be diagnostics for all cases when a
4797 control that does not make sense is encountered.
4799 9. Added the /retain_sender option to "control=submission".
4801 10. $recipients is now available in the predata ACL (oversight).
4803 11. Tidy the search cache before the fork to do a delivery from a message
4804 received from the command line. Otherwise the child will trigger a lookup
4805 failure and thereby defer the delivery if it tries to use (for example) a
4806 cached ldap connection that the parent has called unbind on.
4808 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4809 of $address_data from the recipient verification was clobbered by the
4810 sender verification.
4812 13. The value of address_data from a sender verification is now available in
4813 $sender_address_data in subsequent conditions in the ACL statement.
4815 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4817 15. Added a new option "connect=<time>" to callout options, to set a different
4820 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4821 was its contents. (It was OK if the option was not defined at all.)
4823 17. A "Completed" log line is now written for messages that are removed from
4824 the spool by the -Mrm option.
4826 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4827 information about exactly what failed.
4829 19. Added -dd to debug only the daemon process.
4831 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4832 handles timeouts, both on the server side and network timeouts. Renamed the
4833 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4835 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4836 It is now set to "smtps".
4838 22. $host_address is now set to the target address during the checking of
4839 ignore_target_hosts.
4841 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4842 being passed; this would have caused $sender_host_name to have been used if
4843 matching the list had actually called for a host name (not very likely,
4844 since this list is usually IP addresses). A host name is now passed as
4847 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4848 code that specifies a non-restarting handler (typically sigaction() in
4849 modern systems) in an attempt to fix a rare and obscure crash bug.
4851 25. Narrowed the window for a race in the daemon that could cause it to ignore
4852 SIGCHLD signals. This is not a major problem, because they are used only to
4853 wake it up if nothing else does.
4855 26. A malformed maildirsize file could cause Exim to calculate negative values
4856 for the mailbox size or file count. Odd effects could occur as a result.
4857 The maildirsize information is now recalculated if the size or filecount
4860 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4861 support for a long time. Removed HAVE_SYS_VFS_H.
4863 28. Installed the latest version of exipick from John Jetmore.
4865 29. In an address list, if the pattern was not a regular expression, an empty
4866 subject address (from a bounce message) matched only if the pattern was an
4867 empty string. Non-empty patterns were not even tested. This was the wrong
4868 because it is perfectly reasonable to use an empty address as part of a
4869 database query. An empty address is now tested by patterns that are
4870 lookups. However, all the other forms of pattern expect the subject to
4871 contain a local part and a domain, and therefore, for them, an empty
4872 address still always fails if the pattern is not itself empty.
4874 30. Exim went into a mad DNS loop when attempting to do a callout where the
4875 host was specified on an smtp transport, and looking it up yielded more
4876 than one IP address.
4878 31. Re-factored the code for checking spool and log partition space into a
4879 function that finds that data and another that does the check. The former
4880 is then used to implement four new variables: $spool_space, $log_space,
4881 $spool_inodes, and $log_inodes.
4883 32. The RFC2047 encoding function was originally intended for short strings
4884 such as real names; it was not keeping to the 75-character limit for
4885 encoded words that the RFC imposes. It now respects the limit, and
4886 generates multiple encoded words if necessary. To be on the safe side, I
4887 have increased the buffer size for the ${rfc2047: expansion operator from
4890 33. It is now permitted to omit both strings after an "if" condition; if the
4891 condition is true, the result is "true". As before, when the second string
4892 is omitted, a false condition yields an empty string. This makes it less
4893 cumbersome to write custom ACL and router conditions.
4895 34. Failure to deliver a bounce message always caused it to be frozen, even if
4896 there was an errors_to setting on the router. The errors_to setting is now
4899 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4900 canonical form (fully expanded) before being placed in
4901 $sender_host_address.
4903 36. The table in the code that translates DNS record types into text (T_A to
4904 "A" for instance) was missing entries for NS and CNAME. It is just possible
4905 that this could have caused confusion if both these types were looked up
4906 for the same domain, because the text type is used as part of Exim's
4907 per-process caching. But the chance of anyone hitting this buglet seems
4910 37. The dnsdb lookup has been extended in a number of ways.
4912 (1) There is a new type, "zns", which walks up the domain tree until it
4913 finds some nameserver records. It should be used with care.
4915 (2) There is a new type, "mxh", which is like "mx" except that it returns
4916 just the host names, not the priorities.
4918 (3) It is now possible to give a list of domains (or IP addresses) to be
4919 looked up. The behaviour when one of the lookups defers can be
4920 controlled by a keyword.
4922 (4) It is now possible to specify the separator character for use when
4923 multiple records are returned.
4925 38. The dnslists ACL condition has been extended: it is now possible to supply
4926 a list of IP addresses and/or domains to be looked up in a particular DNS
4929 39. Added log_selector=+queue_time_overall.
4931 40. When running the queue in the test harness, wait just a tad after forking a
4932 delivery process, to get repeatability of debugging output.
4934 41. Include certificate and key file names in error message when GnuTLS fails
4935 to set them up, because the GnuTLS error message doesn't include the name
4936 of the failing file when there is a problem reading it.
4938 42. Allow both -bf and -bF in the same test run.
4940 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4942 44. The "Exiscan patch" is now merged into the mainline Exim source.
4944 45. Sometimes the final signoff response after QUIT could fail to get
4945 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4946 < 0 before doing a fflush(). This bug looks as though it goes back to the
4947 introduction of TLS in release 3.20, but "sometimes" must have been rare
4948 because the tests only now provoked it.
4950 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4951 (this can affect the format of dates).
4953 47. exim_tidydb, when checking for the continued existence of a message for
4954 which it has found a message-specific retry record, was not finding
4955 messages that were in split spool directories. Consequently, it was
4956 deleting retry records that should have stayed in existence.
4958 48. Steve fixed some bugs in eximstats.
4960 49. The SPA authentication driver was not abandoning authentication and moving
4961 on to the next authenticator when an expansion was forced to fail,
4962 contradicting the general specification for all authenticators. Instead it
4963 was generating a temporary error. It now behaves as specified.
4965 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4966 (the order specifies the preference for clients). The order is now AES256,
4967 AES128, 3DES, ARCFOUR128.
4969 51. Small patch to Sieve code - explicitly set From: when generating an
4972 52. Exim crashed if a remote delivery caused a very long error message to be
4973 recorded - for instance if somebody sent an entire SpamAssassin report back
4974 as a large number of 550 error lines. This bug was coincidentally fixed by
4975 increasing the size of one of Exim's internal buffers (big_buffer) that
4976 happened as part of the Exiscan merge. However, to be on the safe side, I
4977 have made the code more robust (and fixed the comments that describe what
4980 53. Now that there can be additional text after "Completed" in log lines (if
4981 the queue_time_overall log selector is set), a one-byte patch to exigrep
4982 was needed to allow it to recognize "Completed" as not the last thing in
4985 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4986 patch that reportedly fixes this has been added. I am not expert enough to
4987 create a test for it. This is what the patch creator wrote:
4989 "I found a little strange behaviour of ldap code when working with
4990 Windows 2003 AD Domain, where users was placed in more than one
4991 Organization Units. When I tried to give exim partial DN, the exit code
4992 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4993 But simultaneously result of request was absolutely normal ldap result,
4994 so I produce this patch..."
4996 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4997 REFERENCE, so I have modified the code to exclude the patch when that macro
5000 55. Some experimental protocols are using DNS PTR records for new purposes. The
5001 keys for these records are domain names, not reversed IP addresses. The
5002 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5003 leaves it alone. Component reversal etc. now happens only for IP addresses.
5006 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5008 57. Double the size of the debug message buffer (to 2048) so that more of very
5009 long debug lines gets shown.
5011 58. The exicyclog utility now does better if the number of log files to keep
5012 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5014 59. Two changes related to the smtp_active_hostname option:
5016 (1) $smtp_active_hostname is now available as a variable.
5017 (2) The default for smtp_banner uses $smtp_active_hostname instead
5018 of $primary_hostname.
5020 60. The host_aton() function is supposed to be passed a string that is known
5021 to be a valid IP address. However, in the case of IPv6 addresses, it was
5022 not checking this. This is a hostage to fortune. Exim now panics and dies
5023 if the condition is not met. A case was found where this could be provoked
5024 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5025 components; fortuitously, this particular loophole had already been fixed
5026 by change 4.50/55 above.
5028 If there are any other similar loopholes, the new check in host_aton()
5029 itself should stop them being exploited. The report I received stated that
5030 data on the command line could provoke the exploit when Exim was running as
5031 exim, but did not say which command line option was involved. All I could
5032 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5033 running as the user.
5036 61. There was a buffer overflow vulnerability in the SPA authentication code
5037 (which came originally from the Samba project). I have added a test to the
5038 spa_base64_to_bits() function which I hope fixes it.
5041 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5042 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5044 63. The daemon start-up calls getloadavg() while still root for those OS that
5045 need the first call to be done as root, but it missed one case: when
5046 deliver_queue_load_max is set with deliver_drop_privilege. This is
5047 necessary for the benefit of the queue runner, because there is no re-exec
5048 when deliver_drop_privilege is set.
5050 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5051 This has been fixed.
5053 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5054 was in use, was not putting the data itself into the right store pool;
5055 consequently, it could be overwritten for a subsequent message in the same
5056 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5059 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5061 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5062 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5063 the function were treating the return as a boolean value, which happened to
5064 work because 0=false and not-0=true, but is not correct code.
5066 68. The host_aton() function was not handling scoped IPv6 addresses (those
5067 with, for example, "%eth0" on the end) correctly.
5069 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5070 negated items (that is, ~something) in unsigned ints. Some compilers
5071 apparently mutter when there is no cast.
5073 70. If an address verification called from an ACL failed, and did not produce a
5074 user-specific message (i.e. there was only a "system" message), nothing was
5075 put in $acl_verify_message. In this situation, it now puts the system
5078 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5079 session; change 4.31/43 added the unwanted input to the log line - except
5080 that it did not do this in the start of session case. It now does.
5082 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5083 This is wrong and can cause the other end to generate a synchronization
5084 error if it is another Exim or anything else that does the synchronization
5085 check. A QUIT command is no longer sent after a timeout.
5087 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5088 during host lookups.
5090 74. The defer_ok option of callout verification was not working if it was used
5091 when verifying addresses in header lines, that is, for this case:
5093 verify = header_sender/callout=defer_ok
5095 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5096 those file descriptors could be used for SMTP connections. If anything
5097 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5098 could be sent to the SMTP client, causing chaos. The daemon now opens
5099 stdin, stdout, and stderr to /dev/null when it puts itself into the
5102 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5103 log by default. The user can override this with suitable Perl magic.
5105 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5106 the log message when discard triggers, was not working for the DATA ACL or
5107 for the non-SMTP ACL.
5109 78. Error message wording change in sieve.c.
5111 79. If smtp_accept_max_per_host was set, the number of connections could be
5112 restricted to fewer than expected, because the daemon was trying to set up
5113 a new connection before checking whether the processes handling previous
5114 connections had finished. The check for completed processes is now done
5115 earlier. On busy systems, this bug wouldn't be noticed because something
5116 else would have woken the daemon, and it would have reaped the completed
5119 80. If a message was submitted locally by a user whose login name contained one
5120 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5121 It caused a spool format error. I have fixed the spool reading code. A
5122 related problem was that the "from" clause in the Received: line became
5123 illegal because of the space(s). It is now covered by ${quote_local_part.
5125 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5128 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5130 83. Patch from Sieve maintainer to fix unterminated string problem in
5131 "vacation" handling.
5133 84. Some minor changes to the Linux configuration files to help with other
5134 OS variants using glibc.
5136 85. One more patch for Sieve to update vacation handling to latest spec.
5139 ----------------------------------------------------
5140 See the note above about the 4.44 and 4.50 releases.
5141 ----------------------------------------------------
5147 1. Change 4.43/35 introduced a bug that caused file counts to be
5148 incorrectly computed when quota_filecount was set in an appendfile
5151 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5152 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5155 3. Give more explanation in the error message when the command for a transport
5156 filter fails to execute.
5158 4. There are several places where Exim runs a non-Exim command in a
5159 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5160 was being done only for the command run by the queryprogram router. It is
5161 now done for all such subprocesses. The other cases are: ${run, transport
5162 filters, and the commands run by the lmtp and pipe transports.
5164 5. Some older OS have a limit of 256 on the maximum number of file
5165 descriptors. Exim was using setrlimit() to set 1000 as a large value
5166 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5167 systems. I've change it so that if it can't get 1000, it tries for 256.
5169 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5170 was an oversight, and furthermore, ever since the addition of extra
5171 controls (e.g. 4.43/32), the checks on when to allow different forms of
5172 "control" were broken. There should now be diagnostics for all cases when a
5173 control that does not make sense is encountered.
5175 7. $recipients is now available in the predata ACL (oversight).
5177 8. Tidy the search cache before the fork to do a delivery from a message
5178 received from the command line. Otherwise the child will trigger a lookup
5179 failure and thereby defer the delivery if it tries to use (for example) a
5180 cached ldap connection that the parent has called unbind on.
5182 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5183 of $address_data from the recipient verification was clobbered by the
5184 sender verification.
5186 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5187 was its contents. (It was OK if the option was not defined at all.)
5189 11. A "Completed" log line is now written for messages that are removed from
5190 the spool by the -Mrm option.
5192 12. $host_address is now set to the target address during the checking of
5193 ignore_target_hosts.
5195 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5196 being passed; this would have caused $sender_host_name to have been used if
5197 matching the list had actually called for a host name (not very likely,
5198 since this list is usually IP addresses). A host name is now passed as
5201 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5202 code that specifies a non-restarting handler (typically sigaction() in
5203 modern systems) in an attempt to fix a rare and obscure crash bug.
5205 15. Narrowed the window for a race in the daemon that could cause it to ignore
5206 SIGCHLD signals. This is not a major problem, because they are used only to
5207 wake it up if nothing else does.
5209 16. A malformed maildirsize file could cause Exim to calculate negative values
5210 for the mailbox size or file count. Odd effects could occur as a result.
5211 The maildirsize information is now recalculated if the size or filecount
5214 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5215 support for a long time. Removed HAVE_SYS_VFS_H.
5217 18. Updated exipick to current release from John Jetmore.
5219 19. Allow an empty sender to be matched against a lookup in an address list.
5220 Previously the only cases considered were a regular expression, or an
5223 20. Exim went into a mad DNS lookup loop when doing a callout where the
5224 host was specified on the transport, if the DNS lookup yielded more than
5227 21. The RFC2047 encoding function was originally intended for short strings
5228 such as real names; it was not keeping to the 75-character limit for
5229 encoded words that the RFC imposes. It now respects the limit, and
5230 generates multiple encoded words if necessary. To be on the safe side, I
5231 have increased the buffer size for the ${rfc2047: expansion operator from
5234 22. Failure to deliver a bounce message always caused it to be frozen, even if
5235 there was an errors_to setting on the router. The errors_to setting is now
5238 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5239 canonical form (fully expanded) before being placed in
5240 $sender_host_address.
5242 24. Updated eximstats to version 1.33
5244 25. Include certificate and key file names in error message when GnuTLS fails
5245 to set them up, because the GnuTLS error message doesn't include the name
5246 of the failing file when there is a problem reading it.
5248 26. Expand error message when OpenSSL has problems setting up cert/key files.
5251 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5252 (this can affect the format of dates).
5254 28. exim_tidydb, when checking for the continued existence of a message for
5255 which it has found a message-specific retry record, was not finding
5256 messages that were in split spool directories. Consequently, it was
5257 deleting retry records that should have stayed in existence.
5259 29. eximstats updated to version 1.35
5260 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5261 1.35 - bugfix such that pie charts by volume are generated correctly
5263 30. The SPA authentication driver was not abandoning authentication and moving
5264 on to the next authenticator when an expansion was forced to fail,
5265 contradicting the general specification for all authenticators. Instead it
5266 was generating a temporary error. It now behaves as specified.
5268 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5269 (the order specifies the preference for clients). The order is now AES256,
5270 AES128, 3DES, ARCFOUR128.
5272 31. Small patch to Sieve code - explicitly set From: when generating an
5275 32. Exim crashed if a remote delivery caused a very long error message to be
5276 recorded - for instance if somebody sent an entire SpamAssassin report back
5277 as a large number of 550 error lines. This bug was coincidentally fixed by
5278 increasing the size of one of Exim's internal buffers (big_buffer) that
5279 happened as part of the Exiscan merge. However, to be on the safe side, I
5280 have made the code more robust (and fixed the comments that describe what
5283 33. Some experimental protocols are using DNS PTR records for new purposes. The
5284 keys for these records are domain names, not reversed IP addresses. The
5285 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5286 leaves it alone. Component reversal etc. now happens only for IP addresses.
5289 34. The host_aton() function is supposed to be passed a string that is known
5290 to be a valid IP address. However, in the case of IPv6 addresses, it was
5291 not checking this. This is a hostage to fortune. Exim now panics and dies
5292 if the condition is not met. A case was found where this could be provoked
5293 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5294 components; fortuitously, this particular loophole had already been fixed
5295 by change 4.50/55 or 4.44/33 above.
5297 If there are any other similar loopholes, the new check in host_aton()
5298 itself should stop them being exploited. The report I received stated that
5299 data on the command line could provoke the exploit when Exim was running as
5300 exim, but did not say which command line option was involved. All I could
5301 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5302 running as the user.
5305 35. There was a buffer overflow vulnerability in the SPA authentication code
5306 (which came originally from the Samba project). I have added a test to the
5307 spa_base64_to_bits() function which I hope fixes it.
5310 36. The daemon start-up calls getloadavg() while still root for those OS that
5311 need the first call to be done as root, but it missed one case: when
5312 deliver_queue_load_max is set with deliver_drop_privilege. This is
5313 necessary for the benefit of the queue runner, because there is no re-exec
5314 when deliver_drop_privilege is set.
5316 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5317 was in use, was not putting the data itself into the right store pool;
5318 consequently, it could be overwritten for a subsequent message in the same
5319 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5322 38. Sometimes the final signoff response after QUIT could fail to get
5323 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5324 < 0 before doing a fflush(). This bug looks as though it goes back to the
5325 introduction of TLS in release 3.20, but "sometimes" must have been rare
5326 because the tests only now provoked it.
5332 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5333 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5334 Now it returns an integer. A number of calls were still expecting a T/F
5335 return. Fortuitously, in all cases, the tests worked in OK situations,
5336 which is the norm. However, things would have gone wrong on any write
5337 failures on the smtp file descriptor. This function is used when sending
5338 messages over SMTP and also when doing verify callouts.
5340 2. When Exim is called to do synchronous delivery of a locally submitted
5341 message (the -odf or -odi options), it no longer closes stderr before doing
5344 3. Implemented the mua_wrapper option.
5346 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5348 5. Implemented the functions header_remove(), header_testname(),
5349 header_add_at_position(), and receive_remove_recipient(), and exported them
5352 6. If an ACL "warn" statement specified the addition of headers, Exim already
5353 inserted X-ACL-Warn: at the start if there was no header name. However, it
5354 was not making this test for the second and subsequent header lines if
5355 there were newlines in the string. This meant that an invalid header could
5356 be inserted if Exim was badly configured.
5358 7. Allow an ACL "warn" statement to add header lines at the start or after all
5359 the Received: headers, as well as at the end.
5361 8. Added the rcpt_4xx retry error code.
5363 9. Added postmaster_mailfrom=xxx to callout verification option.
5365 10. Added mailfrom=xxxx to the callout verification option, for verify=
5368 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5369 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5371 12. Inserted some casts to stop certain compilers warning when using pointer
5372 differences as field lengths or precisions in printf-type calls (mostly
5373 affecting debugging statements).
5375 13. Added optional readline() support for -be (dynamically loaded).
5377 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5378 same clock tick as a message's arrival, so that its received time was the
5379 same as the "first fail" time on the retry record, and that message
5380 remained on the queue past the ultimate address timeout, every queue runner
5381 would try a delivery (because it was past the ultimate address timeout) but
5382 after another failure, the ultimate address timeout, which should have then
5383 bounced the address, did not kick in. This was a "< instead of <=" error;
5384 in most cases the first failure would have been in the next clock tick
5385 after the received time, and all would be well.
5387 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5388 being recognized when the domain list was tested by the match_domain
5389 condition in an expansion string.
5391 16. Added the ${str2b64: operator.
5393 17. Exim was always calling setrlimit() to set a large limit for the number of
5394 processes, without checking whether the existing limit was already
5395 adequate. (It did check for the limit on file descriptors.) Furthermore,
5396 errors from getrlimit() and setrlimit() were being ignored. Now they are
5397 logged to the main and panic logs, but Exim does carry on, to try to do its
5398 job under whatever limits there are.
5400 18. Imported PCRE 5.0.
5402 19. Trivial typo in log message " temporarily refused connection" (the leading
5405 20. If the log selector return_path_on_delivery was set and an address was
5406 redirected to /dev/null, the delivery process crashed because it assumed
5407 that a return path would always be set for a "successful" delivery. In this
5408 case, the whole delivery is bypassed as an optimization, and therefore no
5411 21. Internal re-arrangement: the function for sending a challenge and reading
5412 a response while authentication was assuming a zero-terminated challenge
5413 string. It's now changed to take a pointer and a length, to allow for
5414 binary data in such strings.
5416 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5418 23. Exim was not respecting finduser_retries when seeking the login of the
5419 uid under which it was called; it was always trying 10 times. (The default
5420 setting of finduser_retries is zero.) Also, it was sleeping after the final
5421 failure, which is pointless.
5423 24. Implemented tls_on_connect_ports.
5425 25. Implemented acl_smtp_predata.
5427 26. If the domain in control=submission is set empty, Exim assumes that the
5428 authenticated id is a complete email address when it generates From: or
5429 Sender: header lines.
5431 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5432 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5433 chown and chgrp in /bin and hostname in /usr/bin.
5435 28. Exim was keeping the "process log" file open after each use, just as it
5436 does for the main log. This opens the possibility of it remaining open for
5437 long periods when the USR1 signal hits a daemon. Occasional processlog
5438 errors were reported, that could have been caused by this. Anyway, it seems
5439 much more sensible not to leave this file open at all, so that is what now
5442 29. The long-running daemon process does not normally write to the log once it
5443 has entered its main loop, and it closes the log before doing so. This is
5444 so that log files can straightforwardly be renamed and moved. However,
5445 there are a couple of unusual error situations where the daemon does write
5446 log entries, and I had neglected to close the log afterwards.
5448 30. The text of an SMTP error response that was received during a remote
5449 delivery was being truncated at 512 bytes. This is too short for some of
5450 the long messages that one sometimes sees. I've increased the limit to
5453 31. It is now possible to make retry rules that apply only when a message has a
5454 specific sender, in particular, an empty sender.
5456 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5457 it possible to be selective about when SMTP synchronization is enforced.
5459 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5461 32. Implemented hosts_connection_nolog.
5463 33. Added an ACL for QUIT.
5465 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5468 35. Added mailbox_size and mailbox_filecount to appendfile.
5470 36. Added control = no_multiline_responses to ACLs.
5472 37. There was a bug in the logic of the code that waits for the clock to tick
5473 in the case where the clock went backwards by a substantial amount such
5474 that the microsecond fraction of "now" was more than the microsecond
5475 fraction of "then" (but the whole seconds number was less).
5477 38. Added support for the libradius Radius client library this is found on
5478 FreeBSD (previously only the radiusclient library was supported).
5484 1. When certain lookups returned multiple values in the form name=value, the
5485 quoting of the values was not always being done properly. Specifically:
5486 (a) If the value started with a double quote, but contained no whitespace,
5488 (b) If the value contained whitespace other than a space character (i.e.
5489 tabs or newlines or carriage returns) it was not quoted.
5490 This fix has been applied to the mysql and pgsql lookups by writing a
5491 separate quoting function and calling it from the lookup code. The fix
5492 should probably also be applied to nisplus, ibase and oracle lookups, but
5493 since I cannot test any of those, I have not disturbed their existing code.
5495 2. A hit in the callout cache for a specific address caused a log line with no
5496 reason for rejecting RCPT. Now it says "Previous (cached) callout
5497 verification failure".
5499 3. There was an off-by-one bug in the queryprogram router. An over-long
5500 return line was truncated at 256 instead of 255 characters, thereby
5501 overflowing its buffer with the terminating zero. As well as fixing this, I
5502 have increased the buffer size to 1024 (and made a note to document this).
5504 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5505 when Exim is waiting for an SMTP response from a remote server, Exim
5506 restarts its select() call on the socket, thereby resetting its timeout.
5507 This is not a problem when such interrupts are rare. Somebody set up a cron
5508 job to run exiwhat every 2 minutes, which is less than the normal select()
5509 timeout (5 or 10 minutes). This meant that the select() timeout never
5510 kicked in because it was always reset. I have fixed this by comparing the
5511 time when an interrupt arrives with the time at the start of the first call
5512 to select(). If more time than the timeout has elapsed, the interrupt is
5513 treated as a timeout.
5515 5. Some internal re-factoring in preparation for the addition of Sieve
5516 extensions (by MH). In particular, the "personal" test is moved to a
5517 separate function, and given an option for scanning Cc: and Bcc: (which is
5518 not set for Exim filters).
5520 6. When Exim created an email address using the login of the caller as the
5521 local part (e.g. when creating a From: or Sender: header line), it was not
5522 quoting the local part when it contained special characters such as @.
5524 7. Installed new OpenBSD configuration files.
5526 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5527 try to make them clearer.
5529 9. Callout options, other than the timeout value, were being ignored when
5530 verifying sender addresses in header lines. For example, when using
5532 verify = header_sender/callout=no_cache
5534 the cache was (incorrectly) being used.
5536 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5537 only the Cygwin environment.
5539 11. When return_path_on_delivery was set as a log selector, if different remote
5540 addresses in the same message used different return paths and parallel
5541 remote delivery occurred, the wrong values would sometimes be logged.
5542 (Whenever a remote delivery process finished, the return path value from
5543 the most recently started remote delivery process was logged.)
5545 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5546 lines when AUTH and/or TLS are in use. This is the "received protocol"
5547 field. Exim used to use "asmtp" for authenticated SMTP, without any
5548 indication (in the protocol name) for TLS use. Now it follows the RFC and
5549 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5550 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5551 names appear in log lines as well as in Received: header lines.
5553 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5554 extensions, and comparison tests, and to fix some bugs.
5556 14. Changes to the "personal" filter test:
5558 (1) The test was buggy in that it was just doing the equivalent of
5559 "contains" tests on header lines. For example, if a user's address was
5560 anne@some.where, the "personal" test would incorrectly be true for
5562 To: susanne@some.where
5564 This test is now done by extracting each address from the header in turn,
5565 and checking the entire address. Other tests that are part of "personal"
5566 are now done using regular expressions (for example, to check local parts
5567 of addresses in From: header lines).
5569 (2) The list of non-personal local parts in From: addresses has been
5570 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5571 taken from the Sieve specification recommendations.
5573 (3) If the message contains any header line starting with "List-" it is
5574 treated as non-personal.
5576 (4) The test for "circular" in the Subject: header line has been removed
5577 because it now seems ill-conceived.
5579 15. Minor typos in src/EDITME comments corrected.
5581 16. Installed latest exipick from John Jetmore.
5583 17. If headers_add on a router specified a text string that was too long for
5584 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5585 of string_sprintf() is now avoided.
5587 18. $message_body_size was not set (it was always zero) when running the DATA
5588 ACL and the local_scan() function.
5590 19. For the "mail" command in an Exim filter, no default was being set for
5591 the once_repeat time, causing a random time value to be used if "once" was
5592 specified. (If the value happened to be <= 0, no repeat happened.) The
5593 default is now 0s, meaning "never repeat". The "vacation" command was OK
5594 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5595 (I found it when inspecting the code).
5597 20. There is now an overall timeout for performing a callout verification. It
5598 defaults to 4 times the callout timeout, which applies to individual SMTP
5599 commands during the callout. The overall timeout applies when there is more
5600 than one host that can be tried. The timeout is checked before trying the
5601 next host. This prevents very long delays if there are a large number of
5602 hosts and all are timing out (e.g. when the network connections are timing
5603 out). The value of the overall timeout can be changed by specifying an
5604 additional sub-option for "callout", called "maxwait". For example:
5606 verify = sender/callout=5s,maxwait=20s
5608 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5609 to the end before writing, but this should make it even safer).
5611 22. Exim was forgetting that it had advertised PIPELINING for the second and
5612 subsequent messages on an SMTP connection. It was also not resetting its
5613 memory on STARTTLS and an internal HELO.
5615 23. When Exim logs an SMTP synchronization error within a session, it now
5616 records whether PIPELINING has been advertised or not.
5618 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5619 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5620 rather than long int.
5622 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5624 26. Added the never_mail option to autoreply.
5630 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5631 crash if the getsockname() call failed; this can happen if a connection is
5632 closed very soon after it is established. The problem was simply in the
5633 order in which certain operations were done, causing Exim to try to write
5634 to the SMTP stream before it had set up the file descriptor. The bug has
5635 been fixed by making things happen in the correct order.
5641 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5642 before the connection was closed, thus losing the rejection response.
5644 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5645 some early Solaris releases, but causes trouble in current releases where
5646 socklen_t is defined.
5648 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5651 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5654 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5655 that releasing the top of it at the end releases what was used for sub-
5656 expansions (unless the block got too big). However, discard this block if
5657 the first thing is a variable or header, so that we can use its block when
5658 it is dynamic (useful for very large $message_headers, for example).
5660 6. Lookups now cache *every* query, not just the most recent. A new, separate
5661 store pool is used for this. It can be recovered when all lookup caches are
5662 flushed. Lookups now release memory at the end of their result strings.
5663 This has involved some general refactoring of the lookup sources.
5665 7. Some code has been added to the store_xxx() functions to reduce the amount
5666 of flapping under certain conditions.
5668 8. log_incoming_interface used to affect only the <= reception log lines. Now
5669 it causes the local interface and port to be added to several more SMTP log
5670 lines, for example "SMTP connection from", and rejection lines.
5672 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5674 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5676 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5677 could be overwritten at the end of the current message (or the start of a
5678 new message if it was set in a HELO ACL). The value is now preserved for
5679 the duration of the SMTP connection.
5681 12. If a transport had a headers_rewrite setting, and a matching header line
5682 contained an unqualified address, that address was qualified, even if it
5683 did not match any rewriting rules. The underlying bug was that the values
5684 of the flags that permit the existence of unqualified sender and recipient
5685 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5686 non-local messages, and by -bnq for local messages) were not being
5687 preserved with the message after it was received.
5689 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5690 "next input=" as part of the text comprising the host identity instead of
5691 the correct text. The code was using the same buffer for two different
5692 strings. However, depending on which order the printing function evaluated
5693 its arguments, the bug did not always show up. Under Linux, for example, my
5694 test suite worked just fine.
5696 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5697 doesn't work with some older versions of Perl. It has been changed to "my",
5698 which in any case is probably the better facility to use.
5700 15. A really picky compiler found some instances of statements for creating
5701 error messages that either had too many or two few arguments for the format
5704 16. The size of the buffer for calls to the DNS resolver has been increased
5705 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5706 for addresses that have a lot of PTR records. This alleviates a problem; it
5707 does not fully solve it.
5709 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5710 buffer now truncates the list and logs the incident, which is the same
5711 action as happens when Exim is looking up a host name and its aliases.
5712 Previously in this situation something unpredictable would happen;
5713 sometimes it was "internal error: store_reset failed".
5715 18. If a server dropped the connection unexpectedly when an Exim client was
5716 using GnuTLS and trying to read a response, the client delivery process
5717 crashed while trying to generate an error log message.
5719 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5720 string, for example:
5722 warn message = H1: something\nH2: something
5724 the text was added as a single header line from Exim's point of view
5725 though it ended up OK in the delivered message. However, searching for the
5726 second and subsequent header lines using $h_h2: did not work. This has been
5727 fixed. Similarly, if a system filter added multiple headers in this way,
5728 the routers could not see them.
5730 20. Expanded the error message when iplsearch is called with an invalid key to
5731 suggest using net-iplsearch in a host list.
5733 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5734 ACLs are no longer actually imposed (and a message to that effect is
5737 22. If a "gecos" field in a passwd entry contained escaped characters, in
5738 particular, if it contained a \" sequence, Exim got it wrong when building
5739 a From: or a Sender: header from that name. A second bug also caused
5740 incorrect handling when an unquoted " was present following a character
5741 that needed quoting.
5743 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5744 was not being matched caselessly.
5746 24. Arranged for all hyphens in the exim.8 source to be escaped with
5749 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5750 properly. Recipient callout cache records were still being keyed to include
5751 the sender, even when use_sender was set false. This led to far more
5752 callouts that were necessary. The sender is no longer included in the key
5753 when use_sender is false.
5755 26. Added "control = submission" modifier to ACLs.
5757 27. Added the ${base62d: operator to decode base 62 numbers.
5759 28. dnsdb lookups can now access SRV records.
5761 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5762 the configuration file.
5764 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5765 (-v) mode. This makes the output for a verbose queue run more intelligible.
5767 31. Added a use_postmaster feature to recipient callouts.
5769 32. Added the $body_zerocount variable, containing the number of binary zero
5770 bytes in the message body.
5772 33. The time of last modification of the "new" subdirectory is now used as the
5773 "mailbox time last read" when there is a quota error for a maildir
5776 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5778 35. Added +ignore_unknown as a special item in host lists.
5780 36. Code for decoding IPv6 addresses in host lists is now included, even if
5781 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5782 address was recognized as an IP address, but was then not correctly decoded
5783 into binary, causing unexpected and incorrect effects when compared with
5790 1. Very minor rewording of debugging text in manualroute to say "list of
5791 hosts" instead of "hostlist".
5793 2. If verify=header_syntax was set, and a header line with an unqualified
5794 address (no domain) and a large number of spaces between the end of the
5795 name and the colon was received, the reception process suffered a buffer
5796 overflow, and (when I tested it) crashed. This was caused by some obsolete
5797 code that should have been removed. The fix is to remove it!
5799 3. When running in the test harness, delay a bit after writing a bounce
5800 message to get a bit more predictability in the log output.
5802 4. Added a call to search_tidyup() just before forking a reception process. In
5803 theory, someone could use a lookup in the expansion of smtp_accept_max_
5804 per_host which, without the tidyup, could leave open a database connection.
5806 5. Added the variables $recipient_data and $sender_data which get set from a
5807 lookup success in an ACL "recipients" or "senders" condition, or a router
5808 "senders" option, similar to $domain_data and $local_part_data.
5810 6. Moved the writing of debug_print from before to after the "senders" test
5813 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5814 problems for message scanning, either using a data ACL, or using
5815 local_scan() because the Received: header was not generated till after they
5816 were called (in order to set the time as the time of reception completion).
5817 I have revised the way this works. The header is now generated after the
5818 body is received, but before the ACL or local_scan() are called. After they
5819 are run, the timestamp in the header is updated.
5825 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5826 before starting a queue runner without re-exec. This happened only when
5827 deliver_drop_privilege was set or when the Exim user was set to root. The
5828 effect of the bug was that timeouts during subsequent deliveries caused
5829 crashes instead of being properly handled. The handler is now left at its
5830 default (and expected) setting.
5832 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5833 message, again when deliver_drop_privilege is set or Exim is run as root.
5834 The bug described in (1) was not present in this case, but the tidying up
5835 of the other signals was missing. I have made the two cases consistent.
5837 3. The ignore_target_hosts setting on a manualroute router was being ignored
5838 for hosts that were looked up using the /MX notation.
5840 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5843 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5844 operated on the sender address. After changing the $sender_address to <>
5845 for the sender address verify, Exim was re-instated it as the original
5846 (before rewriting) address, but remembering that it had rewritten it, so it
5847 wasn't rewriting it again. This bug also had the effect of breaking the
5848 sender address verification caching when the sender address was rewritten.
5850 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5851 This has been changed so that if the ip literal address matches
5852 ignore_target_hosts, the router declines.
5854 7. Added expansion conditions match_domain, match_address, and match_local_
5855 part (NOT match_host).
5857 8. The placeholder for the Received: header didn't have a length field set.
5859 9. Added code to Exim itself and to exim_lock to test for a specific race
5860 condition that could lead to file corruption when using MBX delivery. The
5861 issue is with the lockfile that is created in /tmp. If this file is removed
5862 after a process has opened it but before that process has acquired a lock,
5863 there is the potential for a second process to recreate the file and also
5864 acquire a lock. This could lead to two Exim processes writing to the file
5865 at the same time. The added code performs the same test as UW imapd; it
5866 checks after acquiring the lock that its file descriptor still refers to
5867 the same named file.
5869 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5870 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5873 11. Added the smtp_active_hostname option. If used, this will typically be made
5874 to depend on the incoming interface address. Because $interface_address is
5875 not set up until the daemon has forked a reception process, error responses
5876 that can happen earlier (such as "too many connections") no longer contain
5879 12. If an expansion in a condition on a "warn" statement fails because a lookup
5880 defers, the "warn" statement is abandoned, and the next ACL statement is
5881 processed. Previously this caused the whole ACL to be aborted.
5883 13. Added the iplsearch lookup type.
5885 14. Added ident_timeout as a log selector.
5887 15. Added tls_certificate_verified as a log selector.
5889 16. Added a global option tls_require_ciphers (compare the smtp transport
5890 option of the same name). This controls incoming TLS connections.
5892 17. I finally figured out how to make tls_require_ciphers do a similar thing
5893 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5894 before starting the TLS session.
5896 18. Tabs are now shown as \t in -bP output.
5898 19. If the log selector return_path_on_delivery was set, Exim crashed when
5899 bouncing a message because it had too many Received: header lines.
5901 20. If two routers both had headers_remove settings, and the first one included
5902 a superfluous trailing colon, the final name in the first list and the
5903 first name in the second list were incorrectly joined into one item (with a
5904 colon in the middle).
5910 1. Added -C and -D options to the exinext utility, mainly to make it easier
5911 to include in the automated testing, but these could be helpful when
5912 multiple configurations are in use.
5914 2. The exinext utility was not formatting the output nicely when there was
5915 an alternate port involved in the retry record key, nor when there was a
5916 message id as well (for retries that were specific to a specific message
5917 and a specific host). It was also confused by IPv6 addresses, because of
5918 the additional colons they contain. I have fixed the IPv4 problem, and
5919 patched it up to do a reasonable job for IPv6.
5921 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5922 delivery, the log line now contains "pipelined" if PIPELINING was used.
5924 4. An SMTP transport process used to panic and die if the bind() call to set
5925 an explicit outgoing interface failed. This has been changed; it is now
5926 treated in the same way as a connect() failure.
5928 5. A reference to $sender_host_name in the part of a conditional expansion
5929 that was being skipped was still causing a DNS lookup. This no longer
5932 6. The def: expansion condition was not recognizing references to header lines
5933 that used bh_ and bheader_.
5935 7. Added the _cache feature to named lists.
5937 8. The code for checking quota_filecount in the appendfile transport was
5938 allowing one more file than it should have been.
5940 9. For compatibility with Sendmail, the command line option
5948 and sets the incoming protocol and host name (for trusted callers). The
5949 host name and its colon can be omitted when only the protocol is to be set.
5950 Note the Exim already has two private options, -pd and -ps, that refer to
5951 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5952 "s", but I don't think that's a major issue.
5954 10. A number of refactoring changes to the code, none of which should affect
5957 (a) The number of logging options was getting close to filling up the
5958 32-bit word that was used as a bit map. I have split them into two classes:
5959 those that are passed in the argument to log_write(), and those that are
5960 only ever tested independently outside of that function. These are now in
5961 separate 32-bit words, so there is plenty of room for expansion again.
5962 There is no change in the user interface or the logging behaviour.
5964 (b) When building, for example, log lines, the code previously used a
5965 macro that called string_cat() twice, in order to add two strings. This is
5966 not really sufficiently general. Furthermore, there was one instance where
5967 it was actually wrong because one of the argument was used twice, and in
5968 one call a function was used. (As it happened, calling the function twice
5969 did not affect the overall behaviour.) The macro has been replaced by a
5970 function that can join an arbitrary number of extra strings onto a growing
5973 (c) The code for expansion conditions now uses a table and a binary chop
5974 instead of a serial search (which was left over from when there were very
5975 few conditions). Also, it now recognizes conditions like "pam" even when
5976 the relevant support is not compiled in: a suitably worded error message is
5977 given if an attempt is made to use such a condition.
5979 11. Added ${time_interval:xxxxx}.
5981 12. A bug was causing one of the ddress fields not to be passed back correctly
5982 from remote delivery subprocesses. The field in question was not being
5983 subsequently used, so this caused to problems in practice.
5985 13. Added new log selectors queue_time and deliver_time.
5987 14. Might have fixed a bug in maildirsizefile handling that threw up
5988 "unexpected character" debug warnings, and recalculated the data
5989 unnecessarily. In any case, I expanded the warning message to give more
5992 15. Added the message "Restricted characters in address" to the statements in
5993 the default ACL that block characters like @ and % in local parts.
5995 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5996 Three changes have been made:
5998 (a) There was a serious bug; a negative response to MAIL caused the whole
5999 recipient domain to be cached as invalid, thereby blocking all messages
6000 to all local parts at the same domain, from all senders. This bug has
6001 been fixed. The domain is no longer cached after a negative response to
6002 MAIL if the sender used is not empty.
6004 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6007 (c) A new callout option, "use_sender" has been added for people who want
6008 the modified behaviour.
6014 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6017 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6018 indeed breaks things for older releases.
6020 3. Added additional logging to the case where there is a problem reading data
6021 from a filter that is running in a subprocess using a pipe, in order to
6022 try to track down a specific problem.
6024 4. Testing facility fudge: when running in the test harness and attempting
6025 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6026 getting "No route to host". Convert this to a timeout.
6028 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6031 6. Some OS don't have socklen_t but use size_t instead. This affects the
6032 fifth argument of getsockopt() amongst other things. This is now
6033 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6034 can be set for individual OS. I have set it for SunOS5, OSF1, and
6035 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6036 some earlier ones do not.
6038 7. Change 4.30/15 was not doing the test caselessly.
6040 8. The standard form for an IPv6 address literal was being rejected by address
6041 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6042 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6043 this, as well as the form without the "IPv6" on the front (but only when
6044 address literals are enabled, of course).
6046 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6048 10. Exim crashed if a message with an empty sender address specified by -f
6049 encountered a router with an errors_to setting. This could be provoked only
6050 by a command such as
6054 where an empty string was supplied; "<>" did not hit this bug.
6056 11. Installed PCRE release 4.5.
6058 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6059 remained set. It is now erased.
6061 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6062 times from message ids (which are base 36 rather than the normal 62).
6064 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6065 were being counted as actual protocol errors, and logged if the log
6066 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6067 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6068 and DATA following a set of rejected RCPTs do not count as protocol errors.
6069 In other words, Exim assumes they were pipelined, though this may not
6070 actually be the case. Of course, in all cases the client gets an
6071 appropriate error code.
6073 15. If a lookup fails in an ACL condition, a message about the failure may
6074 be available; it is used if testing the ACL cannot continue, because most
6075 such messages specify what the cause of the deferral is. However, some
6076 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6077 that caused an old message to be retained and used if a later statement
6078 caused a defer, replacing the real cause of the deferral.
6080 16. If an IP address had so many PTR records that the DNS lookup buffer
6081 was not large enough to hold them, Exim could crash while trying to process
6082 the truncated data. It now detects and logs this case.
6084 17. Further to 4.21/58, another change has been made: if (and only if) the
6085 first line of a message (the first header line) ends with CRLF, a bare LF
6086 in a subsequent header line has a space inserted after it, so as not to
6087 terminate the header.
6089 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6090 unnecessarily, used atoi() instead of strtol(), and didn't check the
6091 termination when getting file sizes from file names by regex.
6093 19. Completely re-implemented the support for maildirsize files, in the light
6094 of a number of problems with the previous contributed implementation
6095 (4.30/29). In particular:
6097 . If the quota is zero, the maildirsize file is maintained, but no quota is
6100 . If the maildir directory does not exist, it is created before any attempt
6101 to write a maildirsize file.
6103 . The quota value in the file is just a cache; if the quota is changed in
6104 the transport, the new value overrides.
6106 . A regular expression is available for excluding directories from the
6109 20. The autoreply transport checks the characters in options that define the
6110 message's headers; it allows continued headers, but it was checking with
6111 isspace() after an embedded newline instead of explicitly looking for a
6114 21. If all the "regular" hosts to which an address was routed had passed their
6115 expiry times, and had not reached their retry times, the address was
6116 bounced, even if fallback hosts were defined. Now Exim should go on to try
6119 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6120 equivalent code in the SMTP transport. Some hosts send humungous responses
6121 to HELO/EHLO, more than 1024 it seems.
6123 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6124 gives compiler warnings in some environments. I've now done it "properly",
6127 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6128 (because of problems with the built-in one) was declared to return uschar *
6129 instead of char *, causing compiler failure.
6131 25. Fixed a file descriptor leak when processing alias/forward files.
6133 26. Fixed a minor format string issue in dbfn.c.
6135 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6137 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6138 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6141 29. When a sender address is verified, it is cached, to save repeating the test
6142 when there is more than one recipient in a message. However, when the
6143 verification involves a callout, it is possible for different callout
6144 options to be set for different recipients. It is too complicated to keep
6145 track of this in the cache, so now Exim always runs a verification when a
6146 callout is required, relying on the callout cache for the optimization.
6147 The overhead is duplication of the address routing, but this should not be
6150 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6151 to be verified with callout=postmaster, and the main callout worked but the
6152 postmaster check failed, the verification correctly failed. However, if a
6153 subsequent RCPT command asked for sender verification *without* the
6154 postmaster check, incorrect caching caused this verification also to fail,
6157 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6158 it was not caching the DNS options (qualify_single, search_parents) that
6159 were used when the lookup failed. A subsequent lookup with different
6160 options therefore always gave the same answer, though there were cases
6161 where it should not have. (Example: a "domains = !$mx_any" option on a
6162 dnslookup router: the "domains" option is always processed without any
6163 widening, but the router might have qualify_single set.) Now Exim uses the
6164 cached value only when the same options are set.
6166 32. Added John Jetmore's "exipick" utility to the distribution.
6168 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6169 than a timeout (e.g. a certificate is required, and is not provided), an
6170 Exim server now closes the connection immediately. Previously it waited for
6171 the client to close - but if the client is SSL, it seems that they each
6172 wait for each other, leading to a delay before one of them times out.
6174 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6175 maintained 0.8.x compatibility because I don't think many are using it, and
6176 it is clearly obsolete.
6178 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6181 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6182 client certificate was expired. A simple patch fixes this, though I don't
6183 understand the full logic of why the verify callback is called multiple
6186 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6187 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6188 which causes problems with some clients (such as the Certicom SSL Plus
6189 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6190 disables the coutermeasure allowing Eudora to connect."
6192 38. Exim was not checking that a write() to a log file succeeded. This could
6193 lead to Bad Things if a log got too big, in particular if it hit a file
6194 size limit. Exim now panics and dies if it cannot write to a log file, just
6195 as it does if it cannot open a log file.
6197 39. Modified OS/Makefile-Linux so that it now contains
6199 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6201 The two -D definitions ensure that Exim is compiled with large file
6202 support, which makes it possible to handle log files that are bigger than
6205 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6206 instance) a domain was checked against a named list that involved a lookup,
6207 causing $domain_data to be set, then another domain was checked against the
6208 same list, then the first domain was re-checked, the value of $domain_data
6209 after the final check could be wrong. In particular, if the second check
6210 failed, it could be set empty. This bug probably also applied to
6213 41. The strip_trailing_dot option was not being applied to the address given
6214 with the -f command-line option.
6216 42. The code for reading a message's header from the spool was incrementing
6217 $received_count, but never initializing it. This meant that the value was
6218 incorrect (doubled) while delivering a message in the same process in which
6219 it was received. In the most common configuration of Exim, this never
6220 happens - a fresh exec is done - but it can happen when
6221 deliver_drop_privilege is set.
6223 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6224 it now includes up to 150 characters of the unexpected data in the log
6227 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6228 and building data strings. The size of both of these buffers was 10 000
6229 bytes - far larger than anybody would *ever* want, thought I. Needless to
6230 say, somebody hit the limit. I have increased the maximum line length to
6231 20 000 and the maximum data length of concatenated lines to 100 000. I have
6232 also fixed two bugs, because there was no checking on these buffers. Tsk,
6233 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6234 buffer is too small.
6236 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6237 lsearch lookups. Now it does.
6239 46. When parsing a route_list item in a manualroute router, a fixed-length
6240 buffer was used for the list of hosts. I made this 1024 bytes long,
6241 thinking that nobody would ever have a list of hosts that long. Wrong.
6242 Somebody had a whole pile of complicated expansion conditions, and the
6243 string was silently truncated, leading to an expansion error. It turns out
6244 that it is easier to change to an unlimited length (owing to other changes
6245 that have happened since this code was originally written) than to build
6246 structure for giving a limitation error. The length of the item that
6247 expands into the list of hosts is now unlimited.
6249 47. The lsearch lookup could not handle data where the length of text line was
6250 more than 4095 characters. Such lines were truncated, leading to shortened
6251 data being returned. It should now handle lines of any length.
6253 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6254 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6257 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6258 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6259 no longer happens in comment lines. A list of replacements is now placed
6260 at the head of all of the source files, except those whose only change is
6261 to replace PERL_COMMAND in the very first #! line.
6263 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6264 messages on the queue, with a bottom-up merge sort, using code contributed
6265 by Michael Haardt. This should make operations like -bp somewhat faster on
6266 large queues. It won't affect queue runners, except when queue_run_in_order
6269 51. Installed eximstats 1.31 in the distribution.
6271 52. Added support for SRV lookups to the dnslookup router.
6273 53. If an ACL referred to $message_body or $message_body_end, the value was not
6274 reset for any messages that followed in the same SMTP session.
6276 54. The store-handling optimization for building very long strings was not
6277 differentiating between the different store pools. I don't think this
6278 actually made any difference in practice, but I've tidied it.
6280 55. While running the routers to verify a sender address, $sender_address
6281 was still set to the sender address. This is wrong, because when routing to
6282 send a bounce to the sender, it would be empty. Therefore, I have changed
6283 it so that, while verifying a sender address, $sender_address is set to <>.
6284 (There is no change to what happens when verifying a recipient address.)
6286 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6287 target A or AAAA records (if not already returned) without resetting the
6288 qualify_single or search_parents options of the DNS resolver. These are
6289 inappropriate in this case because the targets of MX and SRV records must
6290 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6291 target that, when qualified, matched something in the local domain. These
6292 two options are now turned off when doing these lookups.
6294 57. It seems that at least some releases of Reiserfs (which does not have the
6295 concept of a fixed number of inodes) returns zero and not -1 for the
6296 number of available inodes. This interacted badly with check_spool_inodes,
6297 which assumed that -1 was the "no such thing" setting. What I have done is
6298 to check that the total number of inodes is greater than zero before doing
6299 the test of how many are available.
6301 58. When a "warn" ACL statement has a log_message modifier, the message is
6302 remembered, and not repeated. This is to avoid a lot of repetition when a
6303 message has many recipients that cause the same warning to be written.
6304 However, Exim was preserving the list of already written lines for an
6305 entire SMTP session, which doesn't seem right. The memory is now reset if a
6306 new message is started.
6308 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6309 debugging output unless log_rewrite was also set.
6311 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6312 of (int)(handle) when we know that handle contains (void *)(-1).
6314 61. The Exim daemon panic-logs an error return when it closes the incoming
6315 connection. However "connection reset by peer" seems to be common, and
6316 isn't really an error worthy of noting specially, so that particular error
6319 62. When Exim is trying to find all the local interfaces, it used to panic and
6320 die if the ioctl to get the interface flags failed. However, it seems that
6321 on at least one OS (Solaris 9) it is possible to have an interface that is
6322 included in the list of interfaces, but for which you get a failure error
6323 for this call. This happens when the interface is not "plumbed" into a
6324 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6325 failure of the "get flags" call assumes that the interface is down.
6327 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6328 makes life easier for people who are doing arithmetic on fields extracted
6329 from dates, where you often get leading zeros that should not be
6330 interpreted as octal.
6332 64. Added qualify_domain to the redirect router, to override the global
6335 65. If a pathologically long header line contained very many addresses (the
6336 report of this problem mentioned 10 000) and each of them was rewritten,
6337 Exim could use up a very large amount of memory. (It kept on making new
6338 copies of the header line as it rewrote, and never released the old ones.)
6339 At the expense of a bit more processing, the header rewriting function has
6340 been changed so that it no longer eats memory in this way.
6342 66. The generation of the Received: header has been moved from the time that a
6343 message starts to be received, to the time that it finishes. The timestamp
6344 in the Received: header should now be very close to that of the <= log
6345 line. There are two side-effects of this change:
6347 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6348 logged header lines no longer include the local Received: line, because
6349 it has not yet been created. The same applies to a copy of the message
6350 that is returned to a non-SMTP sender when a message is rejected.
6352 (b) When a filter file is tested using -bf, no additional Received: header
6353 is added to the test message. After some thought, I decided that this
6356 This change does not affect the value of $received_for. It is still set
6357 after address rewriting, but before local_scan() is called.
6359 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6361 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6362 gave an unhelpful panic error message, and a defer error. I have managed to
6363 change this behaviour so that it now rejects any supplied certificate,
6364 which seems right, as the list of acceptable certificates is empty.
6366 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6367 gave an unhelpful defer error. I have not managed to make this reject any
6368 supplied certificates, but the error message it gives is "no certificate
6369 supplied", which is not helpful.
6371 70. exigrep's output now also includes lines that are not associated with any
6372 message, but which match the given pattern. Implemented by a patch from
6373 Martin Sluka, which also tidied up the Perl a bit.
6375 71. Recipient callout verification, like sender verification, was using <> in
6376 the MAIL FROM command. This isn't really the right thing, since the actual
6377 sender may affect whether the remote host accepts the recipient or not. I
6378 have changed it to use the actual sender in the callout; this means that
6379 the cache record is now keyed on a recipient/sender pair, not just the
6380 recipient address. There doesn't seem to be a real danger of callout loops,
6381 since a callout by the remote host to check the sender would use <>.
6382 [SEE ABOVE: changed after hitting problems.]
6384 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6385 temporary errors. However, in the case of such a code being given after
6386 the end of a data transmission (i.e. after ".") Exim was failing to write
6387 a retry record for the message. (Yes, there was some broken host that was
6388 actually sending 8xx at this point.)
6390 73. An unknown lookup type in a host list could cause Exim to panic-die when
6391 the list was checked. (An example that provoked this was putting <; in the
6392 middle of a list instead of at the start.) If this happened during a DATA
6393 ACL check, a -D file could be left lying around. This kind of configuration
6394 error no longer causes Exim to die; instead it causes a defer error. The
6395 incident is still logged to the main and panic logs.
6397 74. Buglet left over from Exim 3 conversion. The message "too many messages
6398 in one connection" was written to the rejectlog but not the mainlog, except
6399 when address rewriting (yes!) was being logged.
6401 75. Added write_rejectlog option.
6403 76. When a system filter was run not as root (that is, when system_filter_user
6404 was set), the values of the $n variables were not being returned to the
6405 main process; thus, they were not subsequently available in the $sn
6408 77. Added +return_path_on_delivery log selector.
6410 78. A connection timeout was being treated differently from recipients deferred
6411 when testing hosts_max_try with a message that was older than the host's
6412 retry timeout. (The host should not be counted, thus allowing all hosts to
6413 be tried at least once before bouncing.) This may have been the cause of an
6414 occasionally reported bug whereby a message would remain on the queue
6415 longer than the retry timeout, but would be bounced if a delivery was
6416 forced. I say "may" because I never totally pinned down the problem;
6417 setting up timeout/retry tests is difficult. See also the next item.
6419 79. The ultimate address timeout was not being applied to errors that involved
6420 a combination of host plus message (for example, a timeout on a MAIL
6421 command). When an address resolved to a number of possible hosts, and they
6422 were not all tried for each delivery (e.g. because of hosts_max_try), a
6423 message could remain on the queue longer than the retry timeout.
6425 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6428 81. Fixed an obscure SMTP outgoing bug which required at least the following
6429 conditions: (a) there was another message waiting for the same server;
6430 (b) the server returned 5xx to all RCPT commands in the first message so
6431 that the message was not completed; (c) the server dropped the connection
6432 or gave a negative response to the RSET that Exim sends to abort the
6433 transaction. The observed case was a dropped connection after DATA that had
6434 been sent in pipelining mode. That is, the server had advertised PIPELINING
6435 but was not implementing it correctly. The effect of the bug was incorrect
6436 behaviour, such as trying another host, and this could lead to a crash.
6442 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6443 and daemon.c were passed as pointers to ints; they should have been
6444 pointers to socklen_t variables (which are typically unsigned ints).
6446 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6449 3. Fixed a really odd bug that affected only the testing scheme; patching a
6450 certain fixed string in the binary changed the value of another string that
6451 happened to be identical to the end of the original first string.
6453 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6454 name", it returns that address as the IP address. On some operating
6455 systems (e.g. Solaris), it also passes back the IP address string as the
6456 "host name". However, on others (e.g. Linux), it passes back an empty
6457 string. Exim wasn't checking for this, and was changing the host name to an
6458 empty string, assuming it had been canonicalized.
6460 5. Although rare, it is permitted to have more than one PTR record for a given
6461 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6462 all the names associated with an address, because they do in Solaris.
6463 However, it seems that they do not in Linux for data that comes from the
6464 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6465 I found this out when I moved to a new Linux workstation and tried to run
6466 the Exim test suite.
6468 To get round this problem I have changed the code so that it now does its
6469 own call to the DNS to look up PTR records when searching for a host name.
6470 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6471 addresses that are only in /etc/hosts are still found.
6473 This behaviour is, however, controlled by an option called host_lookup_
6474 order, which defaults to "bydns:byaddr". If people want to use the other
6475 order, or indeed, just use one or the other means of lookup, they can
6476 specify it in this variable.
6478 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6479 some operating systems, this comes back from gethostbyaddr() as an empty
6480 string, and this is what Exim used to test for. However, it seems that in
6481 other systems, "." is yielded. Exim now tests for this case too.
6483 7. The values of check_spool_space and check_log_space are now held internally
6484 as a number of kilobytes instead of an absolute number of bytes. If a
6485 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6486 kilobyte. This means that much larger values can be stored.
6488 8. Exim monitor: an attempt to get the action menu when not actually pointing
6489 at a message produces an empty menu entitled "No message selected". This
6490 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6491 no entries in it ("Shell widget menu has zero width and/or height"). So I
6492 have added a single, blank menu entry in this case.
6494 9. Added ${quote_local_part.
6496 10. MIME decoding is now applied to the contents of Subject: header lines when
6499 11. Now that a reference to $sender_host_address automatically causes a reverse
6500 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6501 host lookup before query-style lookups in lists that might use this
6502 variable. This has therefore been abolished, and the "net-" prefix is no
6503 longer necessary for query-style lookups.
6505 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6506 have been a typo for LFLAGS, so it has been changed.
6508 13. The install script calls Exim with "-C /dev/null" in order to find the
6509 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6510 to be output. However, since Exim outputs its version number before the
6511 error, it didn't break the script. It just looked ugly. I fixed this by
6512 always allowing "-C /dev/null" if the caller is root.
6514 14. Ignore overlarge ACL variable number when reading spool file - insurance
6515 against a later release with more variables having written the file.
6517 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6518 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6519 this, as well as the form without the "IPv6" on the front.
6521 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6522 OS/Makefile-Darwin file.
6524 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6525 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6527 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6528 variable. However, some OS use ints for inodes. I've added cast to long int
6529 to get rid of the compiler warning.
6531 19. I had forgotten to lock out "/../" in configuration file names when
6532 ALT_CONFIG_PREFIX was set.
6534 20. Routers used for verification do not need to specify transports. However,
6535 if such a router generated a host list, and callout was configured, Exim
6536 crashed, because it could not find a port number from the (non-existent)
6537 transport. It now assumes port 25 in this circumstance.
6539 21. Added the -t option to exigrep.
6541 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6542 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6543 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6544 them is set, LOOKUP_LSEARCH is forced.
6546 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6548 24. Added sender and host information to the "rejected by local_scan()" log
6549 line; previously there was no indication of these.
6551 25. Added .include_if_exists.
6553 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6554 Linux. It turns out that not all file systems support this. Apparently some
6555 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6556 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6557 not supported on the file descriptor, is now ignored when Exim is trying to
6558 sync a directory. This applies only to Linux.
6560 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6562 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6563 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6566 29. Added support for maildirsize files from supplied patch (modified a bit).
6568 30. The use of :fail: followed by an empty string could lead Exim to respond to
6569 sender verification failures with (e.g.):
6571 550 Verification failed for <xxx>
6572 550 Sender verify failed
6574 where the first response line was missing the '-' that indicates it is not
6575 the final line of the response.
6577 31. The loop for finding the name of the user that called Exim had a hardwired
6578 limit of 10; it now uses the value of finduser_retries, which is used for
6579 all other user lookups.
6581 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6584 33. Exim was neglecting to zero errno before one call of strtol() when
6585 expanding a string and expecting an integer value. On some systems this
6586 resulted in spurious "integer overflow" errors. Also, it was casting the
6587 result into an int without checking.
6589 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6590 did not work. The code looks as if it has *never* worked, though it appears
6591 to have been documented since at least release 1.62. I have made it work.
6593 35. The "timeout_DNS" error in retry rules, also documented since at least
6594 1.62, also never worked. As it isn't clear exactly what this means, and
6595 clearly it isn't a major issue, I have abolished the feature by treating it
6596 as "timeout", and writing a warning to the main and panic logs.
6598 36. The display of retry rules for -brt wasn't always showing the error code
6601 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6602 timeout_connect_A, timeout_connect_MX.
6604 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6605 to the empty sender.
6607 39. The daemon was not analysing the content of -oX till after it had closed
6608 stderr and disconnected from the controlling terminal. This meant that any
6609 syntax errors were only noted on the panic log, and the return code from
6610 the command was 0. By re-arranging the code a little, I've made the
6611 decoding happen first, so such errors now appear on stderr, and the return
6612 code is 1. However, the actual setting up of the sockets still happens in
6613 the disconnected process, so errors there are still only recorded on the
6616 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6617 connections (as happens on some IP stacks) was logged at start up time as
6618 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6619 it from "IPv6 and IPv4", which means that two separate sockets are being
6622 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6623 says whether AF_INET or AF_INET6 was passed as an argument.
6625 42. Exiwhat output was messed up when time zones were included in log
6628 43. Exiwhat now gives more information about the daemon's listening ports,
6629 and whether -tls-on-connect was used.
6631 44. The "port" option of the smtp transport is now expanded.
6633 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6634 silently ignored. Now an error message is written to the main and panic
6637 46. There's a new ACL modifier called "logwrite" which writes to a log file
6638 as soon as it is encountered.
6640 47. Added $local_user_uid and $local_user_gid at routing time.
6642 48. Exim crashed when trying to verify a sender address that was being
6645 49. Exim was recognizing only a space character after ".include". It now also
6646 recognizes a tab character.
6648 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6649 extracting the relevant information from the specification. The man page no
6650 longer contains scrambled data for the -d option, and I've added a section
6651 at the front about calling Exim under different names.
6653 51. Added "extra_headers" argument to the "mail" command in filter files.
6655 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6658 53. Installed eximstats 1.29.
6660 54. Added transport_filter_timeout as a generic transport option.
6662 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6663 Cc: header lines. This was required by RFC 822, but it not required by RFC
6666 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6667 incoming messages that did not have them. Now it does so only if the
6668 message originates locally, that is, if there is no associated remote host
6669 address. When Resent- header lines are present, this applies to the Resent-
6670 lines rather than the non-Resent- lines.
6672 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6673 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6675 58. Messages for configuration errors now include the name of the main
6676 configuration file - useful now that there may be more than one file in a
6677 list (.included file names were always shown).
6679 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6680 for those rare installations that do not start the daemon as root or run it
6681 setuid root. I've cut out the call to initgroups() if the daemon is not
6684 60. The Exim user and group can now be bound into the binary as text strings
6685 that are looked up at the start of Exim's processing.
6687 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6689 62. Added $mailstore_basename variable.
6691 63. Installed patch to sieve.c from Michael Haardt.
6693 64. When Exim failed to open the panic log after failing to open the main log,
6694 the original message it was trying to log was written to stderr and debug
6695 output, but if they were not available (the usual case in production), it
6696 was lost. Now it is written to syslog before the two lines that record the
6697 failures to open the logs.
6699 65. Users' Exim filters run in subprocesses under the user's uid. It is
6700 possible for a "deliver" command or an alias in a "personal" command to
6701 provoke an address rewrite. If logging of address rewriting is configured,
6702 this fails because the process is not running as root or exim. There may be
6703 a better way of dealing with this, but for the moment (because 4.30 needs
6704 to be released), I have disabled address rewrite logging when running a
6705 filter in a non-root, non-exim process.
6711 1. The buildconfig auxiliary program wasn't quoting the value set for
6712 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6713 not defined. This bug was masked in 4.22 by the effect that was fixed in
6716 2. Some messages that were rejected after a message id was allocated were
6717 shown as "incomplete" by exigrep. It no longer does this for messages that
6718 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6720 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6721 have allow_domain_literals set, the ID did not get logged in the <= line.
6722 Domain literals are now always recognized in Message-ID: header lines.
6724 4. The first argument for a ${extract expansion item is the key name or field
6725 number. Leading and trailing spaces in this item were not being ignored,
6726 causing some misleading effects.
6728 5. When deliver_drop_privilege was set, single queue runner processes started
6729 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6730 same command in the process it spins off) were not dropping privilege.
6732 6. When the daemon running as "exim" started a queue runner, it always
6733 re-executed Exim in the spun-off process. This is a waste of effort when
6734 deliver_drop_privilege is set. The new process now just calls the
6735 queue-runner function directly.
6741 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6744 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6745 ignored. Though the use of -oP was forcing the writing of a pid file, it
6746 was always written to the default place.
6748 3. If the message "no IP address found for host xxxx" is generated during
6749 incoming verification, it is now followed by identification of the incoming
6750 connection (so you can more easily find what provoked it).
6752 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6754 5. Added some features to "harden" Exim a bit more against certain attacks:
6756 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6757 be put in Local/Makefile. This is like the never_users runtime option,
6758 but it cannot be overridden. The default setting is "root".
6760 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6761 prefix string with which any file named in a -C command line option
6764 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6765 is retained for -C and -D only if the caller of Exim is root. Without
6766 it, the exim user may also use -C and -D and retain privilege.
6768 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6769 command line option is disabled.
6771 6. Macro names set by the -D option must start with an upper case letter, just
6772 like macro names defined in the configuration file.
6774 7. Added "dereference=" facility to LDAP.
6776 8. Two instances of the typo "uknown" in the source files are fixed.
6778 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6779 the Configure-Makefile script screwed up while processing it.
6781 10. Incorporated PCRE 4.4.
6783 11. The SMTP synchronization check was not operating right at the start of an
6784 SMTP session. For example, it could not catch a HELO sent before the client
6785 waited for the greeting. There is now a check for outstanding input at the
6786 point when the greeting is written. Because of the duplex, asynchronous
6787 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6788 way, but not yet received, when the check is performed.
6790 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6791 on TCP/IP sockets, because this apparently causes some broken clients to
6794 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6795 unchanged) from the Cygwin maintainer.
6797 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6798 was supported instead of testing for mbx. Effectively a typo.
6800 15. The spa authenticator server code was not checking that the input it
6801 received was valid base64.
6803 16. The debug output line for the "set" modifier in ACLs was not showing the
6804 name of the variable that was being set.
6806 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6808 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6809 lookup on its own. Something else was needed to trigger the lookup. For
6810 example, a match in host_lookup or the need for a host name in a host list.
6811 Now, if $sender_host_name is referenced and the host name has not yet been
6812 looked up, a lookup is performed. If the lookup fails, the variable remains
6813 empty, and $host_lookup_failed is set to "1".
6815 19. Added "eqi" as a case-independent comparison operator.
6817 20. The saslauthd authentication condition could segfault if neither service
6818 nor realm was specified.
6820 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6821 error message that was logged was misleading, and incoming SMTP
6822 connections were dropped. The message is now more accurate, and temporary
6823 errors are given to SMTP connections.
6825 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6826 (except RSET) with a 421 error, until QUIT is received. However, it was
6827 failing to send a response to QUIT.
6829 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6830 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6831 ACL is now run after the helo_try_verify_hosts code.
6833 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6834 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6835 case-independent, so other case variants are also recognized). Apparently
6836 some systems use these upper case variants.
6838 25. If more than two messages were waiting for the same host, and a transport
6839 filter was specified for the transport, Exim sent two messages over the
6840 same TCP/IP connection, and then failed with "socket operation on non-
6841 socket" when it tried to send the third.
6843 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6845 27. The extern definition of crypt16() in expand.c was not being excluded when
6846 the OS had its own crypt16() function.
6848 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6849 as a preferred synonym for return_size_limit, both as an option and as an
6852 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6854 30. Changed the default configuration ACL to relax the local part checking rule
6855 for addresses that are not in any local domains. For these addresses,
6856 slashes and pipe symbols are allowed within local parts, but the sequence
6857 /../ is explicitly forbidden.
6859 31. SPA server authentication was not clearing the challenge buffer before
6862 32. log_message in a "warn" ACL statement was writing to the reject log as
6863 well as to the main log, which contradicts the documentation and doesn't
6864 seem right (because no rejection is happening). So I have stopped it.
6866 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6867 However, I am unable to do any testing of this.
6869 34. Fixed an infelicity in the appendfile transport. When checking directories
6870 for a mailbox, to see if any needed to be created, it was accidentally
6871 using path names with one or more superfluous leading slashes; tracing
6872 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6874 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6875 contents are added to the log line that is written for every discarded
6876 recipient. (Previously a log_message setting was ignored.)
6878 36. The ${quote: operator now quotes the string if it is empty.
6880 37. The install script runs exim in order to find its version number. If for
6881 some reason other than non-existence or emptiness, which it checks, it
6882 could not run './exim', it was installing it with an empty version number,
6883 i.e. as "exim-". This error state is now caught, and the installation is
6886 38. An argument was missing from the function that creates an error message
6887 when Exim fails to connect to the socket for saslauthd authentication.
6888 This could cause Exim to crash, or give a corrupted message.
6890 39. Added isip, isip4, and isip6 to ${if conditions.
6892 40. The ACL variables $acl_xx are now saved with the message, and can be
6893 accessed later in routers, transports, and filters.
6895 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6896 strings in the file are not string-expanded.
6898 42. If a MAIL command specified a SIZE value that was too large to fit into an
6899 int variable, the check against message_size_limit failed. Such values are
6900 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6901 day this will have to be increased, but I don't think I want to be around
6902 when emails are that large.
6909 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6910 iconv() is not standard in FreeBSD.
6912 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6913 IPv6 enabled. The observed symptom was a segmentation fault on return from
6914 the function os_common_find_running_interfaces() in src/os.c.
6916 3. In the check_special_case() function in daemon.c I had used "errno" as an
6917 argument name, which causes warnings on some systems. This was basically a
6918 typo, since it was named "eno" in the comments!
6920 4. The code that waits for the clock to tick (at a resolution of some fraction
6921 of a second) so as to ensure message-id uniqueness was always waiting for
6922 at least one whole tick, when it could have waited for less. [This is
6923 almost certainly not relevant at current processor speeds, where it is
6924 unlikely to ever wait at all. But we try to future-proof.]
6926 5. The function that sleeps for a time interval that includes fractions of a
6927 second contained a race. It did not block SIGALRM between setting the
6928 timer, and suspending (a couple of lines later). If the interval was short
6929 and the sigsuspend() was delayed until after it had expired, the suspension
6930 never ended. On busy systems this could lead to processes getting stuck for
6933 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6934 process, before it forks any delivery processes. The open lookup caching
6935 mechanism meant that the open file or database connection was passed into
6936 the delivery process. The problem was that delivery processes always tidy
6937 up cached lookup data. This could cause a problem for the next delivery
6938 process started by the queue runner, because the external queue runner
6939 process does not know about the closure. So the next delivery process
6940 still has data in the lookup cache. In the case of a file lookup, there was
6941 no problem because closing a file descriptor in a subprocess doesn't affect
6942 the parent. However, if the lookup was caching a connection to a database,
6943 the connection was closed, and the second delivery process was likely to
6944 see errors such as "PGSQL: query failed: server closed the connection
6945 unexpectedly". The problem has been fixed by closing all cached lookups
6946 in a queue runner before running a delivery process.
6948 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6949 seem to have the "const" qualifier which it has on other OS. I've
6952 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6953 *of the same type* (client or server) with the same public name that an
6954 error should be diagnosed.
6956 9. When Exim looked up a host name for an IP address, but failed to find the
6957 original IP address when looking up the host name (a safety check), it
6958 output the message "<ip address> does not match any IP for NULL", which was
6959 confusing, to say the least. The bug was that the host name should have
6960 appeared instead of "NULL".
6962 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6963 user that is built into the binary, and the -C or -D options is used, root
6964 privilege is dropped before the configuration file is read. In addition,
6965 logging is switched to stderr instead of the normal log files. If the
6966 configuration then re-defines the Exim user, the unprivileged environment
6967 is probably not what is expected, so Exim logs a panic warning message (but
6970 However, if deliver_drop_privilege is set, the unprivileged state may well
6971 be exactly what is intended, so the warning has been cut out in that case,
6972 and Exim is allowed to try to write to its normal log files.
6978 1. smtp_return_error_details was not giving details for temporary sender
6979 or receiver verification errors.
6981 2. Diagnose a configuration error if two authenticators have the same public
6984 3. Exim used not to create the message log file for a message until the first
6985 delivery attempt. This could be confusing when incoming messages were held
6986 for policy or load reasons. The message log file is now created at the time
6987 the message is received, and an initial "Received" line is written to it.
6989 4. The automatically generated man page for command line options had a minor
6990 bug that caused no ill effects; however, a more serious problem was that
6991 the procedure for building the man page automatically didn't always
6992 operate. Consequently, release 4.20 contains an out-of-date version. This
6993 shouldn't happen again.
6995 5. When building Exim with embedded Perl support, the script that builds the
6996 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6997 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6999 6. The freeze_tell option was not being used for messages that were frozen on
7000 arrival, either by an ACL or by local_scan().
7002 7. Added the smtp_incomplete_transaction log selector.
7004 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7005 was accepting AUTH without a new EHLO.
7007 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7008 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7011 10. Logging of TCP/IP connections (when configured) now happens in the main
7012 daemon process instead of the child process, so that the TCP/IP connection
7013 count is more accurate (but it can never be perfect).
7015 11. The use of "drop" in a nested ACL was not being handled correctly in the
7016 outer ACL. Now, if condition failure induced by the nested "drop" causes
7017 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7018 or "require"), the connection is dropped.
7020 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7021 that yield "discard" can now be used with an "accept" or a "discard" verb,
7022 but an error is generated for any others (because I can't see a useful way
7023 to define what should happen).
7025 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7026 are now processed in the same way as lines in the Exim configuration file.
7027 In particular, continuation lines are supported.
7029 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7031 15. Added -ti meaning -t -i.
7033 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7034 domains, and warn by logging if others are found.
7036 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7037 returned from GIFCONF when Exim is trying to find the list of interfaces on
7038 a host. The code in os.c has been modified to copy each ifreq to an aligned
7039 structure in all cases.
7041 Also, in some cases, the returned ifreq's were being copied to a 'struct
7042 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7043 means the last couple of bytes of an IPv6 address could be chopped if the
7044 ifreq contained only a normal sockaddr (14 bytes storage).
7046 18. Named domain lists were not supported in the hosts_treat_as_local option.
7047 An entry such as +xxxx was not recognized, and was treated as a literal
7050 19. Ensure that header lines added by a DATA ACL are included in the reject log
7051 if the ACL subsequently rejects the message.
7053 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7054 MD5 (which is deprecated).
7056 21. When testing a filter file using -bf, Exim was writing a message when it
7057 took the sender from a "From " line in the message, but it was not doing so
7058 when it took $return_path from a Return-Path: header line. It now does.
7060 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7061 with a valid header line field name (a series of printing characters
7062 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7064 23. Changed "disc" in the source to "disk" to conform to the documentation and
7065 the book and for uniformity.
7067 24. Ignore Sendmail's -Ooption=value command line item.
7069 25. When execve() failed while trying to run a command in a pipe transport,
7070 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7071 could be confused with a return value of 69 from the command itself. This
7072 has been changed to 127, the value the shell returns if it is asked to run
7073 a non-existent command. The wording for the related log line suggests a
7074 non-existent command as the problem.
7076 26. If received_header_text expands to an empty string, do not add a Received:
7077 header line to the message. (Well, it adds a token one on the spool, but
7078 marks it "old" so that it doesn't get used or transmitted.)
7080 27. Installed eximstats 1.28 (addition of -nt option).
7082 28. There was no check for failure on the call to getsockname() in the daemon
7083 code. This can fail if there is a shortage of resources on the system, with
7084 ENOMEM, for example. A temporary error is now given on failure.
7086 29. Contrary to the C standard, it seems that in some environments, the
7087 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7088 program. Exim now does this explicitly; it affects the formatting of
7089 timestamps using strftime().
7091 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7092 complaints. I've now initialized all the variables, to avoid this.
7094 32. Header lines added by a system filter were not being "seen" during
7095 transport-time rewrites.
7097 33. The info_callback() function passed to OpenSSL is set up with type void
7098 (*)(SSL *, int, int), as described somewhere. However, when calling the
7099 function (actually a macro) that sets it up, the type void(*)() is
7100 expected. I've put in a cast to prevent warnings from picky compilers.
7102 34. If a DNS black list lookup found a CNAME record, but there were no A
7103 records associated with the domain it pointed at, Exim crashed.
7105 35. If a DNS black list lookup returned more than one A record, Exim ignored
7106 all but the first. It now scans all returned addresses if a particular IP
7107 value is being sought. In this situation, the contents of the
7108 $dnslist_value variable are a list of all the addresses, separated by a
7111 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7112 to accept a host name and all its aliases if the forward lookup for any of
7113 them yielded the IP address of the incoming connection. Now it accepts only
7114 those names whose forward lookup yields the correct IP address. Any other
7115 names are discarded. This closes a loophole whereby a rogue DNS
7116 administrator could create reverse DNS records to break through a
7117 wildcarded host restriction in an ACL.
7119 37. If a user filter or a system filter that ran in a subprocess used any of
7120 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7121 the wrong values were passed to the pipe command ($thisaddress had the
7122 value of $0, $0 had the value of $1, etc). This bug was introduced by
7123 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7125 38. Improved the line breaking for long SMTP error messages from ACLs.
7126 Previously, if there was no break point between 40 and 75 characters, Exim
7127 left the rest of the message alone. Two changes have been made: (a) I've
7128 reduced the minimum length to 35 characters; (b) if it can't find a break
7129 point between 35 and 75 characters, it looks ahead and uses the first one
7130 that it finds. This may give the occasional overlong line, but at least the
7131 remaining text gets split now.
7133 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7134 file descriptors might be low, and that setting 1000 would always raise it.
7135 It turns out that in some environments, the limit is already over 1000 and
7136 that lowering it causes trouble. So now Exim takes care not to decrease it.
7138 40. When delivering a message, the value of $return_path is set to $sender_
7139 address at the start of routing (routers may change the value). By an
7140 oversight, this default was not being set up when an address was tested by
7141 -bt or -bv, which affected the outcome if any router or filter referred to
7144 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7145 log only when "message" or "log_message" are set. However, if one of the
7146 conditions was an address verification, or a call to a nested ACL, the
7147 messages generated by the underlying test were being passed through. This
7148 no longer happens. The underlying message is available in $acl_verify_
7149 message for both "message" and "log_message" expansions, so it can be
7150 passed through if needed.
7152 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7153 new expansion $bh_ to give the encoded byte string without charset
7154 translation. Translation happens only if iconv() is available; HAVE_ICONV
7155 indicates this at build time. HEADERS_CHARSET gives the charset to
7156 translate to; headers_charset can change it in the configuration, and
7157 "headers charset" can change it in an individual filter file.
7159 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7160 that creates RFC 2047 encoded "words" labels them as that charset instead
7161 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7162 expansion operator; (ii) when Exim creates a From: line for a local
7163 message; (iii) when a header line is rewritten to include a "phrase" part.
7165 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7166 buggy, causing it to skip the first lines of messages whose message ID
7167 ended in 'D'. This would not have bitten before Exim release 4.14, because
7168 message IDs were unlikely to end in 'D' before then. The effect was to have
7169 incorrect size information for certain domains.
7171 45. #include "config.h" was missing at the start of the crypt16.c module. This
7172 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7175 46. If there was a timeout during a "random" callout check, Exim treated it as
7176 a failure of the random address, and carried on sending RSET and the real
7177 address. If the delay was just some slowness somewhere, the response to the
7178 original RCPT would be taken as a response to RSET and so on, causing
7179 mayhem of various kinds.
7181 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7182 when I implemented it. It didn't allow for the fact that some option values
7183 may legitimately be negative (e.g. size_addition), and it didn't even do
7184 the right test for positive values.
7186 48. Domain names in DNS records are case-independent. Exim always looks them up
7187 in lower case. Some resolvers return domain names in exactly the case they
7188 appear in the zone file, that is, they may contain uppercase letters. Not
7189 all resolvers do this - some return always lower case. Exim was treating a
7190 change of case by a resolver as a change of domain, similar to a widening
7191 of a domain abbreviation. This triggered its re-routing code and so it was
7192 trying to route what was effectively the same domain again. This normally
7193 caused routing to fail (because the router wouldn't handle the domain
7194 twice). Now Exim checks for this case specially, and just changes the
7195 casing of the domain that it ultimately uses when it transmits the message
7198 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7201 50. If a filter generated a file delivery with a non-absolute name (possible if
7202 no home directory exists for the router), the forbid_file option was not
7205 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7206 the existing equality matching.
7208 52. Exim was using ints instead of ino_t variables in some places where it was
7209 dealing with inode numbers.
7211 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7212 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7213 called TMPDIR, and if it finds it is different, it changes its value.
7215 54. The smtp_printf() function is now made available to local_scan() so
7216 additional output lines can be written before returning. There is also an
7217 smtp_fflush() function to enable the detection of a dropped connection.
7218 The variables smtp_input and smtp_batched_input are exported to
7221 55. Changed the default runtime configuration: the message "Unknown user"
7222 has been removed from the ACL, and instead placed on the localuser router,
7223 using the cannot_route_message feature. This means that any verification
7224 failures that generate their own messages won't get overridden. Similarly,
7225 the "Unrouteable address" message that was in the ACL for unverifiable
7226 relay addresses has also been removed.
7228 56. Added hosts_avoid_esmtp to the smtp transport.
7230 57. The exicyclog script was not checking for the esoteric option
7231 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7232 will work only if exicyclog is run under the appropriate euid.
7234 58. Following a discussion on the list, the rules by which Exim recognises line
7235 endings on incoming messages have been changed. The -dropcr and drop_cr
7236 options are now no-ops, retained only for backwards compatibility. The
7237 following line terminators are recognized: LF CRLF CR. However, special
7238 processing applies to CR:
7240 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7241 nor a local message in the state where . is a terminator.
7243 (ii) If a bare CR is encountered in a header line, an extra space is added
7244 after the line terminator so as not to end the header. The reasoning
7245 behind this is that bare CRs in header lines are most likely either
7246 to be mistakes, or people trying to play silly games.
7248 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7249 was being incorrectly given as 18 bytes larger than it should have been.
7250 This is a VOB (very old bug).
7252 60. This may never have affected anything current, but just in case it has:
7253 When the local host is found other than at the start of a list of hosts,
7254 the local host, those with the same MX, and any that follow, are discarded.
7255 When the list in question was part of a longer list of hosts, the following
7256 hosts (not currently being processed) were also being discarded. This no
7257 longer happens. I'm not sure if this situation could ever has previously
7260 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7263 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7264 header. This is something that is recommended in a new Internet Draft, and
7265 is something that is documented as being done by Sendmail. There are two
7266 possible values. For messages generated by the autoreply transport, Exim
7269 Auto-Submitted: auto-replied
7271 whereas for all other generated messages (e.g. bounces) it adds
7273 Auto-Submitted: auto-generated
7275 63. The "personal" condition in filters now includes a test for the
7276 Auto-Submitted: header. If it contains the string "auto-" the message it
7277 not considered personal.
7279 64. Added rcpt_include_affixes as a generic transport option.
7281 65. Added queue_only_override (default true).
7283 66. Added the syslog_duplication option.
7285 67. If what should have been the first header line of a message consisted of
7286 a space followed by a colon, Exim was mis-interpreting it as a header line.
7287 It isn't of course - it is syntactically invalid and should therefore be
7288 treated as the start of the message body. The misbehaviour could have
7289 caused a number of strange effects, including loss of data in subsequent
7290 header lines, and spool format errors.
7292 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7293 client host had authenticated. This control can now be exercised by an ACL
7294 for more flexibility.
7296 69. By default, callouts do not happen when testing with -bh. There is now a
7297 variant, -bhc, which does actually run the callout code, including
7298 consulting and updating the callout cache.
7300 70. Added support for saslauthd authentication, courtesy of Alexander
7303 71. If statvfs() failed on the spool or log directories while checking their
7304 size for availability, Exim confusingly gave the error "space shortage".
7305 Furthermore, in debugging mode it crashed with a floating point exception.
7306 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7307 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7308 serious problem, Exim now writes to the main and panic logs when this
7309 happens, with details of the failure. It then refuses to accept the
7310 incoming message, giving the message "spool directory problem" or "log
7311 directory problem" with a 421 code for SMTP messages.
7313 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7314 0, 1, and 2 exist, because some OS complain for execs without them (see
7315 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7316 descriptors. However, the code omitted to check that the open succeeded,
7317 causing mysterious errors if for some reason the permissions on /dev/null
7318 got screwed. Now Exim writes a message to the main and panic logs, and
7319 bombs out if it can't open /dev/null.
7321 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7322 interact so that it is all more flexible. It is supposed to remain
7323 backwards compatible. Also added extra_local_interfaces.
7325 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7326 to bomb out with an assertion failure - to the client this appears as a
7327 connection drop. This problem occurs in the part of the code that was taken
7328 from the Samba project. Fortunately, the assertion is in a very simple
7329 function, so I have fixed this by reproducing the function inline in the
7330 one place where it is called, and arranging for authentication to fail
7331 instead of killing the process with assert().
7333 75. The SPA client code was not working when the server requested OEM rather
7334 than Unicode encoding.
7336 76. Added code to make require_files with a specific uid setting more usable in
7337 the case where statting the file as root fails - usually a non-root-mounted
7338 NFS file system. When this happens and the failure is EACCES, Exim now
7339 forks a subprocess and does the per-uid checking as the relevant uid.
7341 77. Added process_log_path.
7343 78. If log_file_path was not explicitly set, a setting of check_log_space or
7344 check_log_inodes was ignored.
7346 79. If a space check for the spool or log partitions fails, the incident is now
7347 logged. Of course, in the latter case the data may get lost...
7349 80. Added the %p formatting code to string_format() so that it can be used to
7350 print addresses in debug_print(). Adjusted all the address printing in the
7351 debugging in store.c to use %p rather than %d.
7353 81. There was a concern that a line of code in smtp_in.c could overflow a
7354 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7355 initially expressed, the concern was not well-founded, because trailing
7356 spaces are removed early. However, if the trailing spaces were followed by
7357 a NULL, they did not get removed, so the overflow was possible. Two fixes
7360 (a) I re-wrote the offending code in a cleaner fashion.
7361 (b) If an incoming SMTP command contains a NULL character, it is rejected
7364 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7365 runs initgroups(), so that if the Exim user is in any additional groups,
7366 they will be used during message reception.
7372 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob