1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
22 JH/01 Remove code calling the customisable local_scan function, unless a new
23 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
25 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
26 non-signal-safe functions being used.
28 JH/03 Bug 2269: When presented with a received message having a stupidly large
29 number of DKIM-Signature headers, disable DKIM verification to avoid
30 a resource-consumption attack. The limit is set at twenty.
32 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
33 report of oldest_pass in ${authres } in consequence, and separate out
34 some descriptions of reasons for verification fail.
36 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
37 files in the spool were present and unlocked. A queue-runner could spot
38 them, resulting in a duplicate delivery. Fix that by doing the unlock
39 after the unlink. Investigation by Tim Stewart. Take the opportunity to
40 add more error-checking on spoolfile handling while that code is being
43 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
44 No known attacks, no CVE, this is defensive hardening.
46 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
47 a queue-runner could start a delivery while other operations were ongoing.
48 Cutthrough delivery was a common victim, resulting in duplicate delivery.
49 Found and investigated by Tim Stewart. Fix by using the open message data
50 file handle rather than opening another, and not locally closing it (which
51 releases a lock) for that case, while creating the temporary .eml format
52 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
54 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
55 $sender_verify_failure/$recipient_verify_failure to "random".
57 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
60 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
61 Previously this would segfault.
63 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
66 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
67 like zero, since the resolver should be doing this for us, But we need one
68 as a CNAME but no MX presence gets the CNAME returned; we need to check
69 that doesn't point to an MX to declare it "no MX returned" rather than
70 "error, loop". A new main option is added so the older capability of
71 following some limited number of chain links is maintained.
73 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
75 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
76 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
77 now-deprecated earlier definitions used only specified the range up to TLS
78 1.2 (in the older-version library docs).
80 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
82 JH/15 Rework TLS client-side context management. Stop using a global, and
83 explicitly pass a context around. This enables future use of TLS for
84 connections to service-daemons (eg. malware scanning) while a client smtp
85 connection is using TLS; with cutthrough connections this is quite likely.
87 JH/16 Fix ARC verification to do AS checks in reverse order.
89 JH/17 Support a "tls" option on the ${readsocket } expansion item.
91 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
92 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
93 Previously the "utf8" would be re-prepended for every additional message.
95 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
96 Previously thery were accepted, resulting in issues when attempting to
97 forward messages to a non-supporting MTA.
99 PP/02 Let -n work with printing macros too, not just options.
101 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
102 one parent address was copied, and bogus data was used at delivery-logging
103 time. Either a crash (after delivery) or bogus log data could result.
104 Discovery and analysis by Tim Stewart.
106 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
107 Previously if the string ended mid-character, we did not insert the
108 promised '?' replacement.
110 PP/04 Documentation: current string operators work on bytes, not codepoints.
112 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
113 should pack well giving a smaller memory footprint so better caching and
114 therefore performance. Group the declarations where this can't be done so
115 that the byte-sized flag variables are not interspersed among pointer
116 variables, giving a better chance of good packing by the compiler.
118 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
119 non-null, to avoid issues with sites running BATV. Previously reports were
120 sent with an empty envelope sender so looked like bounces.
122 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
123 The ignore_error flag wasn't being returned from the filter subprocess so
124 was not set for later routers. Investigation and fix by Matthias Kurz.
126 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
127 and a msg:complete for the whole, when a message is manually removed using
128 -Mrm. Developement by Matthias Kurz, hacked on by JH.
130 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
131 a "Gnu special" function, asprintf() in the DB utility binary builds; I
132 hope that is portable enough.
134 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
135 requiring a known-CA anchor certificate; make it now rely entirely on the
136 TLSA as an anchor. Checking the name on the leaf cert against the name
137 on the A-record for the host is still done for TA (but not for EE mode).
139 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
140 included in delivery lines for non-proxied connections, when compiled with
141 SUPPORT_SOCKS and running with proxy logging enabled.
143 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
144 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
145 move the existing event to fire before the normal logging of message
146 failure so that custom logging is bracketed by normal logging.
148 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
149 msg:fail:internal event. Developement by Matthias Kurz.
151 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
152 far too small for todays use of crypto signatures stored there. Go all
153 the way to the max DNS message size of 64kB, even though this might be
154 overmuch for IOT constrained device use.
156 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
157 copy a string over itself. The library routine is documented as not
158 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
160 JH/32 For main options check_spool_space and check_inode_space, where the
161 platform supports 64b integers, support more than the previous 2^31 kB
162 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
163 the previous G, M, k.
165 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
166 $authenticated_fail_id variable on authentication failure. Previously
169 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
170 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
171 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
172 GNUTLS_SEC_PARAM_MEDIUM.
174 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
175 side. Previously we would continue as if no SNI had been received.
177 JH/36 Harden the handling of string-lists. When a list consisted of a sole
178 "<" character, which should be a list-separator specification, we walked
179 off past the nul-terimation.
181 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
182 causes) even when the retry time is not yet met. Previously they were
183 not, meaning that when (say) an account was over-quota and temp-rejecting,
184 and multiple senders' messages were queued, only one sender would get
185 notified on each configured delay_warning cycle.
187 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they return error
188 codes indicating retry. Under TLS1.3 this becomes required.
194 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
195 When redis_servers is set to a list of > 1 element, and the Redis servers
196 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
197 case of MOVED into a DEFER case instead, thus moving the query onto the
198 next server in the list. For a cluster of N elements, all N servers must
199 be defined in redis_servers.
201 GF/02 Catch and remove uninitialized value warning in exiqsumm
202 Check for existence of @ARGV before looking at $ARGV[0]
204 JH/01 Replace the store_release() internal interface with store_newblock(),
205 which internalises the check required to safely use the old one, plus
206 the allocate and data copy operations duplicated in both (!) of the
207 extant use locations.
209 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
210 modifier. This matches the restriction on the commandline.
212 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
213 Previously only the last row was returned.
215 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
216 we assumed that tags in the header were well-formed, and parsed the
217 element content after inspecting only the first char of the tag.
218 Assumptions at that stage could crash the receive process on malformed
221 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
222 While running the DKIM ACL we operate on the Permanent memory pool so that
223 variables created with "set" persist to the DATA ACL. Also (at any time)
224 DNS lookups that fail create cache records using the Permanent pool. But
225 expansions release any allocations made on the current pool - so a dnsdb
226 lookup expansion done in the DKIM ACL releases the memory used for the
227 DNS negative-cache, and bad things result. Solution is to switch to the
228 Main pool for expansions.
229 While we're in that code, add checks on the DNS cache during store_reset,
230 active in the testsuite.
231 Problem spotted, and debugging aided, by Wolfgang Breyha.
233 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
234 When none of the hosts presented to a transport match an already-open
235 connection, close it and proceed with the list. Previously we would
236 queue the message. Spotted by Lena with Yahoo, probably involving
239 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
240 Previously a spurious "250 OK id=" response was appended to the proper
243 JH/08 The "support for" informational output now, which built with Content
244 Scanning support, has a line for the malware scanner interfaces compiled
245 in. Interface can be individually included or not at build time.
247 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
248 by the template makefile "src/EDITME". The "STREAM" support for an older
249 ClamAV interface method is removed.
251 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
252 rows affected is given instead).
254 JH/11 The runtime Berkeley DB library version is now additionally output by
255 "exim -d -bV". Previously only the compile-time version was shown.
257 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
258 SMTP connection. Previously, when one had more receipients than the
259 first, an abortive onward connection was made. Move to full support for
260 multiple onward connections in sequence, handling cutthrough connection
261 for all multi-message initiating connections.
263 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
264 routers. Previously, a multi-recipient message would fail to match the
265 onward-connection opened for the first recipient, and cause its closure.
267 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
268 a timeout on read on a GnuTLS initiating connection, resulting in the
269 initiating connection being dropped. This mattered most when the callout
270 was marked defer_ok. Fix to keep the two timeout-detection methods
273 JH/15 Relax results from ACL control request to enable cutthrough, in
274 unsupported situations, from error to silently (except under debug)
275 ignoring. This covers use with PRDR, frozen messages, queue-only and
278 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
280 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
281 metadata, resulting in a crash in free().
283 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
284 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
285 Broken also in d185889f4, with init system revamp.
287 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
288 Previously we abruptly closed the connection after reading a malware-
289 found indication; now we go on to read the "scan ok" response line,
292 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
293 ACL. Previously, a crash would result.
295 JH/19 Speed up macro lookups during configuration file read, by skipping non-
296 macro text after a replacement (previously it was only once per line) and
297 by skipping builtin macros when searching for an uppercase lead character.
299 JH/20 DANE support moved from Experimental to mainline. The Makefile control
300 for the build is renamed.
302 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
303 was allocated for every new TLS startup, meaning one per message. Fix
304 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
306 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
307 reported the original. Fix to report (as far as possible) the ACL
308 result replacing the original.
310 JH/23 Fix memory leak during multi-message connections using STARTTLS under
311 OpenSSL. Certificate information is loaded for every new TLS startup,
312 and the resources needed to be freed.
314 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
316 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
319 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
320 DATA response info to the (existing) per-recipient response info for
321 the "C=" log element. It can have useful tracking info from the
322 destination system. Patch from Simon Arlott.
324 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
325 length value. Previously this would segfault.
327 HS/02 Support Avast multiline protoocol, this allows passing flags to
328 newer versions of the scanner.
330 JH/28 Ensure that variables possibly set during message acceptance are marked
331 dead before release of memory in the daemon loop. This stops complaints
332 about them when the debug_store option is enabled. Discovered specifically
333 for sender_rate_period, but applies to a whole set of variables.
334 Do the same for the queue-runner and queue-list loops, for variables set
335 from spool message files. Do the same for the SMTP per-message loop, for
336 certain variables indirectly set in ACL operations.
338 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
339 as a multi-recipient message from a mailinglist manager). The coding had
340 an arbitrary cutoff number of characters while checking for more input;
341 enforced by writing a NUL into the buffer. This corrupted long / fast
342 input. The problem was exposed more widely when more pipelineing of SMTP
343 responses was introduced, and one Exim system was feeding another.
344 The symptom is log complaints of SMTP syntax error (NUL chars) on the
345 receiving system, and refused recipients seen by the sending system
346 (propating to people being dropped from mailing lists).
347 Discovered and pinpointed by David Carter.
349 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
350 replaced by the ${authresults } expansion.
352 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
354 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
355 allows proper process termination in container environments.
357 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
358 Previously the "final dot" had a newline after it; ensure it is CR,LF.
360 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
361 and "err_perm", deprecated since 4.83 when the RFC-defined words
362 "temperror" and "permerror" were introduced.
364 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
365 transport-filters or DKIM-signing. The restriction was lost in the
366 consolidation of verify-callout and delivery SMTP handling.
367 Extend the restriction to also cover ARC-signing.
369 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
370 in defer=pass mode supply a 450 to the initiator. Previously the message
373 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
374 tls_require_ciphers is used as before.
376 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
377 "pass_unscanned". Only tmpfails from the scanner are written to
378 the paniclog, as they may require admin intervention (permission
379 denied, license issues). Other scanner errors (like decompression
380 bombs) do not cause a paniclog entry.
382 JH/36 Fix reinitialisation of DKIM logging variable between messages.
383 Previously it was possible to log spurious information in receive log
386 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
387 triggered odd behaviour from Outlook Express clients.
389 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
392 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
393 since the IETF WG has not yet settled on that versus the original
394 "bare" representation.
396 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
397 Previously the millisecond value corrupted the output.
398 Fix also for syslog_pid=no and log_selector +pid, for which the pid
399 corrupted the output.
405 JH/01 Rework error string handling in TLS interface so that the caller in
406 more cases is responsible for logging. This permits library-sourced
407 string to be attached to addresses during delivery, and collapses
408 pairs of long lines into single ones.
410 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
411 during configuration. Wildcards are allowed and expanded.
413 JH/02 Rework error string handling in DKIM to pass more info back to callers.
414 This permits better logging.
416 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
417 do not close it down and have the child transport start it up again on
418 the passed-on TCP connection. Instead, proxy the child (and any
419 subsequent ones) for TLS via a unix-domain socket channel. Logging is
420 affected: the continued delivery log lines do not have any DNSSEC, TLS
421 Certificate or OCSP information. TLS cipher information is still logged.
423 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
424 identical IP addresses on different listening ports. Will also affect
427 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
428 add noisy ifdef guards to special-case this sillyness.
429 Patch from Bernd Kuhls.
431 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
432 than 255 are no longer allowed.
434 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
435 Disable the session-cache too, which might reduce our load. Since we
436 currrectly use a new context for every connection, both as server and
437 client, there is no benefit for these.
438 GnuTLS appears to not support tickets server-side by default (we don't
439 call gnutls_session_ticket_enable_server()) but client side is enabled
440 by default on recent versions (3.1.3 +) unless the PFS priority string
443 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
444 <https://reproducible-builds.org/specs/source-date-epoch/>.
446 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
447 the check for any unsuccessful recipients did not notice the limit, and
448 erroneously found still-pending ones.
450 JH/08 Pipeline CHUNKING command and data together, on kernels that support
451 MSG_MORE. Only in-clear (not on TLS connections).
453 JH/09 Avoid using a temporary file during transport using dkim. Unless a
454 transport-filter is involved we can buffer the headers in memory for
455 creating the signature, and read the spool data file once for the
456 signature and again for transmission.
458 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
459 4.77 as the kernel support then wasn't solid, having issues in 64bit
460 mode. Now, it's been long enough. Add support for FreeBSD also.
462 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
463 case where the routing stage had gathered several addresses to send to
464 a host before calling the transport for the first, we previously failed
465 to close down TLS in the old transport process before passing the TCP
466 connection to the new process. The new one sent a STARTTLS command
467 which naturally failed, giving a failed delivery and bloating the retry
468 database. Investigation and fix prototype from Wolfgang Breyha.
470 JH/12 Fix check on SMTP command input synchronisation. Previously there were
471 false-negatives in the check that the sender had not preempted a response
472 or prompt from Exim (running as a server), due to that code's lack of
473 awareness of the SMTP input buffering.
475 PP/04 Add commandline_checks_require_admin option.
476 Exim drops privileges sanely, various checks such as -be aren't a
477 security problem, as long as you trust local users with access to their
478 own account. When invoked by services which pass untrusted data to
479 Exim, this might be an issue. Set this option in main configuration
480 AND make fixes to the calling application, such as using `--` to stop
483 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
484 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
485 responses to those, into a single TLS record each way (this usually means
486 a single packet). As a side issue, smtp_enforce_sync now works on TLS
489 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
490 affects you only if you're dancing at the edge of the param size limits.
491 If you are, and this message makes sense to you, then: raise the
492 configured limit or use OpenSSL 1.1. Nothing we can do for older
495 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
496 cmdline element to get the documented default one. Previously it was
499 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
500 are unsuable for non-message-specific reasons - eg. connection timeout,
501 banner-time rejection.
503 JH/16 Fix logging of delivery remote port, when specified by router, under
506 PP/06 Repair manualroute's ability to take options in any order, even if one
507 is the name of a transport.
510 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
512 JH/17 Change the list-building routines interface to use the expanding-string
513 triplet model, for better allocation and copying behaviour.
515 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
516 Previously it was constructed the first time a possibly-matching string
517 was met in the configuration file input during startup; now it is done
520 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
521 compatible one, to avoid the (poorly documented) possibility of a config
522 file in the working directory redirecting the DB files, possibly correpting
523 some existing file. CVE-2017-10140 assigned for BDB.
525 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
526 cache-hot. Previously, although the result was properly cached, the
527 initial verify call returned a defer.
529 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
530 the main verify for receipient in uncached-mode.
532 JH/22 Retire historical build files to an "unsupported" subdir. These are
533 defined as "ones for which we have no current evidence of testing".
535 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
536 if present. Previously it was ignored.
538 JH/24 Start using specified-initialisers in C structure init coding. This is
539 a C99 feature (it's 2017, so now considered safe).
541 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
542 if was a fixed-sized field and bitmask ops via macros; it is now more
545 PP/07 GitHub PR 56: Apply MariaDB build fix.
546 Patch provided by Jaroslav Škarvada.
548 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
549 during Coverity cleanups [4.87 JH/47]
550 Diagnosis and fix provided by Michael Fischer v. Mollard.
552 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
553 the right size to place the terminating semicolon on its own folded
554 line, the header hash was calculated to an incorrect value thanks to
555 the (relaxed) space the fold became.
557 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
558 and confused the parent.
560 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
561 which could crash as a result. This could lead to undeliverable messages.
563 JH/28 Logging: "next input sent too soon" now shows where input was truncated
566 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
567 matters on fast-turnover and PID-randomising systems, which were getting
568 out-of-order delivery.
570 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
571 a possibly-overlapping copy. The symptom was that "Remote host closed
572 connection in response to HELO" was logged instead of the actual 4xx
575 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
576 Previously only that bufferd was discarded, resulting in SYMTP command
579 JH/32 DKIM: when a message has multiple signatures matching an identity given
580 in dkim_verify_signers, run the dkim acl once for each. Previously only
581 one run was done. Bug 2189.
583 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
584 "panic the current process" to "deliberately defer". The panic log is
585 still written with the problem list name; the mail and reject logs now
586 get a temp-reject line for the message that was being handled, saying
587 something like "domains check lookup or other defer". The SMTP 451
588 message is still "Temporary local problem".
590 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
591 A crafted sequence of BDAT commands could result in in-use memory beeing
592 freed. CVE-2017-16943.
594 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
595 from SMTP input. Previously it was always done; now only done for DATA
596 and not BDAT commands. CVE-2017-16944.
598 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
599 to the message (such as an overlong header line). Previously this was
600 not done and we did not exit BDAT mode. Followon from the previous item
601 though a different problem.
607 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
608 than -2003 did; needs libidn2 in addition to libidn.
610 JH/02 The path option on a pipe transport is now expanded before use.
612 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
613 Patch provided by "Björn", documentation fix added too.
615 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
616 missing a wire-to-host endian conversion.
618 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
619 close after a BDAT command line could be taken as a following command,
620 giving a synch failure. Fix by only checking for synch immediately
621 before acknowledging the chunk.
623 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
624 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
625 macro. Patches provided by Josh Soref.
627 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
628 Previously we did not; the RFC seems ambiguous and VRFY is not listed
629 by IANA as a service extension. However, John Klensin suggests that we
632 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
633 the dkim code may be unix-mode line endings rather than smtp wire-format
634 CRLF, so prepend a CR to any bare LF.
636 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
637 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
639 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
640 the first were themselves being wrongly included in the feed into dkim
641 processing; with most chunk sizes in use this resulted in an incorrect
642 body hash calculated value.
644 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
645 DKIM signature block, for verification. Although advised against by
646 standards it is specifically not ruled illegal.
648 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
650 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
651 missing a body hash (the bh= tag).
653 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
654 It seems that HAProxy sends the Proxy Protocol information in clear and
655 only then does a TLS startup, so do the same.
657 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
658 TCP connections (such as for Spamd) unless the daemon successfully set
659 Fast Open mode on its listening sockets. This fixes breakage seen on
660 too-old kernels or those not configured for Fast Open, at the cost of
661 requiring both directions being enabled for TFO, and TFO never being used
662 by non-daemon-related Exim processes.
664 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
665 endings, at least on the first header line. Try to canonify any that get
666 past that check, despite the cost.
668 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
669 now limited to an arbitrary five deep, while parsing addresses with the
670 strip_excess_angle_brackets option enabled.
672 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
673 instead leave the unprompted TLS handshake in socket buffer for the
674 TLS library to consume.
676 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
678 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
680 JH/16 Drop variables when they go out of scope. Memory management drops a whole
681 region in one operation, for speed, and this leaves assigned pointers
682 dangling. Add checks run only under the testsuite which checks all
683 variables at a store-reset and panics on a dangling pointer; add code
684 explicitly nulling out all the variables discovered. Fixes one known
685 bug: a transport crash, where a dangling pointer for $sending_ip_address
686 originally assigned in a verify callout, is re-used.
688 PP/06 Drop '.' from @INC in various Perl scripts.
690 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
692 PP/08 Reduce a number of compilation warnings under clang; building with
693 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
694 should be warning-free.
696 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
698 HS/01 Fix portability problems introduced by PP/08 for platforms where
699 realloc(NULL) is not equivalent to malloc() [SunOS et al].
701 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
702 chunk. This allows us to accept broken chunked messages. We need a more
703 general solution here.
705 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
706 already-broken messages in the queue.
708 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
710 JH/19 Fix reference counting bug in routing-generated-address tracking.
716 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
717 supports it and a size is available (ie. the sending peer gave us one).
719 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
720 years of being deprecated). The replacements are the ACLs
721 acl_smtp_mime and acl_not_smtp_mime.
723 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
724 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
725 or even an in-clear connection were permitted. Now, if the host lookup
726 was dnssec and dane was requested then the host is only used if the
727 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
728 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
729 if one fails this test.
730 This means that a poorly-configured remote DNS will make it incommunicado;
731 but it protects against a DNS-interception attack on it.
733 JH/04 Bug 1810: make continued-use of an open smtp transport connection
734 non-noisy when a race steals the message being considered.
736 JH/05 If main configuration option tls_certificate is unset, generate a
737 self-signed certificate for inbound TLS connections.
739 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
740 in rewrites and routers.
742 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
743 and logged a warning sing 4.83; now they are a configuration file error.
745 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
746 (lacking @domain). Apply the same qualification processing as RCPT.
748 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
750 JH/10 Support ${sha256:} applied to a string (as well as the previous
753 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
754 a cutthrough deliver is pending, as we always want to make a connection.
755 This also avoids re-routing the message when later placing the cutthrough
756 connection after a verify cache hit.
757 Do not update it with the verify result either.
759 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
760 when routing results in more than one destination address.
762 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
763 signing (which inhibits the cutthrough capability). Previously only
764 the presence of an option was tested; now an expansion evaluating as
765 empty is permissible (obviously it should depend only on data available
766 when the cutthrough connection is made).
768 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
769 the relevant preceding SMTP command did not note the pipelining mode.
771 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
772 Previously they were not counted.
774 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
775 as one having no matching records. Previously we deferred the message
776 that needed the lookup.
778 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
779 distinguished as "(=".
781 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
782 for missing MX records. Previously it only worked for missing A records.
784 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
786 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
787 after the data-go-ahead and data-ack. Patch from Jason Betts.
789 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
790 even for a "none" policy. Patch from Tony Meyer.
792 JH/22 Fix continued use of a connection for further deliveries. If a port was
793 specified by a router, it must also match for the delivery to be
796 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
797 When one of the recipients of a message was unsuitable for the connection
798 (has no matching addresses), we lost track of needing to mark it
799 deferred. As a result mail would be lost.
801 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
803 JH/25 Decoding ACL controls is now done using a binary search; the source code
804 takes up less space and should be simpler to maintain. Merge the ACL
805 condition decode tables also, with similar effect.
807 JH/26 Fix problem with one_time used on a redirect router which returned the
808 parent address unchanged. A retry would see the parent address marked as
809 delivered, so not attempt the (identical) child. As a result mail would
812 JH/27 Fix a possible security hole, wherein a process operating with the Exim
813 UID can gain a root shell. Credit to http://www.halfdog.net/ for
814 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
817 JH/28 Enable {spool,log} filesystem space and inode checks as default.
818 Main config options check_{log,spool}_{inodes,space} are now
819 100 inodes, 10MB unless set otherwise in the configuration.
821 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
822 Previously it only applied to the main-section connection policy
825 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
827 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
828 by me. Added RFC7919 DH primes as an alternative.
830 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
831 are not in the system include path.
833 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
834 GnuTLS, when a session startup failed (eg because the client disconnected)
835 Exim did stdio operations after fclose. This was exposed by a recent
836 change which nulled out the file handle after the fclose.
838 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
839 signed directly by the cert-signing cert, rather than an intermediate
840 OCSP-signing cert. This is the model used by LetsEncrypt.
842 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
844 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
845 an incoming connection.
847 HS/02 Bug 1802: Do not half-close the connection after sending a request
850 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
851 fallback to "prime256v1".
853 JH/34 SECURITY: Use proper copy of DATA command in error message.
854 Could leak key material. Remotely exploitable. CVE-2016-9963.
860 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
861 and 3.4.4 - once the server is enabled to respond to an OCSP request
862 it does even when not requested, resulting in a stapling non-aware
863 client dropping the TLS connection.
865 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
866 support variable-length bit vectors. No functional change.
868 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
869 The I= interface field on outgoing lines is now after the H= remote
870 host field, same as incoming lines. There is a separate
871 outgoing_interface log selector which allows you to disable the
874 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
875 If not running log_selector +smtp_connection the mainlog would be held
876 open indefinitely after a "too many connections" event, including to a
877 deleted file after a log rotate. Leave the per net connection logging
878 leaving it open for efficiency as that will be quickly detected by the
879 check on the next write.
881 HS/01 Bug 1671: Fix post transport crash.
882 Processing the wait-<transport> messages could crash the delivery
883 process if the message IDs didn't exist for some reason. When
884 using 'split_spool_directory=yes' the construction of the spool
885 file name failed already, exposing the same netto behaviour.
887 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
888 mime_regex ACL conditions.
890 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
891 to DSN fail messages (bounces): remote IP, remote greeting, remote response
892 to HELO, local diagnostic string.
894 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
895 log line to debug. Even when configured with a tls authenticator many
896 client connections are expected to not authenticate in this way, so
897 an authenticate fail is not an error.
899 HS/02 Add the Exim version string to the process info. This way exiwhat
900 gives some more detail about the running daemon.
902 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
903 matter for fast-change records such as DNSBLs.
905 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
906 retry record, even if constant. There may be multiple transports with
907 different interface settings and the retry behaviour needs to be kept
910 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
912 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
914 JH/10 Bug 840: fix log_defer_output option of pipe transport
916 JH/11 Bug 830: use same host for all RCPTS of a message, even under
917 hosts_randomize. This matters a lot when combined with mua_wrapper.
919 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
920 ${quote_pgsql:<string>} operator.
922 JH/13 Bug 1708: avoid misaligned access in cached lookup.
924 JH/14 Change header file name for freeradius-client. Relevant if compiling
925 with Radius support; from the Gentoo tree and checked under Fedora.
927 JH/15 Bug 1712: Introduce $prdr_requested flag variable
929 JH/16 Bug 1714: Permit an empty string as expansion result for transport
930 option transport_filter, meaning no filtering.
932 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
934 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
935 defaults to "*" (all hosts). The variable is now available when not built
936 with TLS, default unset, mainly to enable keeping the testsuite sane.
937 If a server certificate is not supplied (via tls_certificate) an error is
938 logged, and clients will find TLS connections fail on startup. Presumably
939 they will retry in-clear.
940 Packagers of Exim are strongly encouraged to create a server certificate
941 at installation time.
943 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
944 with the $config_file variable.
946 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
947 in transport context, after the attempt, and per-recipient. The latter type
948 is per host attempted. The event data is the error message, and the errno
949 information encodes the lookup type (A vs. MX) used for the (first) host,
950 and the trailing two digits of the smtp 4xx response.
952 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
953 to write to mainlog (or rejectlog, paniclog) in the window between file
954 creation and permissions/ownership being changed. Particularly affects
955 installations where exicyclog is run as root, rather than exim user;
956 result is that the running daemon panics and dies.
958 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
960 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
961 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
962 "pri" and "weight". Note that the previous implicit priority given by the
963 list order is no longer honoured.
965 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
968 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
969 by defining SUPPORT_SOCKS.
971 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
972 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
973 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
974 variables are renamed to proxy_{local,external}_{address,port}.
976 JH/27 Move Internationalisation support from Experimental to mainline, enabled
977 for a build by defining SUPPORT_I18N
979 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
980 of the query string, and make ${quote_redis:} do that quoting.
982 JH/29 Move Events support from Experimental to mainline, enabled by default
983 and removable for a build by defining DISABLE_EVENT.
985 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
987 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
988 cached by the daemon.
990 JH/32 Move Redis support from Experimental to mainline, enabled for a build
991 by defining LOOKUP_REDIS. The libhiredis library is required.
993 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
994 keys are given for lookup.
996 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
997 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
998 only supported when built with TLS support. The PolarSSL SHA routines
999 are still used when the TLS library is too old for convenient support.
1001 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1002 openssl_options), for security. OpenSSL forces this from version 1.1.0
1003 server-side so match that on older versions.
1005 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1006 allocation for $value could be released as the expansion processing
1007 concluded, but leaving the global pointer active for it.
1009 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1010 and to use the domains and local_parts ACL conditions.
1012 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1013 incorrectly not doubled on cutthrough transmission, hence seen as a
1014 body-termination at the receiving system - resulting in truncated mails.
1015 Commonly the sender saw a TCP-level error, and retransmitted the message
1016 via the normal store-and-forward channel. This could result in duplicates
1017 received - but deduplicating mailstores were liable to retain only the
1018 initial truncated version.
1020 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1022 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1024 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1025 we're in there, support oversigning also; bug 1309.
1027 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1029 HS/04 Add support for keep_environment and add_environment options.
1031 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1032 either intentional arithmetic overflow during PRNG, or testing config-
1035 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1036 delivery resulted in actual delivery. Cancel cutthrough before DATA
1039 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1040 rejects a recipient: pass the reject to the originator.
1042 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1043 Many were false-positives and ignorable, but it's worth fixing the
1046 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1047 for the new environment-manipulation done at startup. Move the routines
1048 from being local to tls.c to being global via the os.c file.
1050 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1051 an extract embedded as result-arg for a map, the first arg for extract
1052 is unavailable so we cannot tell if this is a numbered or keyed
1053 extraction. Accept either.
1059 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1062 JH/02 The smtp transport option "multi_domain" is now expanded.
1064 JH/03 The smtp transport now requests PRDR by default, if the server offers
1067 JH/04 Certificate name checking on server certificates, when exim is a client,
1068 is now done by default. The transport option tls_verify_cert_hostnames
1069 can be used to disable this per-host. The build option
1070 EXPERIMENTAL_CERTNAMES is withdrawn.
1072 JH/05 The value of the tls_verify_certificates smtp transport and main options
1073 default to the word "system" to access the system default CA bundle.
1074 For GnuTLS, only version 3.0.20 or later.
1076 JH/06 Verification of the server certificate for a TLS connection is now tried
1077 (but not required) by default. The verification status is now logged by
1078 default, for both outbound TLS and client-certificate supplying inbound
1081 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1084 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1085 Status Notification (bounce) messages are now MIME format per RFC 3464.
1086 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1087 under the control of the dsn_advertise_hosts option, and routers may
1088 have a dsn_lasthop option.
1090 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1091 default, modifiable by a malware= option. The list separator for
1092 the options can now be changed in the usual way. Bug 68.
1094 JH/10 The smtp_receive_timeout main option is now expanded before use.
1096 JH/11 The incoming_interface log option now also enables logging of the
1097 local interface on delivery outgoing connections.
1099 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1100 if the interface and destination host and port all match.
1102 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1105 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1106 Patch from Andrew Lewis.
1108 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1109 now supports optional time-restrictions, weighting, and priority
1110 modifiers per server. Patch originally by <rommer@active.by>.
1112 JH/16 The spamd_address main option now supports a mixed list of local
1113 and remote servers. Remote servers can be IPv6 addresses, and
1114 specify a port-range.
1116 JH/17 Bug 68: The spamd_address main option now supports an optional
1117 timeout value per server.
1119 JH/18 Bug 1581: Router and transport options headers_add/remove can
1120 now have the list separator specified.
1122 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1125 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1128 JH/21 Support for the A6 type of dns record is withdrawn.
1130 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1131 rather than the verbs used.
1133 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1134 from 255 to 1024 chars.
1136 JH/24 Verification callouts now attempt to use TLS by default.
1138 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1139 are generic router options now. The defaults didn't change.
1141 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1142 Original patch from Alexander Shikoff, worked over by JH.
1144 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1145 files if ZCAT_COMMAND is not executable.
1147 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1149 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1151 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1152 Normally benign, it bites when the pair was led to by a CNAME;
1153 modern usage is to not canonicalize the domain to a CNAME target
1154 (and we were inconsistent anyway for A-only vs AAAA+A).
1156 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1158 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1159 when evaluating $sender_host_dnssec.
1161 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1162 $sender_helo_dnssec variable.
1164 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1166 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1168 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1170 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1171 documented as working, but never had. Support all but $spam_report.
1173 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1174 added for tls authenticator.
1176 HS/03 Add perl_taintmode main config option
1182 TL/01 When running the test suite, the README says that variables such as
1183 no_msglog_check are global and can be placed anywhere in a specific
1184 test's script, however it was observed that placement needed to be near
1185 the beginning for it to behave that way. Changed the runtest perl
1186 script to read through the entire script once to detect and set these
1187 variables, reset to the beginning of the script, and then run through
1188 the script parsing/test process like normal.
1190 TL/02 The BSD's have an arc4random API. One of the functions to induce
1191 adding randomness was arc4random_stir(), but it has been removed in
1192 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1193 function when detected.
1195 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1196 cause callback expansion.
1198 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1199 syntax errors in an expansion can be treated as a string instead of
1200 logging or causing an error, due to the internal use of bool_lax
1201 instead of bool when processing it.
1203 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1204 server certificates when making smtp deliveries.
1206 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1208 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1210 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1212 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1213 Merged patch from Sebastian Wiedenroth.
1215 JH/05 Fix results-pipe from transport process. Several recipients, combined
1216 with certificate use, exposed issues where response data items split
1217 over buffer boundaries were not parsed properly. This eventually
1218 resulted in duplicates being sent. This issue only became common enough
1219 to notice due to the introduction of connection certificate information,
1220 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1222 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1223 size buffer was used, resulting in syntax errors when an expansion
1226 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1227 version 3.3.6 or later.
1229 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1230 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1231 both become "event_action", the variables become $event_name, $event_data
1232 and $event_defer_errno. There is a new variable $verify_mode, usable in
1233 routers, transports and related events. The tls:cert event is now also
1234 raised for inbound connections, if the main configuration event_action
1237 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1238 early OCSP support, but no stapling (appears to be less than 1.0.0).
1240 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1241 server certificate names available under the smtp transport option
1242 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1245 JH/10 Time-related extraction expansions from certificates now use the main
1246 option "timezone" setting for output formatting, and are consistent
1247 between OpenSSL and GnuTLS compilations. Bug 1541.
1249 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1250 encoded parameter in the incoming message. Bug 1558.
1252 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1253 include certificate info, eximon was claiming there were spoolfile
1256 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1258 JH/14 Log delivery-related information more consistently, using the sequence
1259 "H=<name> [<ip>]" wherever possible.
1261 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1262 are problematic for Debian distribution, omit them from the release
1265 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1267 JH/16 Fix string representation of time values on 64bit time_t architectures.
1270 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1271 output list separator was used.
1276 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1277 checkers that were complaining about end of non-void function with no
1280 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1281 This was a regression introduced in 4.83 by another bugfix.
1283 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1285 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1286 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1292 TF/01 Correctly close the server side of TLS when forking for delivery.
1294 When a message was received over SMTP with TLS, Exim failed to clear up
1295 the incoming connection properly after forking off the child process to
1296 deliver the message. In some situations the subsequent outgoing
1297 delivery connection happened to have the same fd number as the incoming
1298 connection previously had. Exim would try to use TLS and fail, logging
1299 a "Bad file descriptor" error.
1301 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1302 utilities have not been installed.
1304 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1305 temporary space as the ACL may create new global variables.
1307 TL/01 LDAP support uses per connection or global context settings, depending
1308 upon the detected version of the libraries at build time.
1310 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1311 to extract and use the src ip:port in logging and expansions as if it
1312 were a direct connection from the outside internet. PPv2 support was
1313 updated based on HAProxy spec change in May 2014.
1315 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1317 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1318 Properly escape header and check for NULL return.
1320 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1323 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1325 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1326 characters in header names, implemented as a verify condition.
1327 Contributed by Michael Fischer v. Mollard.
1329 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1330 results permerror and temperror. Previous values are deprecated but
1331 still accepted. In a future release, err_perm and err_temp will be
1332 completely removed, which will be a backward incompatibility if the
1333 ACL tests for either of these two old results. Patch contributed by
1334 user bes-internal on the mailing list.
1336 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1338 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1339 selectors, in both main and reject logs.
1341 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1344 JH/07 Add malware type "sock" for talking to simple daemon.
1346 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1348 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1349 routers/transports under cutthrough routing.
1351 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1352 numbers. Touch up "bool" conditional to keep the same definition.
1354 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1356 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1358 JH/12 Expand items in router/transport headers_add or headers_remove lists
1359 individually rather than the list as a whole. Bug 1452.
1361 Required for reasonable handling of multiple headers_ options when
1362 they may be empty; requires that headers_remove items with embedded
1363 colons must have them doubled (or the list-separator changed).
1365 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1366 view the policy declared in the DMARC record. Currently, $dmarc_status
1367 is a combined value of both the record presence and the result of the
1370 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1372 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1373 dnslookup router and the smtp transport (applying to the forward
1376 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1377 of ldap servers used for a specific lookup. Patch provided by Heiko
1380 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1381 New variable $lookup_dnssec_authenticated for observability.
1383 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1384 Patch submitted by Lars Timman.
1386 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1388 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1389 Requires trusted mode and valid format message id, aborts otherwise.
1390 Patch contributed by Heiko Schlichting.
1392 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1393 certextract with support for various fields. Bug 1358.
1395 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1396 is requested by default, modifiable by smtp transport option
1399 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1400 operate on certificate variables to give certificate fingerprints
1401 Also new ${sha256:cert_variable}.
1403 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1405 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1406 Christian Aistleitner.
1408 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1410 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1411 file. Patch from Wolfgang Breyha.
1413 JH/25 Expand the coverage of the delivery $host and $host_address to
1414 client authenticators run in verify callout. Bug 1476.
1416 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1417 align with daemon_smtp_ports. Bug 72.
1419 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1420 support and error reporting did not work properly.
1422 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1423 and is readable. Patch from Andrew Colin Kissa.
1425 TL/14 Enhance documentation of ${run expansion and how it parses the
1426 commandline after expansion, particularly in the case when an
1427 unquoted variable expansion results in an empty value.
1429 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1431 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1432 by renaming away from it. Addresses GH issue 15, reported by
1435 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1437 TL/15 SECURITY: prevent double expansion in math comparison functions
1438 (can expand unsanitized data). Not remotely exploitable.
1445 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1447 PP/02 Make -n do something, by making it not do something.
1448 When combined with -bP, the name of an option is not output.
1450 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1453 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1454 $sender_host_name and config options to manage this, and basic check
1457 PP/05 DSCP support for outbound connections and control modifier for inbound.
1459 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1460 (Only plugin which currently uses this is kerberos4, which nobody should
1461 be using, but we should make it available and other future plugins might
1462 conceivably use it, even though it would break NAT; stuff *should* be
1463 using channel bindings instead).
1465 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1466 name; added for Sendmail compatibility; requires admin caller.
1467 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1468 just ignore it); requires trusted caller.
1469 Also parse but ignore: -Ac -Am -X<logfile>
1472 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1474 TL/02 Add +smtp_confirmation as a default logging option.
1476 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1477 Patch by Magnus Holmgren from 2007-02-20.
1479 TL/04 Bugzilla 1281 - Spec typo.
1480 Bugzilla 1283 - Spec typo.
1481 Bugzilla 1290 - Spec grammar fixes.
1483 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1485 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1487 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1488 mailing list by by Dmitry Isaikin.
1490 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1492 JH/02 Support "G" suffix to numbers in ${if comparisons.
1494 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1496 NM/01 Bugzilla 1197 - Spec typo
1497 Bugzilla 1196 - Spec examples corrections
1499 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1501 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1502 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1505 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1506 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1508 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1509 "acl {{name}{arg}...}", and optional args on acl condition
1512 JH/05 Permit multiple router/transport headers_add/remove lines.
1514 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1516 JH/07 Avoid using a waiting database for a single-message-only transport.
1517 Performance patch from Paul Fisher. Bugzilla 1262.
1519 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1522 JH/09 Add $headers_added variable, with content from use of ACL modifier
1523 add_header (but not yet added to the message). Bugzilla 199.
1525 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1526 Pulled from Bugzilla 817 by Wolfgang Breyha.
1528 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1530 (nb: this is the same fix as in Exim 4.80.1)
1532 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1535 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1538 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1540 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1541 advertise SMTP AUTH mechanism to us, instead of a generic
1542 protocol violation error. Also, make Exim more robust to bad
1543 data from the Dovecot auth socket.
1545 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1547 When a queue runner is handling a message, Exim first routes the
1548 recipient addresses, during which it prunes them based on the retry
1549 hints database. After that it attempts to deliver the message to
1550 any remaining recipients. It then updates the hints database using
1553 So if a recipient address works intermittently, it can get repeatedly
1554 deferred at routing time. The retry hints record remains fresh so the
1555 address never reaches the final cutoff time.
1557 This is a fairly common occurrence when a user is bumping up against
1558 their storage quota. Exim had some logic in its local delivery code
1559 to deal with this. However it did not apply to per-recipient defers
1560 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1562 This change adds a proper retry rule check during routing so that the
1563 final cutoff time is checked against the message's age. We only do
1564 this check if there is an address retry record and there is not a
1565 domain retry record; this implies that previous attempts to handle
1566 the address had the retry_use_local_parts option turned on. We use
1567 this as an approximation for the destination being like a local
1568 delivery, as in LMTP.
1570 I suspect this new check makes the old local delivery cutoff check
1571 redundant, but I have not verified this so I left the code in place.
1573 TF/02 Correct gecos expansion when From: is a prefix of the username.
1575 Test 0254 submits a message to Exim with the header
1579 When I ran the test suite under the user fanf2, Exim expanded
1580 the header to contain my full name, whereas it should have added
1581 a Resent-Sender: header. It erroneously treats any prefix of the
1582 username as equal to the username.
1584 This change corrects that bug.
1586 GF/01 DCC debug and logging tidyup
1587 Error conditions log to paniclog rather than rejectlog.
1588 Debug lines prefixed by "DCC: " to remove any ambiguity.
1590 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1592 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1593 Bug spotted by Jeremy Harris; was flawed since initial commit.
1594 Would have resulted in OCSP responses post-SNI triggering an Exim
1595 NULL dereference and crash.
1597 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1599 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1600 Bug detection, analysis and fix by Samuel Thibault.
1601 Bugzilla 1331, Debian bug #698092.
1603 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1605 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1606 Server implementation by Todd Lyons, client by JH.
1607 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1608 config variable "prdr_enable" controls whether the server
1609 advertises the facility. If the client requests PRDR a new
1610 acl_data_smtp_prdr ACL is called once for each recipient, after
1611 the body content is received and before the acl_smtp_data ACL.
1612 The client is controlled by both of: a hosts_try_prdr option
1613 on the smtp transport, and the server advertisement.
1614 Default client logging of deliveries and rejections involving
1615 PRDR are flagged with the string "PRDR".
1617 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1618 fclose(). Diagnosis by Todd Lyons.
1620 PP/17 Update configure.default to handle IPv6 localhost better.
1621 Patch by Alain Williams (plus minor tweaks).
1624 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1625 This is now consistent with GnuTLS, and is now documented: the
1626 previous undocumented portable approach to treating the option as
1627 unset was to force an expansion failure. That still works, and
1628 an empty string is now equivalent.
1630 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1631 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1632 not performing validation itself.
1634 PP/20 Added force_command boolean option to pipe transport.
1635 Patch from Nick Koston, of cPanel Inc.
1637 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1640 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1642 PP/21 Fix eximon continuous updating with timestamped log-files.
1643 Broken in a format-string cleanup in 4.80, missed when I repaired the
1644 other false fix of the same issue.
1645 Report and fix from Heiko Schlichting.
1648 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1649 Report from Prashanth Katuri.
1651 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1652 It's SecureTransport, so affects any MacOS clients which use the
1653 system-integrated TLS libraries, including email clients.
1655 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1656 using a MIME ACL for non-SMTP local injection.
1657 Report and assistance in diagnosis by Warren Baker.
1659 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1661 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1663 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1664 last id that failed so it may be referenced in subsequent ACL's.
1666 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1669 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1670 ldap library initialization, allowing self-signed CA's to be
1671 used. Also properly sets require_cert option later in code by
1672 using NULL (global ldap config) instead of ldap handle (per
1673 session). Bug diagnosis and testing by alxgomz.
1675 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1676 the src/util/ subdirectory.
1678 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1679 renamed to Transport Post Delivery Action by Jeremy Harris, as
1682 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1683 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1684 redis_servers = needs to be configured which will be used by the redis
1685 lookup. Patch from Warren Baker, of The Packet Hub.
1687 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1689 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1690 hostname or reverse DNS when processing a host list. Used suggestions
1691 from multiple comments on this bug.
1693 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1695 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1696 Missed a few lines, added it to make the runtest require no keyboard
1699 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1700 contains upper case chars. Make router use caseful_local_part.
1702 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1703 support when GnuTLS has been built with p11-kit.
1709 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1711 This, or similar/improved, will also be change PP/11 of 4.82.
1717 PP/01 Handle short writes when writing local log-files.
1718 In practice, only affects FreeBSD (8 onwards).
1719 Bugzilla 1053, with thanks to Dmitry Isaikin.
1721 NM/01 Bugzilla 949 - Documentation tweak
1723 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1726 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1728 PP/02 Implemented gsasl authenticator.
1730 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1732 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1733 `pkg-config foo` for cflags/libs.
1735 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1736 with rest of GSASL and with heimdal_gssapi.
1738 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1739 `pkg-config foo` for cflags/libs for the TLS implementation.
1741 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1742 properties get this fed in as external SSF. A number of robustness
1743 and debugging improvements to the cyrus_sasl authenticator.
1745 PP/08 cyrus_sasl server now expands the server_realm option.
1747 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1748 Patch by Jeremy Harris.
1750 PP/10 Added dbmjz lookup type.
1752 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1754 PP/12 MAIL args handles TAB as well as SP, for better interop with
1755 non-compliant senders.
1756 Analysis and variant patch by Todd Lyons.
1758 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1759 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1760 Patch from Dirk Mueller <dmueller@suse.com>
1762 PP/13 tls_peerdn now print-escaped for spool files.
1763 Observed some $tls_peerdn in wild which contained \n, which resulted
1764 in spool file corruption.
1766 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1767 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1768 or write after TLS renegotiation, which otherwise led to messages
1771 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1772 as a tracking header (ie: a signed header comes before the signature).
1773 Patch from Wolfgang Breyha.
1775 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1776 comma-sep list; embedded commas doubled.
1778 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1780 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1782 Report and patch from Dmitry Banschikov.
1784 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1785 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1786 failures. We appear to now support TLS1.1+ with Exim.
1788 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1789 lets Exim select keys and certificates based upon TLS SNI from client.
1790 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1791 before an outbound SMTP session. New log_selector, +tls_sni.
1793 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1794 NULL dereference. Report and patch from Alun Jones.
1796 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1797 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1798 Not seeing resolver debug output on NetBSD, but suspect this is a
1799 resolver implementation change.
1801 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1802 Left warnings. Added "eximon gdb" invocation mode.
1804 PP/21 Defaulting "accept_8bitmime" to true, not false.
1806 PP/22 Added -bw for inetd wait mode support.
1808 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1809 locate the relevant includes and libraries. Made this the default.
1811 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1812 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1814 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1815 This may cause build issues on older platforms.
1817 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1818 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1819 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1820 Added SNI support via GnuTLS too.
1821 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1823 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1825 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1826 Applied second patch from Janne, implementing suggestion to default
1827 multiple-strings-in-record handling to match SPF spec.
1829 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1831 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1832 read-only, out of scope).
1833 Patch from Wolfgang Breyha, report from Stuart Northfield.
1835 PP/29 Fix three issues highlighted by clang analyser static analysis.
1836 Only crash-plausible issue would require the Cambridge-specific
1837 iplookup router and a misconfiguration.
1838 Report from Marcin Mirosław.
1840 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1842 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1843 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1844 As part of this, removing so much warning spew let me fix some minor
1845 real issues in debug logging.
1847 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1848 assignment on my part. Fixed.
1850 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1851 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1852 Janne Snabb (who went above and beyond: thank you).
1854 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1855 string otherwise requires a connection and a bunch more work and it's
1856 relatively easy to get wrong. Should also expose TLS library linkage
1859 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1860 64-bit ${eval} (JH/03).
1862 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1863 GNU libc to support some of the 64-bit stuff, should not lead to
1864 conflicts. Defined before os.h is pulled in, so if a given platform
1865 needs to override this, it can.
1867 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1868 protection layer was required, which is not implemented.
1869 Bugzilla 1254, patch from Wolfgang Breyha.
1871 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1872 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1873 tls_dhparam take prime identifiers. Also unbreak combination of
1874 OpenSSL+DH_params+TLSSNI.
1876 PP/39 Disable SSLv2 by default in OpenSSL support.
1882 PP/01 Solaris build fix for Oracle's LDAP libraries.
1883 Bugzilla 1109, patch from Stephen Usher.
1885 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1887 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1890 TF/02 Fix a couple more cases where we did not log the error message
1891 when unlink() failed. See also change 4.74-TF/03.
1893 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1894 lock up or crash if it happened to be inside a call to libc when it
1895 got a SIGUSR1 from exiwhat.
1897 The SIGUSR1 handler appends the current process status to the process
1898 log which is later printed by exiwhat. It used to use the general
1899 purpose logging code to do this, but several functions it calls are
1900 not safe for signals.
1902 The new output code in the SIGUSR1 handler is specific to the process
1903 log, and simple enough that it's easy to inspect for signal safety.
1904 Removing some special cases also simplifies the general logging code.
1905 Removing the spurious timestamps from the process log simplifies
1908 TF/04 Improved ratelimit ACL condition.
1910 The /noupdate option has been deprecated in favour of /readonly which
1911 has clearer semantics. The /leaky, /strict, and /readonly update modes
1912 are mutually exclusive. The update mode is no longer included in the
1913 database key; it just determines when the database is updated. (This
1914 means that when you upgrade Exim will forget old rate measurements.)
1916 Exim now checks that the per_* options are used with an update mode that
1917 makes sense for the current ACL. For example, when Exim is processing a
1918 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1919 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1920 must specify per_mail/readonly. If you omit the update mode it defaults to
1921 /leaky where that makes sense (as before) or /readonly where required.
1923 The /noupdate option is now undocumented but still supported for
1924 backwards compatibility. It is equivalent to /readonly except that in
1925 ACLs where /readonly is required you may specify /leaky/noupdate or
1926 /strict/noupdate which are treated the same as /readonly.
1928 A useful new feature is the /count= option. This is a generalization
1929 of the per_byte option, so that you can measure the throughput of other
1930 aggregate values. For example, the per_byte option is now equivalent
1931 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1933 The per_rcpt option has been generalized using the /count= mechanism
1934 (though it's more complicated than the per_byte equivalence). When it is
1935 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1936 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1937 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1938 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1939 Note that using per_rcpt with a non-readonly update mode in more than
1940 one ACL will cause the recipients to be double-counted. (The per_mail
1941 and per_byte options don't have this problem.)
1943 The handling of very low rates has changed slightly. If the computed rate
1944 is less than the event's count (usually one) then this event is the first
1945 after a long gap. In this case the rate is set to the same as this event's
1946 count, so that the first message of a spam run is counted properly.
1948 The major new feature is a mechanism for counting the rate of unique
1949 events. The new per_addr option counts the number of different
1950 recipients that someone has sent messages to in the last time period. It
1951 behaves like per_rcpt if all the recipient addresses are different, but
1952 duplicate recipient addresses do not increase the measured rate. Like
1953 the /count= option this is a general mechanism, so the per_addr option
1954 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1955 example, measure the rate that a client uses different sender addresses
1956 with the options per_mail/unique=$sender_address. There are further
1957 details in the main documentation.
1959 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1961 TF/06 Removed a few PCRE remnants.
1963 TF/07 Automatically extract Exim's version number from tags in the git
1964 repository when doing development or release builds.
1966 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1967 Bugzilla 879. Patch from Paul Fisher.
1969 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1970 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1973 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1975 PP/05 Variable $av_failed, true if the AV scanner deferred.
1976 Bugzilla 1078. Patch from John Horne.
1978 PP/06 Stop make process more reliably on build failure.
1979 Bugzilla 1087. Patch from Heiko Schlittermann.
1981 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1982 Bugzilla 1089. Patch from Heiko Schlittermann.
1984 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1985 Bugzilla 1131. Patch from Holger Weiß.
1987 PP/09 Handle IPv6 addresses with SPF.
1988 Bugzilla 860. Patch from Wolfgang Breyha.
1990 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1992 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1995 PP/11 match_* no longer expand right-hand-side by default.
1996 New compile-time build option, EXPAND_LISTMATCH_RHS.
1997 New expansion conditions, "inlist", "inlisti".
1999 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2001 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2003 PP/14 fix log_write() format string regression from TF/03.
2004 Bugzilla 1152. Patch from Dmitry Isaikin.
2010 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2012 PP/02 Harmonised TLS library version reporting; only show if debugging.
2013 Layout now matches that introduced for other libraries in 4.74 PP/03.
2015 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2017 PP/04 New "dns_use_edns0" global option.
2019 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2022 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2023 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2025 TK/01 Updated PolarSSL code to 0.14.2.
2026 Bugzilla 1097. Patch from Andreas Metzler.
2028 PP/07 Catch divide-by-zero in ${eval:...}.
2029 Fixes bugzilla 1102.
2031 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2034 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2035 format-string attack -- SECURITY: remote arbitrary code execution.
2037 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2038 time unintentionally subject to list matching rules, letting the header
2039 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2040 arbitrary string expansion). This allowed for information disclosure.
2042 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2043 INT_MIN/-1 -- value coerced to INT_MAX.
2049 NM/01 Workaround for PCRE version dependency in version reporting
2052 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2053 This fixes portability to compilers other than gcc, notably
2054 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2056 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2057 makefiles for portability to HP-UX and POSIX correctness.
2059 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2060 Also via indented variable definition in the Makefile.
2061 (Debugging by Oliver Heesakkers).
2063 PP/02 Restore caching of spamd results with expanded spamd_address.
2064 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2066 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2067 Improves build reliability. Fix from: Frank Elsner
2069 NM/02 Fix wide character breakage in the rfc2047 coding
2070 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2072 NM/03 Allow underscore in dnslist lookups
2073 Fixes bug 1026. Patch from Graeme Fowler
2075 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2076 Code patches from Adam Ciarcinski of NetBSD.
2078 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2081 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2082 is logged, to avoid truncation. Patch from John Horne.
2084 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2085 Patch from Jakob Hirsch.
2087 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2088 SQL string expansion failure details.
2089 Patch from Andrey Oktyabrski.
2091 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2092 Patch from Simon Arlott.
2094 PP/09 New lookups functionality failed to compile on old gcc which rejects
2095 extern declarations in function scope.
2096 Patch from Oliver Fleischmann
2098 PP/10 Use sig_atomic_t for flags set from signal handlers.
2099 Check getgroups() return and improve debugging.
2100 Fixed developed for diagnosis in bug 927 (which turned out to be
2103 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2104 Patch from Mark Zealey.
2106 PP/12 Bugzilla 1056: Improved spamd server selection.
2107 Patch from Mark Zealey.
2109 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2110 Based on patch from Heiko Schlittermann.
2112 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2113 Patch from Uwe Doering, sign-off by Michael Haardt.
2115 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2116 variable declaration deep within a block. Bug and patch from
2119 PP/15 lookups-Makefile IRIX compatibility coercion.
2121 PP/16 Make DISABLE_DKIM build knob functional.
2123 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2124 Patch by Simon Arlott
2126 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2127 variable argument macros. Our copy now differs from upstream.
2133 TF/01 Failure to get a lock on a hints database can have serious
2134 consequences so log it to the panic log.
2136 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2137 controlled using the smtp_confirmation log selector.
2139 TF/03 Include the error message when we fail to unlink a spool file.
2141 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2142 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2143 for maintaining out-of-tree patches for some time.
2145 PP/01 Bugzilla 139: Documentation and portability issues.
2146 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2147 Handle per-OS dynamic-module compilation flags.
2149 PP/02 Let /dev/null have normal permissions.
2150 The 4.73 fixes were a little too stringent and complained about the
2151 permissions on /dev/null. Exempt it from some checks.
2152 Reported by Andreas M. Kirchwitz.
2154 PP/03 Report version information for many libraries, including
2155 Exim version information for dynamically loaded libraries. Created
2156 version.h, now support a version extension string for distributors
2157 who patch heavily. Dynamic module ABI change.
2159 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2160 privilege escalation vulnerability whereby the Exim run-time user
2161 can cause root to append content of the attacker's choosing to
2164 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2167 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2168 If dropping privileges for untrusted macros, we disabled normal logging
2169 on the basis that it would fail; for the Exim run-time user, this is not
2170 the case, and it resulted in successful deliveries going unlogged.
2171 Fixed. Reported by Andreas Metzler.
2177 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2178 only prepend for the Resent-* case. Fixes regression introduced in
2179 Exim 4.70 by NM/22 for Bugzilla 607.
2181 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2182 increasing numbers of administrators be bitten by this.
2184 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2186 PP/03 Bugzilla 994: added openssl_options main configuration option.
2188 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2190 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2192 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2194 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2195 without a peer certificate, leading to a segfault because of an
2196 assumption that peers always have certificates. Be a little more
2197 paranoid. Problem reported by Martin Tscholak.
2199 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2200 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2201 NB: ClamAV planning to remove STREAM in "middle of 2010".
2202 CL also introduces -bmalware, various -d+acl logging additions and
2203 more caution in buffer sizes.
2205 PP/09 Implemented reverse_ip expansion operator.
2207 PP/10 Bugzilla 937: provide a "debug" ACL control.
2209 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2211 PP/12 Bugzilla 973: Implement --version.
2213 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2215 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2217 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2219 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2220 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2221 ignore trailing whitespace.
2223 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2225 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2228 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2229 Notification from Dr Andrew Aitchison.
2231 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2232 ExtendedDetectionInfo response format.
2233 Notification from John Horne.
2235 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2238 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2239 XSL and documented dependency on system catalogs, with examples of how
2242 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2245 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2246 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2247 configuration file which is writeable by the Exim user or group.
2249 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2250 of configuration files to cover files specified with the -C option if
2251 they are going to be used with root privileges, not just the default
2254 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2255 option (effectively making it always true).
2257 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2258 files to be used while preserving root privileges.
2260 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2261 that rogue child processes cannot use them.
2263 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2264 run-time user, instead of root.
2266 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2267 Exim run-time user without dropping privileges.
2269 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2270 result string, instead of calling string_vformat() twice with the same
2273 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2274 for other users. Others should always drop root privileges if they use
2275 -C on the command line, even for a whitelisted configure file.
2277 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2279 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2285 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2286 $data_path, and $header_path variables; fixed documentation bugs and
2289 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2290 exipick to access non-standard spools, including the "frozen" queue
2293 NM/01 Bugzilla 965: Support mysql stored procedures.
2294 Patch from Alain Williams
2296 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2298 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2299 Patch from Andreas Metzler
2301 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2302 Patch from Kirill Miazine
2304 NM/05 Bugzilla 671: Added umask to procmail example.
2306 JJ/03 installed exipick 20100323.0, fixing doc bug
2308 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2309 directory. Notification and patch from Dan Rosenberg.
2311 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2313 TK/02 Improve log output when DKIM signing operation fails.
2315 MH/01 Treat the transport option dkim_domain as a colon separated
2316 list, not as a single string, and sign the message with each element,
2317 omitting multiple occurences of the same signer.
2319 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2320 Bugzilla 985, 986. Patch by Simon Arlott
2322 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2323 Patch by Simon Arlott
2325 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2326 MBX locking. Notification from Dan Rosenberg.
2332 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2334 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2336 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2338 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2340 NM/04 Bugzilla 915: Flag broken perl installation during build.
2346 TK/01 Added patch by Johannes Berg that expands the main option
2347 "spamd_address" if it starts with a dollar sign.
2349 TK/02 Write list of recipients to X-Envelope-Sender header when building
2350 the mbox-format spool file for content scanning (suggested by Jakob
2353 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2354 (http://www.dcc-servers.net/) support via dccifd. Activated by
2355 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2357 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2358 by Mark Daniel Reidel <mr@df.eu>.
2360 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2361 When building exim an external PCRE library is now needed -
2362 PCRE is a system library on the majority of modern systems.
2363 See entry on PCRE_LIBS in EDITME file.
2365 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2366 conversation. Added nologin parameter to request.
2367 Patch contributed by Kirill Miazine.
2369 TF/01 Do not log submission mode rewrites if they do not change the address.
2371 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2373 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2374 log files in place. Contributed by Roberto Lima.
2376 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2378 TF/03 Bugzilla 615: When checking the local_parts router precondition
2379 after a local_part_suffix or local_part_prefix option, Exim now
2380 does not use the address's named list lookup cache, since this
2381 contains cached lookups for the whole local part.
2383 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2384 Robert Millan. Documentation is in experimental-spec.txt.
2386 TF/04 Bugzilla 668: Fix parallel build (make -j).
2388 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2390 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2391 Patch provided by Jan Srzednicki.
2393 TF/05 Leading white space used to be stripped from $spam_report which
2394 wrecked the formatting. Now it is preserved.
2396 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2397 that they are available at delivery time.
2399 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2401 TF/08 TLS error reporting now respects the incoming_interface and
2402 incoming_port log selectors.
2404 TF/09 Produce a more useful error message if an SMTP transport's hosts
2405 setting expands to an empty string.
2407 NM/06 Bugzilla 744: EXPN did not work under TLS.
2408 Patch provided by Phil Pennock.
2410 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2411 Patch provided by Richard Godbee.
2413 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2414 acl_smtp_notquit, added index entry.
2416 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2417 Patch provided by Eugene Bujak.
2419 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2420 accept(). Patch provided by Maxim Dounin.
2422 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2423 Patch provided by Phil Pennock.
2425 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2427 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2428 Patch provided by Brad "anomie" Jorsch.
2430 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2431 Patch provided by Dean Brooks.
2433 TK/05 Add native DKIM support (does not depend on external libraries).
2435 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2436 Patch provided by Graeme Fowler.
2438 NM/16 Bugzilla 851: Documentation example syntax fix.
2440 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2442 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2445 NM/19 Bugzilla 745: TLS version reporting.
2446 Patch provided by Phil Pennock.
2448 NM/20 Bugzilla 167: bool: condition support.
2449 Patch provided by Phil Pennock.
2451 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2452 clients. Patch provided by Phil Pennock.
2454 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2455 Patch provided by Brad "anomie" Jorsch.
2457 NM/23 Bugzilla 687: Fix misparses in eximstats.
2458 Patch provided by Heiko Schlittermann.
2460 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2461 Patch provided by Heiko Schlittermann.
2463 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2464 plus update to original patch.
2466 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2468 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2469 Patch provided by David Brownlee.
2471 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2473 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2475 NM/30 Bugzilla 888: TLS documentation bugfixes.
2477 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2479 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2480 Unlike the original bugzilla I have changed all shell scripts in src tree.
2482 NM/33 Bugzilla 898: Transport filter timeout fix.
2483 Patch by Todd Rinaldo.
2485 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2486 Patch by Serge Demonchaux.
2488 NM/35 Bugzilla 39: Base64 decode bug fixes.
2489 Patch by Jakob Hirsch.
2491 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2493 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2495 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2497 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2503 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2504 ALT-N's libdkim that I have put here:
2505 http://duncanthrax.net/exim-experimental/
2507 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2508 were called 'true' and it seems that C99 defines that as a reserved
2509 keyword to be used with 'bool' variable types. That means you could
2510 not include C99-style headers which use bools without triggering
2511 build errors in sieve.c.
2513 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2514 as mailq or other aliases. Changed the --help handling significantly
2515 to do whats expected. exim_usage() emits usage/help information.
2517 SC/01 Added the -bylocaldomain option to eximstats.
2519 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2521 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2523 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2529 PH/01 Another patch from the Sieve maintainer.
2531 PH/02 When an IPv6 address is converted to a string for single-key lookup
2532 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2533 dots are used instead of colons so that keys in lsearch files need not
2534 contain colons. This was done some time before quoting was made available
2535 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2536 (notated using the quote facility) so as to distinguish them from IPv4
2537 keys. This meant that lookups for IP addresses in host lists did not work
2538 for iplsearch lookups.
2540 This has been fixed by arranging for IPv6 addresses to be expressed with
2541 colons if the lookup type is iplsearch. This is not incompatible, because
2542 previously such lookups could never work.
2544 The situation is now rather anomalous, since one *can* have colons in
2545 ordinary lsearch keys. However, making the change in all cases is
2546 incompatible and would probably break a number of configurations.
2548 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2551 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2552 conversion specification without a maximum field width, thereby enabling
2553 a rogue spamd server to cause a buffer overflow. While nobody in their
2554 right mind would setup Exim to query an untrusted spamd server, an
2555 attacker that gains access to a server running spamd could potentially
2556 exploit this vulnerability to run arbitrary code as the Exim user.
2558 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2559 $primary_hostname instead of what libspf2 thinks the hosts name is.
2561 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2562 a directory entry by the name of the lookup key. Previously, if a
2563 symlink pointed to a non-existing file or a file in a directory that
2564 Exim lacked permissions to read, a lookup for a key matching that
2565 symlink would fail. Now it is enough that a matching directory entry
2566 exists, symlink or not. (Bugzilla 503.)
2568 PH/03 The body_linecount and body_zerocount variables are now exported in the
2571 PH/04 Added the $dnslist_matched variable.
2573 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2574 This means they are set thereafter only if the connection becomes
2577 PH/06 Added the client_condition to authenticators so that some can be skipped
2578 by clients under certain conditions.
2580 PH/07 The error message for a badly-placed control=no_multiline_responses left
2581 "_responses" off the end of the name.
2583 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2585 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2586 (without spaces) instead of just copying the configuration text.
2588 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2590 PH/11 Added $max_received_linelength.
2592 PH/12 Added +ignore_defer and +include_defer to host lists.
2594 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2595 way in which PCRE > 7.0 is built.
2597 PH/14 Implemented queue_only_load_latch.
2599 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2600 MAIL command. The effect was to mangle the value on 64-bit systems.
2602 PH/16 Another patch from the Sieve maintainer.
2604 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2606 PH/18 If a system quota error occurred while trying to create the file for
2607 a maildir delivery, the message "Mailbox is full" was not appended to the
2608 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2609 only to a quota excession during the actual writing of the file.
2611 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2612 characters?) which causes problems in log lines. The DN values are now
2613 passed through string_printing() before being added to log lines.
2615 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2616 and InterBase are left for another time.)
2618 PH/21 Added message_body_newlines option.
2620 PH/22 Guard against possible overflow in moan_check_errorcopy().
2622 PH/23 POSIX allows open() to be a macro; guard against that.
2624 PH/24 If the recipient of an error message contained an @ in the local part
2625 (suitably quoted, of course), incorrect values were put in $domain and
2626 $local_part during the evaluation of errors_copy.
2632 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2633 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2636 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2637 issue a MAIL command.
2639 PH/02 In an ACL statement such as
2641 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2643 if a client was not listed at all, or was listed with a value other than
2644 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2645 the condition was not true (as it should be), so access was not denied.
2646 The bug was that the ! inversion was incorrectly passed on to the second
2647 item. This has been fixed.
2649 PH/03 Added additional dnslists conditions == and =& which are different from
2650 = and & when the dns lookup returns more than one IP address.
2652 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2653 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2655 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2656 FSYNC, which compiles an option called disable_fsync that allows for
2657 bypassing fsync(). The documentation is heavily laced with warnings.
2659 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2661 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2662 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2663 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2664 including adding "make clean"; (3) Added -fPIC when compiling the test
2665 dynamically loaded module, to get rid of a warning.
2667 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2668 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2669 = 0s. The bug is otherwise harmless.
2671 PH/07 There was a bug in the dovecot authenticator such that the value of
2672 $auth1 could be overwritten, and so not correctly preserved, after a
2673 successful authentication. This usually meant that the value preserved by
2674 the server_setid option was incorrect.
2676 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2678 PH/09 Installed PCRE release 7.0.
2680 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2681 run for batched SMTP input. It is now run at the start of every message
2682 in the batch. While fixing this I discovered that the process information
2683 (output by running exiwhat) was not always getting set for -bs and -bS
2684 input. This is fixed, and it now also says "batched" for BSMTP.
2686 PH/11 Added control=no_pipelining.
2688 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2689 patch, slightly modified), and move the expansion of helo_data till after
2690 the connection is made in the smtp transport (so it can use these
2693 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2695 PH/14 Added log_selector = +pid.
2697 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2699 PH/16 Add ${if forany and ${if forall.
2701 PH/17 Added dsn_from option to vary the From: line in DSNs.
2703 PH/18 Flush SMTP output before performing a callout, unless control =
2704 no_callout_flush is set.
2706 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2707 was true (the default) a successful delivery failed to delete the retry
2708 item, thus causing premature timeout of the address. The bug is now
2711 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2713 PH/21 Long custom messages for fakedefer and fakereject are now split up
2714 into multiline responses in the same way that messages for "deny" and
2715 other ACL rejections are.
2717 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2718 with slight modification.
2720 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2721 draft, changing the syntax and factoring some duplicate code".
2723 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2724 for deliveries of the second and subsequent messages over the same SMTP
2727 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2728 ${reduce, with only minor "tidies".
2730 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2732 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2733 expansion side effects.
2735 PH/27 When a message times out after an over-quota error from an Exim-imposed
2736 quota, the bounce message says "mailbox is full". This message was not
2737 being given when it was a system quota that was exceeded. It now should
2740 MH/03 Made $recipients available in local_scan(). local_scan() already has
2741 better access to the recipient list through recipients_list[], but
2742 $recipients can be useful in postmaster-provided expansion strings.
2744 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2745 in the case of a MAIL command with additional options following the
2746 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2747 were accidentally chopped off.
2749 PH/29 SMTP synchronization checks are implemented when a command is read -
2750 there is a check that no more input is waiting when there shouldn't be
2751 any. However, for some commands, a delay in an ACL can mean that it is
2752 some time before the response is written. In this time, more input might
2753 arrive, invalidly. So now there are extra checks after an ACL has run for
2754 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2755 pipelining has not been advertised.
2757 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2759 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2760 returned in the SMTP response when smtp_return_error_details was false.
2761 This has been fixed.
2763 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2764 instead of the C I/O that was originally supplied, because problems were
2765 reported on Solaris.
2767 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2768 Exim which did not show up earlier: it was assuming that a call to
2769 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2770 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2771 was a macro that became an assignment, so it seemed to work. This has
2772 changed to a proper function call with a void return, hence the compile
2773 error. Exim's code has been fixed.
2775 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2778 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2780 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2782 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2783 because it turns out to be too big or there is a timeout) there is no
2784 "Completed" line in the log. When some messages of this type were
2785 selected by exigrep, they were listed as "not completed". Others were
2786 picked up by some special patterns. I have improved the selection
2787 criteria to be more general.
2789 PH/38 The host_find_failed option in the manualroute router can now be set
2790 to "ignore", to completely ignore a host whose IP address cannot be
2791 found. If all hosts are ignored, the behaviour is controlled by the new
2792 host_all_ignored option.
2794 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2795 homing or because of multiple MX records with /mx) generated more than
2796 one IP address, and the following item turned out to be the local host,
2797 all the secondary addresses of the first item were incorrectly removed
2798 from the list, along with the local host and any following hosts (which
2799 is what is supposed to happen).
2801 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2802 whoever called Exim into the -H file. In the case of the daemon it was
2803 behaving confusingly. When first started, it used values for whoever
2804 started the daemon, but after a SIGHUP it used the Exim user (because it
2805 calls itself on a restart). I have changed the code so that it now always
2808 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2809 message are rejected with the same error (e.g. no authentication or bad
2810 sender address), and a DATA command is nevertheless sent (as can happen
2811 with PIPELINING or a stupid MUA), the error message that was given to the
2812 RCPT commands is included in the rejection of the DATA command. This is
2813 intended to be helpful for MUAs that show only the final error to their
2816 PH/42 Another patch from the Sieve maintainer.
2818 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2819 Eximstats - Fixed some broken HTML links and added missing column headers
2821 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2822 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2824 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2826 PH/43 Yet another patch from the Sieve maintainer.
2828 PH/44 I found a way to check for a TCP/IP connection going away before sending
2829 the response to the final '.' that terminates a message, but only in the
2830 case where the client has not sent further data following the '.'
2831 (unfortunately, this is allowed). However, in many cases there won't be
2832 any further data because there won't be any more messages to send. A call
2833 to select() can be used: if it shows that the input is "ready", there is
2834 either input waiting, or the socket has been closed. An attempt to read
2835 the next input character can distinguish the two cases. Previously, Exim
2836 would have sent an OK response which the client would never have see.
2837 This could lead to message repetition. This fix should cure that, at
2838 least in a lot of common cases.
2840 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2841 advertised in response to EHLO.
2847 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2848 fixed by 4.65/MH/01 (is this a record?) are fixed:
2850 (i) An empty string was always treated as zero by the numeric comparison
2851 operators. This behaviour has been restored.
2853 (ii) It is documented that the numeric comparison operators always treat
2854 their arguments as decimal numbers. This was broken in that numbers
2855 starting with 0 were being interpreted as octal.
2857 While fixing these problems I realized that there was another issue that
2858 hadn't been noticed. Values of message_size_limit (both the global option
2859 and the transport option) were treated as octal if they started with 0.
2860 The documentation was vague. These values are now always treated as
2861 decimal, and I will make that clear in the documentation.
2867 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2868 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2871 MH/01 Don't check that the operands of numeric comparison operators are
2872 integers when their expansion is in "skipping" mode (fixes bug
2873 introduced by 4.64-PH/07).
2875 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2876 child addresses, Exim now panics and dies. Previously, because the count
2877 is held in a short int, deliveries were likely to be lost. As such a
2878 large number of recipients for a single message is ridiculous
2879 (performance will be very, very poor), I have chosen to impose a limit
2880 rather than extend the field.
2886 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2887 leftover -K file (the existence of which was triggered by #402).
2888 While we were at it, introduced process PID as part of the -K
2889 filename. This should rule out race conditions when creating
2892 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2893 processing considerably. Previous code took too long for large mails,
2894 triggering a timeout which in turn triggers #401.
2896 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2897 in the DK code in transports.c. sendfile() is not really portable,
2898 hence the _LINUX specificness.
2900 TF/01 In the add_headers option to the mail command in an Exim filter,
2901 there was a bug that Exim would claim a syntax error in any
2902 header after the first one which had an odd number of characters
2905 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2906 callout verification, Exim cached a "reject" for the entire domain. This
2907 is correct for most verifications, but it is not correct for a recipient
2908 verification with use_sender or use_postmaster set, because in that case
2909 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2910 case of MAIL FROM:<> rejection from other early rejections (e.g.
2911 rejection of HELO). When verifying a recipient using a non-null MAIL
2912 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2913 Whatever the result of the callout, the value of the domain cache is
2914 left unchanged (for any other kind of callout, getting as far as trying
2915 RCPT means that the domain itself is ok).
2917 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2920 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2921 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2922 session, the daemon ignores ECONNECT errors and logs others; it now
2923 ignores EPIPE as well.
2925 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2926 (quoted-printable decoding).
2928 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2929 later the small subsequent patch to fix an introduced bug.
2931 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2933 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2935 PH/08 An error is now given if message_size_limit is specified negative.
2937 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2938 to be given (somewhat) arbitrary names.
2940 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2943 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2944 miscellaneous code fixes
2946 PH/10 Added the log_reject_target ACL modifier to specify where to log
2949 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2950 hostname. This is wrong, because it relates to the incoming message (and
2951 probably the interface on which it is arriving) and not to the outgoing
2952 callout (which could be using a different interface). This has been
2953 changed to use the value of the helo_data option from the smtp transport
2954 instead - this is what is used when a message is actually being sent. If
2955 there is no remote transport (possible with a router that sets up host
2956 addresses), $smtp_active_hostname is used.
2958 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2959 tweaks were necessary in order to get it to work (see also 21 below):
2960 (a) The code assumed that strncpy() returns a negative number on buffer
2961 overflow, which isn't the case. Replaced with Exim's string_format()
2963 (b) There were several signed/unsigned issues. I just did the minimum
2964 hacking in of casts. There is scope for a larger refactoring.
2965 (c) The code used strcasecmp() which is not a standard C function.
2966 Replaced with Exim's strcmpic() function.
2967 (d) The code set only $1; it now sets $auth1 as well.
2968 (e) A simple test gave the error "authentication client didn't specify
2969 service in request". It would seem that Dovecot has changed its
2970 interface. Fortunately there's a specification; I followed it and
2971 changed what the client sends and it appears to be working now.
2973 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2976 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2977 address A is aliased to B and C, where B exists and C does not. Without
2978 -v the output is "A verified" because verification stops after a
2979 successful redirection if more than one address is generated. However,
2980 with -v the child addresses are also verified. Exim was outputting "A
2981 failed to verify" and then showing the successful verification for C,
2982 with its parentage. It now outputs "B failed to verify", showing B's
2983 parentage before showing the successful verification of C.
2985 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2986 look up a TXT record in a specific list after matching in a combined
2989 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2990 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2991 they consult the DNS. I had assumed they would set it the way they
2992 wanted; and indeed my experiments on Linux seem to show that in some
2993 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2994 To be on the safe side, however, I have now made the interface to
2995 host_find_byname() similar to host_find_bydns(), with an argument
2996 containing the DNS resolver options. The host_find_byname() function now
2997 sets these options at its start, just as host_find_bydns() does. The smtp
2998 transport options dns_qualify_single and dns_search_parents are passed to
2999 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3000 of host_find_byname() use the default settings of RES_DEFNAMES
3001 (qualify_single) but not RES_DNSRCH (search_parents).
3003 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3004 spool_read_header() do less string testing, by means of a preliminary
3005 switch on the second character of optional "-foo" lines. (This is
3006 overdue, caused by the large number of possibilities that now exist.
3007 Originally there were few.) While I was there, I also converted the
3008 str(n)cmp tests so they don't re-test the leading "-" and the first
3009 character, in the hope this might squeeze out yet more improvement.
3011 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3012 flag allowing group syntax was set by the header_syntax check but not
3013 turned off, possible causing trouble later; (2) The flag was not being
3014 set at all for the header_verify test, causing "group"-style headers to
3015 be rejected. I have now set it in this case, and also caused header_
3016 verify to ignore an empty address taken from a group. While doing this, I
3017 came across some other cases where the code for allowing group syntax
3018 while scanning a header line wasn't quite right (mostly, not resetting
3019 the flag correctly in the right place). These bugs could have caused
3020 trouble for malformed header lines. I hope it is now all correct.
3022 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3023 with the "reply" argument non-NULL. The code, however (which originally
3024 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3025 but it didn't always do it. This confused somebody who was copying the
3026 code for some other use. I have removed all the tests.
3028 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3029 feature that was used to support insecure browsers during the U.S. crypto
3030 embargo. It requires special client support, and Exim is probably the
3031 only MTA that supported it -- and would never use it because real RSA is
3032 always available. This code has been removed, because it had the bad
3033 effect of slowing Exim down by computing (never used) parameters for the
3034 RSA_EXPORT functionality.
3036 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3037 authenticator to fail if there's a tab character in the incoming data
3038 (there should never be unless someone is messing about, as it's supposed
3039 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3040 if the connection is using TLS or if the remote IP is the same as the
3041 local IP, and the "valid-client-cert option" if a client certificate has
3044 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3045 authenticators. This can be used for authorization after authentication
3046 succeeds. (In the case of plaintext, it servers for both authentication
3049 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3050 if any retry times were supplied.
3052 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3053 connection, where there is no client IP address to check. In this
3054 situation, the verify now always succeeds.
3056 PH/25 Applied John Jetmore's -Mset patch.
3058 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3060 PH/27 In a string expansion for a processed (not raw) header when multiple
3061 headers of the same name were present, leading whitespace was being
3062 removed from all of them, but trailing whitespace was being removed only
3063 from the last one. Now trailing whitespace is removed from each header
3064 before concatenation. Completely empty headers in a concatenation (as
3065 before) are ignored.
3067 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3068 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3070 PH/29 [Removed. This was a change that I later backed out, and forgot to
3071 correct the ChangeLog entry (that I had efficiently created) before
3072 committing the later change.]
3074 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3075 address errors (4xx response to RCPT) over the same connection as other
3076 messages routed to the same hosts. Such deliveries are always "forced",
3077 so retry times are not inspected. This resulted in far too many retries
3078 for the affected addresses. The effect occurred only when there were more
3079 hosts than the hosts_max_try setting in the smtp transport when it had
3080 the 4xx errors. Those hosts that it had tried were not added to the list
3081 of hosts for which the message was waiting, so if all were tried, there
3082 was no problem. Two fixes have been applied:
3084 (i) If there are any address or message errors in an SMTP delivery, none
3085 of the hosts (tried or untried) are now added to the list of hosts
3086 for which the message is waiting, so the message should not be a
3087 candidate for sending over the same connection that was used for a
3088 successful delivery of some other message. This seems entirely
3089 reasonable: after all the message is NOT "waiting for some host".
3090 This is so "obvious" that I'm not sure why it wasn't done
3091 previously. Hope I haven't missed anything, but it can't do any
3092 harm, as the worst effect is to miss an optimization.
3094 (ii) If, despite (i), such a delivery is accidentally attempted, the
3095 routing retry time is respected, so at least it doesn't keep
3096 hammering the server.
3098 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3099 in ${readsocket because some servers need this prod.
3101 PH/32 Added some extra debug output when updating a wait-xxx database.
3103 PH/33 The hint "could be header name not terminated by colon", which has been
3104 given for certain expansion errors for a long time, was not being given
3105 for the ${if def:h_colon_omitted{... case.
3107 PH/34 The spec says: "With one important exception, whenever a domain list is
3108 being scanned, $domain contains the subject domain." There was at least
3109 one case where this was not true.
3111 PH/35 The error "getsockname() failed: connection reset by peer" was being
3112 written to the panic log as well as the main log, but it isn't really
3113 panic-worthy as it just means the connection died rather early on. I have
3114 removed the panic log writing for the ECONNRESET error when getsockname()
3117 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3118 runs only) independently of the message's sender address. This meant
3119 that, if the 4xx error was in fact related to the sender, a different
3120 message to the same recipient with a different sender could confuse
3121 things. In particular, this can happen when sending to a greylisting
3122 server, but other circumstances could also provoke similar problems.
3123 I have changed the default so that the retry time for these errors is now
3124 based a combination of the sender and recipient addresses. This change
3125 can be overridden by setting address_retry_include_sender=false in the
3128 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3129 remote server are returned as part of bounce messages. This was not
3130 happening for LMTP over a pipe (the lmtp transport), but now it is the
3131 same for both kinds of LMTP.
3133 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3134 in header lines that were in fact CNAMEs. This is no longer the case.
3136 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3137 and queue runs started by the daemon processed all messages. This has
3138 been fixed so that -R and -S can now usefully be given with -q<time>.
3140 PH/40 Import PCRE release 6.7 (fixes some bugs).
3142 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3144 PH/42 Give an error if -q is specified more than once.
3146 PH/43 Renamed the variables $interface_address and $interface_port as
3147 $received_ip_address and $received_port, to make it clear that these
3148 values apply to message reception, and not to the outgoing interface when
3149 a message is delivered. (The old names remain recognized, of course.)
3151 PH/44 There was no timeout on the connect() call when using a Unix domain
3152 socket in the ${readsocket expansion. There now is.
3154 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3155 be meaningful with "accept".
3157 SC/01 Eximstats V1.43
3158 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3160 SC/02 Eximstats V1.44
3161 Use a glob alias rather than an array ref in the generated
3162 parser. This improves both readability and performance.
3164 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3165 Collect SpamAssassin and rejection statistics.
3166 Don't display local sender or destination tables unless
3167 there is data to show.
3168 Added average volumes into the top table text output.
3170 SC/04 Eximstats V1.46
3171 Collect data on the number of addresses (recipients)
3172 as well as the number of messages.
3174 SC/05 Eximstats V1.47
3175 Added 'Message too big' to the list of mail rejection
3176 reasons (thanks to Marco Gaiarin).
3178 SC/06 Eximstats V1.48
3179 Mainlog lines which have GMT offsets and are too short to
3180 have a flag are now skipped.
3182 SC/07 Eximstats V1.49 (Alain Williams)
3183 Added the -emptyok flag.
3185 SC/08 Eximstats V1.50
3186 Fixes for obtaining the IP address from reject messages.
3188 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3189 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3190 whitespace changes from 4.64-PH/27
3192 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3195 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3196 are found, allow negative numbers in numeric criteria)
3198 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3200 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3201 to match changes made in 4.64-PH/43
3203 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3205 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3206 "domain" is actually an IP address, and force a failure. This locks out
3207 those revolvers/nameservers that support "A-for-A" lookups, in
3208 contravention of the specifications.
3210 PH/48 When a host name was looked up from an IP address, and the subsequent
3211 forward lookup of the name timed out, the host name was left in
3212 $sender_host_name, contrary to the specification.
3214 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3215 restricted to single-key lookups, Exim was not diagnosing an error if
3216 * or *@ was used with a query-style lookup.
3218 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3220 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3221 long ago, but noone interested enough thought of it. Let's just say that
3222 the "1.1" means that there are some new functions that weren't there at
3223 some point in the past.
3225 PH/51 Error processing for expansion failure of helo_data from an smtp
3226 transport during callout processing was broken.
3228 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3229 tested/used via the -bh/-bhc/-bs options.
3231 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3232 bug, fixed in subsequent PCRE releases).
3234 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3235 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3237 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3243 SC/01 Use a glob alias rather than an array ref in eximstats generated
3244 parser. This improves both readability and performance.
3246 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3247 Don't display local sender or destination tables in eximstats unless
3248 there is data to show.
3249 Added average volumes into the eximstats top table text output.
3251 SC/03 Collect data on the number of addresses (recipients) as well
3252 as the number of messages in eximstats.
3254 TF/01 Correct an error in the documentation for the redirect router. Exim
3255 does (usually) call initgroups() when daemonizing.
3257 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3258 with consistent privilege compared to when running as a daemon.
3260 TF/03 Note in the spec that $authenticated_id is not set for local
3261 submissions from trusted users.
3263 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3264 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3266 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3267 by adding some example configuration directives to the default
3268 configuration file. A little bit of work is required to uncomment the
3269 directives and define how usernames and passwords are checked, but
3270 there is now a framework to start from.
3272 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3273 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3274 without this. I don't know how relevant this is to other LDAP libraries.
3276 PH/02 Add the verb name to the "unknown ACL verb" error.
3278 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3280 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3282 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3283 directory not expanded when it should be if an expanded home directory
3284 was set for the address (which is overridden by the transport).
3286 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3289 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3290 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3291 because it is too late at that time, and has no effect.
3293 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3294 security issue with \' (bugzilla #107). I could not use the
3295 PQescapeStringConn() function, because it needs a PGconn value as one of
3298 PH/08 When testing addresses using -bt, indicate those final addresses that
3299 are duplicates that would not cause an additional delivery. At least one
3300 person was confused, thinking that -bt output corresponded to deliveries.
3301 (Suppressing duplicates isn't a good idea as you lose the information
3302 about possibly different redirections that led to the duplicates.)
3304 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3305 systems where poll() doesn't work, in particular OS X.
3307 PH/10 Added more information to debugging output for retry time not reached.
3309 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3310 operations in malware.c.
3312 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3315 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3316 syslog_duplication set false, log lines that would normally be written
3317 both the the main log and to the reject log were not written to syslog at
3320 PH/14 In the default configuration, change the use of "message" in ACL warn
3321 statements to "add_header".
3323 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3324 not followed by a command (e.g. "seen endif").
3326 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3327 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3330 PH/17 Added extra conditions to the default value of delay_warning_condition
3334 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3335 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3336 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3339 The Auto-Submitted: and various List- headers are standardised, whereas I
3340 don't think Precedence: ever was.
3342 PH/18 Refactored debugging code in route_finduser() to show more information,
3343 in particular, the error code if getpwnam() issues one.
3345 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3346 This is apparently needed in addition to the PH/07 change above to avoid
3347 any possible encoding problems.
3349 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3350 but not after initializing Perl.
3352 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3353 output them only if debugging. By default they are written stderr,
3354 apparently, which is not desirable.
3356 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3359 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3362 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3364 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3365 authenticated or an ident call has been made. Suppress the default
3366 values for $authenticated_id and $authenticated_sender (but permit -oMai
3367 and -oMas) when testing with -bh.
3369 PH/24 Re-jigged the order of the tests in the default configuration so that the
3370 tests for valid domains and recipients precede the DNS black list and CSA
3371 tests, on the grounds that those ones are more expensive.
3373 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3374 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3375 This bug exists in every version of Exim that I still have, right back to
3378 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3379 However, an attempt to turn on case-sensitivity in a regex key by
3380 including (?-i) didn't work because the subject string was already
3381 lowercased, and the effects were non-intuitive. It turns out that a
3382 one-line patch can be used to allow (?-i) to work as expected.
3388 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3389 other effects) broke the use of negated acl sub-conditions.
3391 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3394 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3395 "Deny" causes Exim to reject the incoming connection with a 554 error.
3396 Unfortunately, if there is a major crisis, such as a disk failure,
3397 tcp-wrappers gives "deny", whereas what one would like would be some
3398 kind of temporary error. A kludge has been added to help with this.
3399 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3400 554 error is used if errno is still zero or contains ENOENT (which occurs
3401 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3404 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3406 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3407 errors. Otherwise a message that provokes a temporary error (when other
3408 messages do not) can cause a whole host to time out.
3410 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3411 addresses were routed directly to files or pipes from a redirect router.
3412 File deliveries just didn't batch; pipe deliveries might have suffered
3415 PH/06 A failure to get a lock for a hints database would erroneously always say
3416 "Failed to get write lock", even when it was really a read lock.
3418 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3419 of 0600. This has been changed to use the value of the lockfile_mode
3420 option (which defaults to 0600).
3422 PH/08 Applied small patch from the Sieve maintainer.
3424 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3425 folder from quota calculations, a direct delivery into this folder messed
3426 up the contents of the maildirsize file. This was because the regex was
3427 used only to exclude .Trash (or whatever) when the size of the mailbox
3428 was calculated. There was no check that a delivery was happening into an
3429 excluded directory. This bug has been fixed by ignoring all quota
3430 processing for deliveries into excluded directories.
3432 PH/10 Added the maildirfolder_create_regex option to appendfile.
3438 PH/01 The code for finding all the local interface addresses on a FreeBSD
3439 system running IPv6 was broken. This may well have applied to all BSD
3440 systems, as well as to others that have similar system calls. The broken
3441 code found IPv4 interfaces correctly, but gave incorrect values for the
3442 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3443 that it would not match correctly against @[] and not recognize the IPv6
3446 PH/02 The ipliteral router was not recognizing addresses of the form user@
3447 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3449 PH/03 Added disable_ipv6.
3451 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3452 decoded form, because it is most often used to construct To: headers
3453 lines in autoreplies, and the decoded form may well be syntactically
3454 invalid. However, $reply_address has leading white space removed, and all
3455 newlines turned into spaces so that the autoreply transport does not
3458 PH/05 If group was specified without a user on a router, and no group or user
3459 was specified on a transport, the group from the router was ignored.
3461 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3462 for visible compile-time settings that can be used to change these
3463 numbers, for those that want even more. Backwards compatibility with old
3464 spool files has been maintained. However, going back to a previous Exim
3465 release will lost any variables that are in spool files.
3467 PH/07 Two small changes when running in the test harness: increase delay when
3468 passing a TCP/IP connection to a new process, in case the original
3469 process has to generate a bounce, and remove special handling of
3470 127.0.0.2 (sic), which is no longer necessary.
3472 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3473 be the same on different OS.
3475 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3478 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3479 whether --show-vars was specified or not
3481 JJ/02 exipick: Added support for new ACL variable spool format introduced
3484 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3485 syntactically invalid From: or Reply-to: line, and a filter used this to
3486 generate an autoreply, and therefore failed to obtain an address for the
3487 autoreply, Exim could try to deliver to a non-existent relative file
3488 name, causing unrelated and misleading errors. What now happens is that
3489 it logs this as a hard delivery error, but does not attempt to create a
3492 PH/11 The exinext utility has a -C option for testing purposes, but although
3493 the given file was scanned by exinext itself; it wasn't being passed on
3494 when Exim was called.
3496 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3497 an end-of-file indication when reading a command response.
3499 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3500 compiled. In many other places in Exim, IPv6 addresses are always
3501 recognized, so I have changed this. It also means that IPv4 domain
3502 literals of the form [IPV4:n.n.n.n] are now always recognized.
3504 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3505 used if the router is not running as root, for example, when verifying at
3506 ACL time, or when using -bh. The debugging output from this situation was
3507 non-existent - all you got was a failure to exec. I have made two
3510 (a) Failures to set uid/gid, the current directory, or a process leader
3511 in a subprocess such as that created by queryprogram now generate
3512 suitable debugging output when -d is set.
3514 (b) The queryprogram router detects when it is not running as root,
3515 outputs suitable debugging information if -d is set, and then runs
3516 the subprocess without attempting to change uid/gid.
3518 PH/15 Minor change to Makefile for building test_host (undocumented testing
3521 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3522 additional section of a DNS packet that returns MX or SRV records.
3523 Instead, it always explicitly searches for A/AAAA records. This avoids
3524 major problems that occur when a DNS server includes only records of one
3525 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3526 fixed another bug: if SRV records were looked up and the corresponding
3527 address records were *not* found in the additional section, the port
3528 values from the SRV records were lost.
3530 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3531 using the correct key (the original address) when searching the retry
3532 rules in order to find which one to use for generating the retry hint.
3534 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3535 adding the default one. Similarly, if it contains a Reply-To: header, the
3536 errors_reply_to option, if set, is not used.
3538 PH/19 When calculating a retry time, Exim used to measure the "time since
3539 failure" by looking at the "first failed" field in the retry record. Now
3540 it does not use this if it is later than than the arrival time of the
3541 message. Instead it uses the arrival time. This makes for better
3542 behaviour in cases where some deliveries succeed, thus re-setting the
3543 "first failed" field. An example is a quota failure for a huge message
3544 when small messages continue to be delivered. Without this change, the
3545 "time since failure" will always be short, possible causing more frequent
3546 delivery attempts for the huge message than are intended.
3547 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3549 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3550 $1, $2, $3) because the numerical variables can be reset during some
3551 expansion items (e.g. "match"), thereby losing the authentication data.
3553 PH/21 Make -bV show the size of off_t variables so that the test suite can
3554 decide whether to run tests for quotas > 2G.
3556 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3557 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3558 filecount value is greater than 2G or if a quota value is greater than 2G
3559 on a system where the size of off_t is not greater than 4, a panic error
3562 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3563 never match. The debug and -bh output now contains an explicit error
3564 message indicating a malformed IPv4 address or mask.
3566 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3567 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3568 PH/23 above applies.
3570 PH/25 Do not write to syslog when running in the test harness. The only
3571 occasion when this arises is a failure to open the main or panic logs
3572 (for which there is an explicit test).
3574 PH/26 Added the /no_tell option to "control=freeze".
3576 PH/27 If a host name lookup failed very early in a connection, for example, if
3577 the IP address matched host_lookup and the reverse lookup yielded a name
3578 that did not have a forward lookup, an error message of the form "no IP
3579 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3580 could be logged. Now it outputs the IP address instead of "NULL".
3582 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3583 allows the sender and the authenticated sender to be set when
3584 submitting a message from within Exim. Since child_open_exim() is
3585 documented for local_scan(), the new function should be too.
3587 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3588 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3589 results in an empty string is now treated as unset.
3591 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3593 PH/31 Added sender_verify_fail logging option.
3595 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3596 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3597 tidied the source and removed it altogether.
3599 PH/33 When a queue run was abandoned because the load average was too high, a
3600 log line was always written; now it is written only if the queue_run log
3601 selector is set. In addition, the log line for abandonment now contains
3602 information about the queue run such as the pid. This is always present
3603 in "start" and "stop" lines but was omitted from the "abandon" line.
3605 PH/34 Omit spaces between a header name and the colon in the error message that
3606 is given when verify = headers_syntax fails (if there are lots of them,
3607 the message gets confusing).
3609 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3610 names, as there are now some PTR records that contain slashes. This check
3611 is only to protect against broken name servers that fall over on strange
3612 characters, so the fact that it applies to all lookups doesn't matter.
3614 PH/36 Now that the new test suite is complete, we can remove some of the
3615 special code in Exim that was needed for the old test suite. For example,
3616 sorting DNS records because real resolvers return them in an arbitrary
3617 order. The new test suite's fake resolver always returns records in the
3620 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3621 bounces) except when queue_only is set, to avoid logging races between
3622 the different processes.
3624 PH/38 Panic-die if .include specifies a non-absolute path.
3626 PH/39 A tweak to the "H" retry rule from its user.
3628 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3629 a label. They prevented compilation on older perls.
3631 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3632 a warning to be raised on newish perls.
3634 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3635 on queue. Changes to match documented behaviour of showing count of
3636 messages matching specified criteria.
3638 PH/40 Changed the default ident timeout from 30s to 5s.
3640 PH/41 Added support for the use of login_cap features, on those BSD systems
3641 that have them, for controlling the resources used by pipe deliveries.
3643 PH/42 The content-scanning code uses fopen() to create files in which to put
3644 message data. Previously it was not paying any attention to the mode of
3645 the files. Exim runs with umask(0) because the rest of the code creates
3646 files with open(), and sets the required mode explicitly. Thus, these
3647 files were ending up world-writeable. This was not a big issue, because,
3648 being within the spool directory, they were not world-accessible. I have
3649 created a function called modefopen, which takes an additional mode
3650 argument. It sets umask(777), creates the file, chmods it to the required
3651 mode, then resets the umask. All the relevant calls to fopen() in the
3652 content scanning code have been changed to use this function.
3654 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3655 to 24 hours. This avoids potential overflow problems when processing G
3656 and H retry rules. I suspect nobody ever tinkers with this value.
3658 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3660 PH/45 When the plaintext authenticator is running as a client, the server's
3661 challenges are checked to ensure they are valid base64 strings. By
3662 default, the authentication attempt is cancelled if an invalid string is
3663 received. Setting client_ignore_invalid_base64 true ignores these errors.
3664 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3665 they are received. Thus, the responses can be made to depend on the
3666 challenges. If an invalid string is ignored, an empty string is placed in
3669 PH/46 Messages that are created by the autoreply transport now contains a
3670 References: header, in accordance with RFCs 2822 and 3834.
3672 PH/47 Added authenticated_sender_force to the smtp transport.
3674 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3676 PH/49 Installed latest patch from the Sieve maintainer.
3678 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3679 was also set, the appendfile transport was unnecessarily scanning a
3680 directory of message files (e.g. for maildir delivery) to find the count
3681 of files (along with the size), even though it did not need this
3682 information. It now does the scan only if it needs to find either the
3683 size of the count of files.
3685 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3687 PH/52 Two bugs concerned with error handling when the smtp transport is
3690 (i) Exim was not creating retry information for temporary errors given
3691 for individual recipients after the DATA command when the smtp transport
3692 was used in LMTP mode. This meant that they could be retried too
3693 frequently, and not timed out correctly.
3695 (ii) Exim was setting the flag that allows error details to be returned
3696 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3697 recipients that were returned after the DATA command.
3699 PH/53 This is related to PH/52, but is more general: for any failing address,
3700 when detailed error information was permitted to be returned to the
3701 sender, but the error was temporary, then after the final timeout, only
3702 "retry timeout exceeded" was returned. Now it returns the full error as
3703 well as "retry timeout exceeded".
3705 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3706 do this, and (what is worse) MTAs that accept it.
3708 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3709 will now be deprecated.
3711 PH/56 New os.c-cygwin from the Cygwin maintainer.
3713 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3714 formats (previously only available in exim formats via -bpr, -bpru,
3715 and -bpra. Now also available in native and exiqgrep formats)
3717 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3718 with very large, slow to parse queues
3720 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3722 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3724 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3725 responses to authentication challenges, though it was showing the
3726 challenges; (ii) I've removed the CR characters from the debug output for
3729 PH/58 Allow for the insertion of a newline as well as a space when a string
3730 is turned into more than one encoded-word during RFC 2047 encoding. The
3731 Sieve code now uses this.
3733 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3734 data_4xx, lost_connection, tls_required.
3736 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3737 message was being sent as an SMTP response.
3739 PH/61 Add -l and -k options to exicyclog.
3741 PH/62 When verifying, if an address was redirected to one new address, so that
3742 verification continued, and the new address failed or deferred after
3743 having set something in $address_data, the value of $address_data was not
3744 passed back to the ACL. This was different to the case when no
3745 redirection occurred. The value is now passed back in both cases.
3747 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3748 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3749 use login_cap.h, so on its own it isn't the distinguishing feature. The
3750 new name refers directly to the setclassresources() function.
3752 PH/65 Added configuration files for NetBSD3.
3754 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3756 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3757 is preferred over IPv4.
3759 PH/68 The bounce_return_message and bounce_return_body options were not being
3760 honoured for bounces generated during the reception of non-SMTP messages.
3761 In particular, this applied to messages rejected by the ACL. This bug has
3762 been fixed. However, if bounce_return_message is true and bounce_return_
3763 body is false, the headers that are returned for a non-SMTP message
3764 include only those that have been read before the error was detected.
3765 (In the case of an ACL rejection, they have all been read.)
3767 PH/69 The HTML version of the specification is now built in a directory called
3768 spec_html instead of spec.html, because the latter looks like a path with
3769 a MIME-type, and this confuses some software.
3771 PH/70 Catch two compiler warnings in sieve.c.
3773 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3774 function verify_get_ident() calls ip_connect() to connect a socket, but
3775 if the "connect()" function timed out, ip_connect() used to close the
3776 socket. However, verify_get_ident() also closes the socket later, and in
3777 between Exim writes to the log, which may get opened at this point. When
3778 the socket was closed in ip_connect(), the log could get the same file
3779 descriptor number as the socket. This naturally causes chaos. The fix is
3780 not to close the socket in ip_connect(); the socket should be closed by
3781 the function that creates it. There was only one place in the code where
3782 this was missing, in the iplookup router, which I don't think anybody now
3783 uses, but I've fixed it anyway.
3785 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3786 well as to direct DNS lookups. Otherwise the handling of names in host
3787 lists is inconsistent and therefore confusing.
3793 PH/01 Two changes to the default runtime configuration:
3795 (1) Move the checks for relay_from_hosts and authenticated clients from
3796 after to before the (commented out) DNS black list checks.
3798 (2) Add control=submission to the relay_from_hosts and authenticated
3799 clients checks, on the grounds that messages accepted by these
3800 statements are most likely to be submissions.
3802 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3804 (1) Generate an error if the third argument for the ${prvs expansion is
3807 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3810 (3) Reset the variables that are obtained from the first argument of
3811 ${prvscheck and used in the second argument before leaving the code,
3812 because their memory is reclaimed, so using them afterwards may do
3815 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3816 one (it's much easier than Tom thought :-).
3818 (5) Because of (4), we can now allow for the use of $prvscheck_result
3819 inside the third argument.
3821 PH/03 For some reason, the default setting of PATH when running a command from
3822 a pipe transport was just "/usr/bin". I have changed it to
3825 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3826 anything to be listed in the output from -bV.
3828 PH/05 When a filter generated an autoreply, the entire To: header line was
3829 quoted in the delivery log line, like this:
3831 => >A.N.Other <ano@some.domain> <original@ddress> ...
3833 This has been changed so that it extracts the operative address. There
3834 may be more than one such address. If so, they are comma-separated, like
3837 => >ano@some.domain,ona@other.domain <original@ddress> ...
3839 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3840 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3841 not being looked up in the rDNS to get a host name, Exim was showing the
3842 IP address twice in Received: lines, even though the IP addresses were
3843 identical. For example:
3845 Received: from [1.2.3.4] (helo=[1.2.3.4])
3847 However, if the real host name was known, it was omitting the HELO data
3848 if it matched the actual IP address. This has been tidied up so that it
3849 doesn't show the same IP address twice.
3851 PH/07 When both +timestamp and +memory debugging was on, the value given by
3852 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3853 called by the debug printing, thereby overwriting the timestamp buffer.
3854 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3856 PH/08 When the original message was included in an autoreply transport, it
3857 always said "this is a copy of the message, including all the headers",
3858 even if body_only or headers_only was set. It now gives an appropriate
3861 PH/09 Applied a patch from the Sieve maintainer which:
3863 o fixes some comments
3864 o adds the (disabled) notify extension core
3865 o adds some debug output for the result of if/elsif tests
3866 o points to the current vacation draft in the documentation
3867 and documents the missing references header update
3871 o fixes a bug in processing the envelope test (when testing
3872 multiple envelope elements, the last element determined the
3875 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3876 Electronic Mail") by including:
3878 Auto-submitted: auto-generated
3880 in the messages that it generates (bounce messages and others, such as
3881 warnings). In the case of bounce messages for non-SMTP messages, there was
3882 also a typo: it was using "Auto_submitted" (underscore instead of
3883 hyphen). Since every message generated by Exim is necessarily in response
3884 to another message, thes have all been changed to:
3886 Auto-Submitted: auto-replied
3888 in accordance with these statements in the RFC:
3890 The auto-replied keyword:
3892 - SHOULD be used on messages sent in direct response to another
3893 message by an automatic process,
3895 - MUST NOT be used on manually-generated messages,
3897 - MAY be used on Delivery Status Notifications (DSNs) and Message
3898 Disposition Notifications (MDNs),
3900 - MUST NOT be used on messages generated by automatic or periodic
3901 processes, except for messages which are automatic responses to
3904 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3905 to the default Received: header definition.
3907 PH/12 Added log selector acl_warn_skipped (default on).
3909 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3910 variables because (a) they are in the wrong storage pool and (b) even if
3911 they were copied, it wouldn't work properly because of the caching.
3913 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3914 checking when decoding. Apparently there are clients that generate
3915 overlong encoded strings. Why am I not surprised?
3917 PH/15 If the first argument of "${if match_address" was not empty, but did not
3918 contain an "@" character, Exim crashed. Now it writes a panic log message
3919 and treats the condition as false.
3921 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3923 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3924 extension "envelope-auth". The code is finished and in agreement with
3925 other implementations, but there is no documentation so far and in fact,
3926 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3927 not changing the active code.
3929 Print executed "if" and "elsif" statements when debugging is used. This
3930 helps a great deal to understand what a filter does.
3932 Document more things not specified clearly in RFC3028. I had all this
3933 sorted out, when out of a sudden new issues came to my mind. Oops."
3935 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3938 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3939 canonical form (as documented). However, after a host name lookup from
3940 the IP address, check_host() was doing a simple string comparison with
3941 addresses acquired from the DNS when checking that the found name did
3942 have the original IP as one of its addresses. Since any found IPv6
3943 addresses are likely to be in abbreviated form, the comparison could
3944 fail. Luckily, there already exists a function for doing the comparison
3945 by converting both addresses to binary, so now that is used instead of
3946 the text comparison.
3948 PH/20 There was another similar case to PH/19, when a complete host name was
3949 given in a host list; looking up its IP address could give an abbreviated
3950 form, whereas the current host's name might or might not be abbreviated.
3951 The same fix has been applied.
3957 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3958 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3961 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3962 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3964 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3966 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3967 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3968 Also, for "Auto-Submitted", treat anything other than "no" as
3969 non-personal, in accordance with RFC 3834. (Previously it treated
3970 anything starting "auto-" as non-personal.)
3972 TF/01 The control=submission/name=... option had a problem with syntax
3973 errors if the name included a slash character. The /name= option
3974 now slurps the rest of the string, so it can include any characters
3975 but it must come last in the list of options (after /sender_retain
3978 PH/05 Some modifications to the interface to the fake nameserver for the new
3986 TK/01 Added the "success_on_redirect" address verification option. See
3987 NewStuff for rationale and an example.
3989 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3991 PH/02 Patch to exigrep to allow it to work on syslog lines.
3993 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3994 fread() to skip over the body file's header line, because in Cygwin the
3995 header line is locked and is inaccessible.
3997 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3998 co-exist for some time) to make it clear that it is the Exim ID that is
3999 referenced, not the Message-ID: header line.
4001 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4002 string_format() function, because snprintf() does not exist on all
4005 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4006 ${stat: expansion item.
4008 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4009 protocol synchronization error", to keep the pedants happy.
4011 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4012 well as for IRIX systems, when gcc is being used. See the host.c source
4015 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4017 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4020 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4021 transport and to the smtp transport in LMTP mode.
4023 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4025 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4026 run a filter in a subprocess. This could lead to confusion in subsequent
4027 lookups in the parent process. There should also be a search_tidyup() at
4028 the end of the subprocess.
4030 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4031 only if the host matched helo_try_verify_hosts, which caused the
4032 verification to occur when the EHLO/HELO command was issued. The ACL just
4033 tested the remembered result. Now, if a previous verification attempt has
4034 not happened, "verify = helo" does it there and then.
4036 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4038 TK/03 Fix log output including CR from clamd.
4040 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4041 exist provoked a memory error which could cause a segfault.
4043 PH/15 Installed PCRE 6.2
4045 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4047 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4048 of the problem. Specifically, suggested +O2 rather than +O1 for the
4051 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4053 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4054 record, the port was not correctly logged when the outgoing_port log
4055 selector was set (it logged the transort's default port).
4057 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4058 fallback_hosts, and "hosts" in the smtp transport.
4060 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4061 host errors such as "Connection refused".
4063 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4064 authentication with radiusclient 0.4.9:
4066 - Error returned from rc_read_config was caught wrongly
4067 - Username/password not passed on to radius server due to wrong length.
4069 The presumption is that some radiusclient API changes for 4.51/PH/17
4070 were not taken care of correctly. The code is still untested by me (my
4071 Linux distribution still has 0.3.2 of radiusclient), but it was
4072 contributed by a Radius user.
4074 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4075 expanding the "port" option of the smtp transport.
4077 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4078 while reading a MIME header. Thanks to Tom Hughes for a patch.
4080 PH/24 Include config.h inside local_scan.h so that configuration settings are
4083 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4084 that in an ACL for RCPT (for example), you can examine exactly what was
4087 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4088 commands, but it was not correctly comparing the address with the actual
4089 client host address. Thus, it would show the EHLO address in Received:
4090 header lines when this was not necessary.
4092 PH/27 Added the % operator to ${eval:}.
4094 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4095 it should be ignoring failures (because with -C, for example, it has lost
4096 privilege). It wasn't ignoring creation failures other than "already
4099 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4102 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4103 because an input line was too long, either on its own, or by virtue of
4104 too many continuations, the temporary file was not being removed, and the
4105 return code was incorrect.
4107 PH/31 Missing "BOOL" in function definition in filtertest.c.
4109 PH/32 Applied Sieve patches from the maintainer.
4111 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4113 PH/33 Added "verify = not_blind".
4115 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4116 Local/Makefile (with some defaults set). These are used in built scripts
4117 such as exicyclog, but they have never been used in the exim_install
4118 script (though there are many overriding facilities there). I have
4119 arranged that the exim_install script now takes note of these two
4122 PH/35 Installed configuration files for Dragonfly.
4124 PH/36 When a locally submitted message by a trusted user did not contain a
4125 From: header, and the sender address was obtained from -f or from an SMTP
4126 MAIL command, and the trusted user did not use -F to supply a sender
4127 name, $originator_name was incorrectly used when constructing a From:
4128 header. Furthermore, $originator_name was used for submission mode
4129 messages from external hosts without From: headers in a similar way,
4130 which is clearly wrong.
4132 PH/37 Added control=suppress_local_fixups.
4134 PH/38 When log_selector = +received_sender was set, and the addition of the
4135 sender made the log line's construction buffer exactly full, or one byte
4136 less than full, an overflow happened when the terminating "\n" was
4139 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4140 when the result of a list match is failure because a DNS lookup failed.
4142 PH/40 RM_COMMAND is now used in the building process.
4144 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4145 the "build-* directories that it finds.
4147 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4148 address in a domain literal was a prefix of an interface address.
4150 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4151 when verifying a sender address, unless rewrite_headers is false.
4153 PH/44 Wrote a long comment about why errors_to addresses are verified as
4154 recipients, not senders.
4156 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4157 the ratelimit ACL was added.
4159 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4161 PH/46 Added extra information about PostgreSQL errors to the error string.
4163 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4164 avoiding the need to install special zones in a real server. This is
4165 backwards compatible; if it can't find the fake resolver, it drops back.
4166 Thus, both old and new test suites can be run.
4168 TF/02 Added util/ratelimit.pl
4170 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4171 clock is set back in time.
4173 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4174 Candler <B.Candler@pobox.com>.
4176 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4177 OK for addresses that are the result of redirections.
4179 PH/48 A number of further additions for the benefit of the new test suite,
4180 including a fake gethostbyname() that interfaces to the fake DNS resolver
4183 TF/06 The fix for widen_domains has also been applied to qualify_single and
4184 search_parents which are the other dnslookup options that can cause
4187 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4190 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4192 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4193 service" if it encountered an MX record with an empty target hostname.
4194 The message is now "an MX or SRV record indicated no SMTP service".
4196 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4197 if the DNS of the sending site is misconfigured. This is quite a
4198 common situation. This change restores the behaviour of treating a
4199 helo verification defer as a failure.
4201 PH/51 If self=fail was set on a router, the bounce message did not include the
4202 actual error message.
4208 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4210 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4211 command itself ended in error, the underlying message about the transport
4212 filter timeout was being overwritten with the pipe command error. Now the
4213 underlying error message should be appended to the second error message.
4215 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4217 PH/02 Reduce the amount of output that "make" produces by default. Full output
4218 can still be requested.
4220 PH/03 The warning log line about a condition test deferring for a "warn" verb
4221 was being output only once per connection, rather than after each
4222 occurrence (because it was using the same function as for successful
4223 "warn" verbs). This seems wrong, so I have changed it.
4225 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4226 it should not have, which might have caused a crash in the right
4227 circumstances, but probably never did.
4229 PH/04 Installed a modified version of Tony Finch's patch to make submission
4230 mode fix the return path as well as the Sender: header line, and to
4231 add a /name= option so that you can make the user's friendly name appear
4234 TF/03 Added the control = fakedefer ACL modifier.
4236 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4237 Mark Lowes for thorough testing.
4239 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4241 TK/03 Merged latest SRS patch from Miles Wilton.
4243 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4244 with the definition in sysexits.h (which is #included earlier).
4245 Fortunately, Exim does not actually use EX_OK. The code used to try to
4246 preserve the sysexits.h value, by assuming that macro definitions were
4247 scanned for macro replacements. I have been disabused of this notion,
4248 so now the code just undefines EX_OK before #including unistd.h.
4250 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4251 in the smtp transport. When a block could not be written in a single
4252 write() function, the timeout was being re-applied to each part-write.
4253 This seems wrong - if the receiver was accepting one byte at a time it
4254 would take for ever. The timeout is now adjusted when this happens. It
4255 doesn't have to be particularly precise.
4257 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4258 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4260 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4261 without a domain if the check to <postmaster@domain> fails.
4263 SC/01 Eximstats: added -xls and the ability to specify output files
4264 (patch written by Frank Heydlauf).
4266 SC/02 Eximstats: use FileHandles for outputting results.
4268 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4270 SC/04 Eximstats: fixed display of large numbers with -nvr option
4272 SC/05 Eximstats: fixed merging of reports with empty tables.
4274 SC/06 Eximstats: added the -include_original_destination flag
4276 SC/07 Eximstats: removed tabs and trailing whitespace.
4278 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4280 TK/06 MBOX spool code: Add real "From " MBOX separator line
4281 so the .eml file is really in mbox format (even though
4282 most programs do not really care). Patch from Alex Miller.
4284 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4285 The latter is generated from $received_to and is only set if the
4286 message has one envelope recipient. SA can use these headers,
4287 obviously out-of-the-box. Patch from Alex Miller.
4289 PH/08 The ${def test on a variable was returning false if the variable's
4290 value was "0", contrary to what the specification has always said!
4291 The result should be true unless the variable is empty.
4293 PH/09 The syntax error of a character other than { following "${if
4294 def:variable_name" (after optional whitespace) was not being diagnosed.
4295 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4296 accidental colon was present, for example, could give incorrect results.
4298 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4299 result is used (not including appendfile, where other changes are about
4302 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4303 This involved changing a lot of size variables from int to off_t. It
4304 should work with maildirs and everything.
4306 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4307 spamd dying while we are connected to it.
4309 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4312 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4313 function for BDB 4.3.
4315 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4317 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4318 the sources for PCRE have been re-arranged and more files are now
4321 PH/15 The code I had for printing potentially long long variables in PH/11
4322 above was not the best (it lost precision). The length of off_t variables
4323 is now inspected at build time, and an appropriate printing format (%ld
4324 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4325 to be "long long int" or "long int". This is needed for the internal
4326 formatting function string_vformat().
4328 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4329 the configuration file to be ":syslog", then the script "guesses" where
4330 the logs files are, rather than using the compiled in default. In our
4331 case the guess is not the same as the compiled default, so the script
4332 suddenly stopped working when I started to use syslog. The patch checks
4333 to see if log_file_path is "". If so, it attempts to read it from exim
4334 with no configuration file to get the compiled in version, before it
4335 falls back to the previous guessing code."
4337 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4338 implementing BATV in an Exim configuration. See NewStuff for the gory
4341 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4342 Makefile that are specific to HP-UX.
4344 PH/18 If the "use_postmaster" option was set for a recipient callout together
4345 with the "random" option, the postmaster address was used as the MAIL
4346 FROM address for the random test, but not for the subsequent recipient
4347 test. It is now used for both.
4349 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4350 patch removes a few documentation additions to RFC 3028, because the
4351 latest draft now contains them. It adds the new en;ascii-case comparator
4352 and a new error check for 8bit text in MIME parts. Comparator and
4353 require names are now matched exactly. I enabled the subaddress
4354 extension, but it is not well tested yet (read: it works for me)."
4356 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4357 rework some of the code of TK/09 above to avoid the hardwired use of
4358 "%lld" and "long long". Replaced the call to snprintf() with a call to
4361 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4362 records point to non-existent hosts", "retry timeout exceeded", and
4363 "retry time not reached for any host after a long failure period".
4365 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4366 experimental DomainKeys support:
4368 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4369 (2) On an error such as an illegally used "control", the wrong name for
4370 the control was given.
4372 These problems did NOT occur unless DomainKeys support was compiled.
4374 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4376 PH/24 Added ${if match_ip condition.
4378 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4379 databases so that it will be absolutely obvious if a crash occurs in the
4380 DB library. This is a regular occurrence (often caused by mis-matched
4383 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4384 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4385 picked up on a user's system that detects such things. There doesn't seem
4386 to be a gcc warning option for this - only an attribute that has to be
4387 put on the function's prototype. It seems that in Fedora Core 4 they have
4388 set this on a number of new functions. No doubt there will be more in due
4391 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4392 specify a transport. However, if an address that was verified by such a
4393 router was the subject of a callout, Exim crashed because it tried to
4394 read the rcpt_include_affixes from the non-existent transport. Now it
4395 just assumes that the setting of that option is false. This bug was
4396 introduced by 4.51/PH/31.
4398 PH/28 Changed -d+all to exclude +memory, because that information is very
4399 rarely of interest, but it makes the output a lot bigger. People tend to
4400 do -d+all out of habit.
4402 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4403 code in os-type was giving problems when libc.so lives in lib64, like on
4406 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4407 aren't the modern standard, and it seems that some systems' include files
4408 don't always have them. Exim was already checking for some of the newer
4409 ones like T_AAAA, and defining it itself. I've added checks for all the
4410 record types that Exim uses.
4412 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4413 not automatically generating a new one, as it is supposed to. This
4414 prevented TLS from working. If the file did exist, but contained invalid
4415 data, a new version was generated, as expected. It was only the case of a
4416 non-existent file that was broken.
4418 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4419 with a change in libdomainkeys > 0.64.
4421 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4422 from DNS. If the selector record carries the flag, it now has
4423 precedence over the domain-wide flag.
4425 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4427 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4428 the use of a port name that isn't defined in /etc/services) occurred, the
4429 message was deferred as in a normal delivery, and thus remained on the
4430 spool, instead of being failed because of the mua_wrapper setting. This
4431 is now fixed, and I tidied up some of the mua_wrapper messages at the
4434 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4435 the messages in a hash of arrays rather than using individual hashes.
4436 This is a bit cleaner and results in dramatic memory savings, albeit
4437 at a slight CPU cost.
4439 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4440 as requested by Marc Sherman.
4442 SC/10 Eximstats: added histograms for user specified patterns as requested
4445 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4447 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4448 fopen() in the content-scanning modules that did not already have it.
4454 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4455 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4457 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4459 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4461 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4462 to test Sieve filters that use "vacation".
4464 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4465 that changes the way the GnuTLS parameters are stored in the cache file.
4466 The new format can be generated externally. For backward compatibility,
4467 if the data in the cache doesn't make sense, Exim assumes it has read an
4468 old-format file, and it generates new data and writes a new file. This
4469 means that you can't go back to an older release without removing the
4472 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4473 work if there are any delivery delays because "one_time" forces the
4474 parent to be marked "delivered", so its unseen clone is never tried
4475 again. For this reason, Exim now forbids the simultaneous setting of
4478 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4479 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4480 entry if you want to know the details. The fix, however, neglected to
4481 consider the case when local delivery batching is involved. The test for
4482 "previously delivered" was not happening when checking to see if an
4483 address could be batched with a previous (undelivered) one; under
4484 certain circumstances this could lead to multiple deliveries to the same
4487 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4488 in its include files, and this causes problems building Exim.
4490 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4491 header_syntax) but Exim was just ignoring anything given after a slash.
4492 In particular, this caused confusion with an attempt to use "verify =
4493 reverse_host_lookup/defer_ok". An error is now given when options are
4494 supplied for verify items that do not have them. (Maybe reverse_host_
4495 lookup should have a defer_ok option, but that's a different point.)
4497 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4498 defined by RFC 821) to 2048, because there were problems with some AUTH
4499 commands, and RFC 1869 says the size should be increased for extended
4500 SMTP commands that take arguments.
4502 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4505 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4506 "unknown" error; now it says that the functionality isn't in the binary.
4508 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4509 an address' error message when a string expansion fails (syntax or
4510 whatever). Otherwise the password may appear in the log. Following change
4511 PH/42 below, there is no longer a chance of it appearing in a bounce
4514 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4516 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4517 of its addresses was ever tried. (Bugzilla bug #2.)
4519 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4520 the result incorrectly in the debug output. (It correctly added a newline
4521 to what was transported.)
4523 TF/01 Added $received_time.
4525 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4526 commented out examples of how to interface to a virus scanner and to
4527 SpamAssassin. Also added commented examples of av_scanner and
4528 spamd_address settings.
4530 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4531 and controls are allowed in which ACLs. There were a couple of minor
4532 errors. Some of the entries in the conditions table (which is a table of
4533 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4534 negation of where the condition IS allowed.
4536 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4538 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4539 header file does not have a version number, so I've had to invent a new
4540 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4541 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4542 radiusclient), but it was contributed by a Radius user.
4544 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4545 files or directories, for OpenSSL.
4547 PH/19 When an Exim process that is running as root has to create an Exim log
4548 file, it does so in a subprocess that runs as exim:exim so as to get the
4549 ownership right at creation (otherwise, other Exim processes might see
4550 the file with the wrong ownership). There was no test for failure of this
4551 fork() call, which would lead to the process getting stuck as it waited
4552 for a non-existent subprocess. Forks do occasionally fail when resources
4553 run out. I reviewed all the other calls to fork(); they all seem to check
4556 PH/20 When checking for unexpected SMTP input at connect time (before writing
4557 the banner), Exim was not dealing correctly with a non-positive return
4558 from the read() function. If the client had disconnected by this time,
4559 the result was a log entry for a synchronization error with an empty
4560 string after "input=" when read() returned zero. If read() returned -1
4561 (an event I could not check), uninitialized data bytes were printed.
4562 There were reports of junk text (parts of files, etc) appearing after
4565 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4567 PH/22 Added support for macro redefinition, and (re)definition in between
4568 driver and ACL definitions.
4570 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4571 forgetting to use the resulting value; it was using the unexpanded value.
4573 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4574 hadn't been configured. The fix is from Juergen Kreileder, who
4575 understands it better than I do:
4577 "Here's what I see happening with three configured cyrus_sasl
4578 authenticators configured (plain, login, cram-md5):
4580 On startup auth_cyrus_sasl_init() gets called for each of these.
4581 This means three calls to sasl_listmech() without a specified mech_list.
4582 => SASL tests which mechs of all available mechs actually work
4583 => three warnings about OTP not working
4584 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4586 With the patch, sasl_listmech() also gets called three times. But now
4587 SASL's mech_list option is set to the server_mech specified in the the
4588 authenticator. Or in other words, the answer from sasl_listmech()
4589 gets limited to just the mech you're testing for (which is different
4591 => the return list contains just 'plain' or 'login', 'cram-md5' or
4592 nothing depending on the value of ob->server_mech.
4594 I've just tested the patch: Authentication still works fine,
4595 unavailable mechs specified in the exim configuration are still
4596 caught, and the auth.log warnings about OTP are gone."
4598 PH/25 When debugging is enabled, the contents of the command line are added
4599 to the debugging output, even when log_selector=+arguments is not
4602 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4603 answer is "GNU", and only if the return is "GNU/something" is the answer
4606 PH/27 $acl_verify_message is now set immediately after the failure of a
4607 verification in an ACL, and so is available in subsequent modifiers. In
4608 particular, the message can be preserved by coding like this:
4610 warn !verify = sender
4611 set acl_m0 = $acl_verify_message
4613 Previously, $acl_verify_message was set only while expanding "message"
4614 and "log_message" when a very denied access.
4616 PH/28 Modified OS/os.c-Linux with
4618 -#ifndef OS_LOAD_AVERAGE
4619 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4621 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4622 nomenclature these days.)
4624 PH/29 Installed patch from the Sieve maintainer that adds the options
4625 sieve_useraddress and sieve_subaddress to the redirect router.
4627 PH/30 In these circumstances:
4628 . Two addresses routed to the same list of hosts;
4629 . First host does not offer TLS;
4630 . First host accepts first address;
4631 . First host gives temporary error to second address;
4632 . Second host offers TLS and a TLS session is established;
4633 . Second host accepts second address.
4634 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4635 and peerdn, if requested) that were in fact used only for the second
4638 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4639 attention to any local part prefix or suffix that was matched by the
4640 router that accepted the address. It now behaves in the same way as it
4641 does for delivery: the affixes are removed from the local part unless
4642 rcpt_include_affixes is set on the transport.
4644 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4645 timeout during the DATA phase of an incoming message.
4647 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4648 applied a patch sanctioned by the Sieve maintainer.
4650 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4651 the uid or gid is negative. A case of a negative gid caused this to be
4652 noticed. The fix allows for either to be negative.
4654 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4655 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4658 PH/36 The change PH/12 above was broken. Fixed it.
4660 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4661 the grounds that routing the same address twice would always produce the
4662 same answer. This might have been true once, but it is certainly no
4663 longer true now. Routing a child address may depend on the previous
4664 routing that produced that child. Some complicated redirection strategies
4665 went wrong when messages had multiple recipients, and made Exim's
4666 behaviour dependent on the order in which the addresses were given.
4668 I have moved the duplicate checking until after the routing is complete.
4669 Exim scans the addresses that are assigned to local and remote
4670 transports, and removes any duplicates. This means that more work will be
4671 done, as duplicates will always all be routed, but duplicates are
4672 presumably rare, so I don't expect this is of any significance.
4674 For deliveries to pipes, files, and autoreplies, the duplicate checking
4675 still happens during the routing process, since they are not going to be
4678 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4679 It corrects a timeout issue with spamd. This is Ian's comment: "The
4680 background is that sometimes spamd either never reads data from a
4681 connection it has accepted, or it never writes response data. The exiscan
4682 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4683 blindly assumes that writes won't block so it may never time out."
4685 PH/39 Allow G after quota size as well as K and M.
4687 PH/40 The value set for $authenticated_id in an authenticator may not contain
4688 binary zeroes or newlines because the value is written to log lines and
4689 to spool files. There was no check on this. Now the value is run through
4690 the string_printing() function so that such characters are converted to
4691 printable escape sequences.
4693 PH/41 $message_linecount is a new variable that contains the total number of
4694 lines in the message. Compare $body_linecount, which is the count for the
4697 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4698 bounce and delay warning messages, except in certain special cases, which
4701 (a) An SMTP error message from a remote host;
4702 (b) A message specified in a :fail: redirection;
4703 (c) A message specified in a "fail" command in a system filter;
4704 (d) A message specified in a FAIL return from the queryprogram router;
4705 (e) A message specified by the cannot_route_message router option.
4707 In these cases only, Exim does include the error details in bounce and
4708 warning messages. There are also a few cases where bland messages such
4709 as "unrouteable address" or "local delivery error" are given.
4711 PH/43 $value is now also set for the "else" part of a ${run expansion.
4713 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4714 being worked on, but at least Exim now implements the latest version to
4717 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4718 process to complete was treated as a delivery failure, a timeout while
4719 writing the message to the pipe was logged, but erroneously treated as a
4720 successful delivery. Such timeouts include transport filter timeouts. For
4721 consistency with the overall process timeout, these timeouts are now
4722 treated as errors, giving rise to delivery failures by default. However,
4723 there is now a new Boolean option for the pipe transport called
4724 timeout_defer, which, if set TRUE, converts the failures into defers for
4725 both kinds of timeout. A transport filter timeout is now identified in
4728 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4729 systems where "make" and "gmake" are different, calling "gmake" at top
4730 level broke things. I've arranged for the value of $(MAKE) to be passed
4731 from the Makefile to this script so that it can call the same version of
4735 A note about Exim versions 4.44 and 4.50
4736 ----------------------------------------
4738 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4739 changes of various kinds. As a consequence, a big documentation update was
4740 needed. This delayed the release for rather longer than seemed good, especially
4741 in the light of a couple of (minor) security issues. Therefore, the changes
4742 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4743 release. So 4.44 and 4.50 are in effect two different branches that both start
4746 I have left the 4.50 change log unchanged; it contains all the changes since
4747 4.43. The change log for 4.44 is below; many of its items are identical to
4748 those for 4.50. This seems to be the most sensible way to preserve the
4749 historical information.
4755 1. Minor wording change to the doc/README.SIEVE file.
4757 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4758 computation of the current number of files was incorrect.
4760 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4761 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4764 4. Give more explanation in the error message when the command for a transport
4765 filter fails to execute.
4767 5. There are several places where Exim runs a non-Exim command in a
4768 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4769 was being done only for the command run by the queryprogram router. It is
4770 now done for all such subprocesses. The other cases are: ${run, transport
4771 filters, and the commands run by the lmtp and pipe transports.
4773 6. Added CONFIGURE_GROUP build-time option.
4775 7. Some older OS have a limit of 256 on the maximum number of file
4776 descriptors. Exim was using setrlimit() to set 1000 as a large value
4777 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4778 systems. I've change it so that if it can't get 1000, it tries for 256.
4780 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4781 was an oversight, and furthermore, ever since the addition of extra
4782 controls (e.g. 4.43/32), the checks on when to allow different forms of
4783 "control" were broken. There should now be diagnostics for all cases when a
4784 control that does not make sense is encountered.
4786 9. Added the /retain_sender option to "control=submission".
4788 10. $recipients is now available in the predata ACL (oversight).
4790 11. Tidy the search cache before the fork to do a delivery from a message
4791 received from the command line. Otherwise the child will trigger a lookup
4792 failure and thereby defer the delivery if it tries to use (for example) a
4793 cached ldap connection that the parent has called unbind on.
4795 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4796 of $address_data from the recipient verification was clobbered by the
4797 sender verification.
4799 13. The value of address_data from a sender verification is now available in
4800 $sender_address_data in subsequent conditions in the ACL statement.
4802 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4804 15. Added a new option "connect=<time>" to callout options, to set a different
4807 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4808 was its contents. (It was OK if the option was not defined at all.)
4810 17. A "Completed" log line is now written for messages that are removed from
4811 the spool by the -Mrm option.
4813 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4814 information about exactly what failed.
4816 19. Added -dd to debug only the daemon process.
4818 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4819 handles timeouts, both on the server side and network timeouts. Renamed the
4820 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4822 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4823 It is now set to "smtps".
4825 22. $host_address is now set to the target address during the checking of
4826 ignore_target_hosts.
4828 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4829 being passed; this would have caused $sender_host_name to have been used if
4830 matching the list had actually called for a host name (not very likely,
4831 since this list is usually IP addresses). A host name is now passed as
4834 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4835 code that specifies a non-restarting handler (typically sigaction() in
4836 modern systems) in an attempt to fix a rare and obscure crash bug.
4838 25. Narrowed the window for a race in the daemon that could cause it to ignore
4839 SIGCHLD signals. This is not a major problem, because they are used only to
4840 wake it up if nothing else does.
4842 26. A malformed maildirsize file could cause Exim to calculate negative values
4843 for the mailbox size or file count. Odd effects could occur as a result.
4844 The maildirsize information is now recalculated if the size or filecount
4847 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4848 support for a long time. Removed HAVE_SYS_VFS_H.
4850 28. Installed the latest version of exipick from John Jetmore.
4852 29. In an address list, if the pattern was not a regular expression, an empty
4853 subject address (from a bounce message) matched only if the pattern was an
4854 empty string. Non-empty patterns were not even tested. This was the wrong
4855 because it is perfectly reasonable to use an empty address as part of a
4856 database query. An empty address is now tested by patterns that are
4857 lookups. However, all the other forms of pattern expect the subject to
4858 contain a local part and a domain, and therefore, for them, an empty
4859 address still always fails if the pattern is not itself empty.
4861 30. Exim went into a mad DNS loop when attempting to do a callout where the
4862 host was specified on an smtp transport, and looking it up yielded more
4863 than one IP address.
4865 31. Re-factored the code for checking spool and log partition space into a
4866 function that finds that data and another that does the check. The former
4867 is then used to implement four new variables: $spool_space, $log_space,
4868 $spool_inodes, and $log_inodes.
4870 32. The RFC2047 encoding function was originally intended for short strings
4871 such as real names; it was not keeping to the 75-character limit for
4872 encoded words that the RFC imposes. It now respects the limit, and
4873 generates multiple encoded words if necessary. To be on the safe side, I
4874 have increased the buffer size for the ${rfc2047: expansion operator from
4877 33. It is now permitted to omit both strings after an "if" condition; if the
4878 condition is true, the result is "true". As before, when the second string
4879 is omitted, a false condition yields an empty string. This makes it less
4880 cumbersome to write custom ACL and router conditions.
4882 34. Failure to deliver a bounce message always caused it to be frozen, even if
4883 there was an errors_to setting on the router. The errors_to setting is now
4886 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4887 canonical form (fully expanded) before being placed in
4888 $sender_host_address.
4890 36. The table in the code that translates DNS record types into text (T_A to
4891 "A" for instance) was missing entries for NS and CNAME. It is just possible
4892 that this could have caused confusion if both these types were looked up
4893 for the same domain, because the text type is used as part of Exim's
4894 per-process caching. But the chance of anyone hitting this buglet seems
4897 37. The dnsdb lookup has been extended in a number of ways.
4899 (1) There is a new type, "zns", which walks up the domain tree until it
4900 finds some nameserver records. It should be used with care.
4902 (2) There is a new type, "mxh", which is like "mx" except that it returns
4903 just the host names, not the priorities.
4905 (3) It is now possible to give a list of domains (or IP addresses) to be
4906 looked up. The behaviour when one of the lookups defers can be
4907 controlled by a keyword.
4909 (4) It is now possible to specify the separator character for use when
4910 multiple records are returned.
4912 38. The dnslists ACL condition has been extended: it is now possible to supply
4913 a list of IP addresses and/or domains to be looked up in a particular DNS
4916 39. Added log_selector=+queue_time_overall.
4918 40. When running the queue in the test harness, wait just a tad after forking a
4919 delivery process, to get repeatability of debugging output.
4921 41. Include certificate and key file names in error message when GnuTLS fails
4922 to set them up, because the GnuTLS error message doesn't include the name
4923 of the failing file when there is a problem reading it.
4925 42. Allow both -bf and -bF in the same test run.
4927 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4929 44. The "Exiscan patch" is now merged into the mainline Exim source.
4931 45. Sometimes the final signoff response after QUIT could fail to get
4932 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4933 < 0 before doing a fflush(). This bug looks as though it goes back to the
4934 introduction of TLS in release 3.20, but "sometimes" must have been rare
4935 because the tests only now provoked it.
4937 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4938 (this can affect the format of dates).
4940 47. exim_tidydb, when checking for the continued existence of a message for
4941 which it has found a message-specific retry record, was not finding
4942 messages that were in split spool directories. Consequently, it was
4943 deleting retry records that should have stayed in existence.
4945 48. Steve fixed some bugs in eximstats.
4947 49. The SPA authentication driver was not abandoning authentication and moving
4948 on to the next authenticator when an expansion was forced to fail,
4949 contradicting the general specification for all authenticators. Instead it
4950 was generating a temporary error. It now behaves as specified.
4952 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4953 (the order specifies the preference for clients). The order is now AES256,
4954 AES128, 3DES, ARCFOUR128.
4956 51. Small patch to Sieve code - explicitly set From: when generating an
4959 52. Exim crashed if a remote delivery caused a very long error message to be
4960 recorded - for instance if somebody sent an entire SpamAssassin report back
4961 as a large number of 550 error lines. This bug was coincidentally fixed by
4962 increasing the size of one of Exim's internal buffers (big_buffer) that
4963 happened as part of the Exiscan merge. However, to be on the safe side, I
4964 have made the code more robust (and fixed the comments that describe what
4967 53. Now that there can be additional text after "Completed" in log lines (if
4968 the queue_time_overall log selector is set), a one-byte patch to exigrep
4969 was needed to allow it to recognize "Completed" as not the last thing in
4972 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4973 patch that reportedly fixes this has been added. I am not expert enough to
4974 create a test for it. This is what the patch creator wrote:
4976 "I found a little strange behaviour of ldap code when working with
4977 Windows 2003 AD Domain, where users was placed in more than one
4978 Organization Units. When I tried to give exim partial DN, the exit code
4979 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4980 But simultaneously result of request was absolutely normal ldap result,
4981 so I produce this patch..."
4983 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4984 REFERENCE, so I have modified the code to exclude the patch when that macro
4987 55. Some experimental protocols are using DNS PTR records for new purposes. The
4988 keys for these records are domain names, not reversed IP addresses. The
4989 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4990 leaves it alone. Component reversal etc. now happens only for IP addresses.
4993 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4995 57. Double the size of the debug message buffer (to 2048) so that more of very
4996 long debug lines gets shown.
4998 58. The exicyclog utility now does better if the number of log files to keep
4999 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5001 59. Two changes related to the smtp_active_hostname option:
5003 (1) $smtp_active_hostname is now available as a variable.
5004 (2) The default for smtp_banner uses $smtp_active_hostname instead
5005 of $primary_hostname.
5007 60. The host_aton() function is supposed to be passed a string that is known
5008 to be a valid IP address. However, in the case of IPv6 addresses, it was
5009 not checking this. This is a hostage to fortune. Exim now panics and dies
5010 if the condition is not met. A case was found where this could be provoked
5011 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5012 components; fortuitously, this particular loophole had already been fixed
5013 by change 4.50/55 above.
5015 If there are any other similar loopholes, the new check in host_aton()
5016 itself should stop them being exploited. The report I received stated that
5017 data on the command line could provoke the exploit when Exim was running as
5018 exim, but did not say which command line option was involved. All I could
5019 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5020 running as the user.
5023 61. There was a buffer overflow vulnerability in the SPA authentication code
5024 (which came originally from the Samba project). I have added a test to the
5025 spa_base64_to_bits() function which I hope fixes it.
5028 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5029 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5031 63. The daemon start-up calls getloadavg() while still root for those OS that
5032 need the first call to be done as root, but it missed one case: when
5033 deliver_queue_load_max is set with deliver_drop_privilege. This is
5034 necessary for the benefit of the queue runner, because there is no re-exec
5035 when deliver_drop_privilege is set.
5037 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5038 This has been fixed.
5040 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5041 was in use, was not putting the data itself into the right store pool;
5042 consequently, it could be overwritten for a subsequent message in the same
5043 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5046 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5048 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5049 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5050 the function were treating the return as a boolean value, which happened to
5051 work because 0=false and not-0=true, but is not correct code.
5053 68. The host_aton() function was not handling scoped IPv6 addresses (those
5054 with, for example, "%eth0" on the end) correctly.
5056 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5057 negated items (that is, ~something) in unsigned ints. Some compilers
5058 apparently mutter when there is no cast.
5060 70. If an address verification called from an ACL failed, and did not produce a
5061 user-specific message (i.e. there was only a "system" message), nothing was
5062 put in $acl_verify_message. In this situation, it now puts the system
5065 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5066 session; change 4.31/43 added the unwanted input to the log line - except
5067 that it did not do this in the start of session case. It now does.
5069 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5070 This is wrong and can cause the other end to generate a synchronization
5071 error if it is another Exim or anything else that does the synchronization
5072 check. A QUIT command is no longer sent after a timeout.
5074 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5075 during host lookups.
5077 74. The defer_ok option of callout verification was not working if it was used
5078 when verifying addresses in header lines, that is, for this case:
5080 verify = header_sender/callout=defer_ok
5082 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5083 those file descriptors could be used for SMTP connections. If anything
5084 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5085 could be sent to the SMTP client, causing chaos. The daemon now opens
5086 stdin, stdout, and stderr to /dev/null when it puts itself into the
5089 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5090 log by default. The user can override this with suitable Perl magic.
5092 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5093 the log message when discard triggers, was not working for the DATA ACL or
5094 for the non-SMTP ACL.
5096 78. Error message wording change in sieve.c.
5098 79. If smtp_accept_max_per_host was set, the number of connections could be
5099 restricted to fewer than expected, because the daemon was trying to set up
5100 a new connection before checking whether the processes handling previous
5101 connections had finished. The check for completed processes is now done
5102 earlier. On busy systems, this bug wouldn't be noticed because something
5103 else would have woken the daemon, and it would have reaped the completed
5106 80. If a message was submitted locally by a user whose login name contained one
5107 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5108 It caused a spool format error. I have fixed the spool reading code. A
5109 related problem was that the "from" clause in the Received: line became
5110 illegal because of the space(s). It is now covered by ${quote_local_part.
5112 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5115 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5117 83. Patch from Sieve maintainer to fix unterminated string problem in
5118 "vacation" handling.
5120 84. Some minor changes to the Linux configuration files to help with other
5121 OS variants using glibc.
5123 85. One more patch for Sieve to update vacation handling to latest spec.
5126 ----------------------------------------------------
5127 See the note above about the 4.44 and 4.50 releases.
5128 ----------------------------------------------------
5134 1. Change 4.43/35 introduced a bug that caused file counts to be
5135 incorrectly computed when quota_filecount was set in an appendfile
5138 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5139 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5142 3. Give more explanation in the error message when the command for a transport
5143 filter fails to execute.
5145 4. There are several places where Exim runs a non-Exim command in a
5146 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5147 was being done only for the command run by the queryprogram router. It is
5148 now done for all such subprocesses. The other cases are: ${run, transport
5149 filters, and the commands run by the lmtp and pipe transports.
5151 5. Some older OS have a limit of 256 on the maximum number of file
5152 descriptors. Exim was using setrlimit() to set 1000 as a large value
5153 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5154 systems. I've change it so that if it can't get 1000, it tries for 256.
5156 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5157 was an oversight, and furthermore, ever since the addition of extra
5158 controls (e.g. 4.43/32), the checks on when to allow different forms of
5159 "control" were broken. There should now be diagnostics for all cases when a
5160 control that does not make sense is encountered.
5162 7. $recipients is now available in the predata ACL (oversight).
5164 8. Tidy the search cache before the fork to do a delivery from a message
5165 received from the command line. Otherwise the child will trigger a lookup
5166 failure and thereby defer the delivery if it tries to use (for example) a
5167 cached ldap connection that the parent has called unbind on.
5169 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5170 of $address_data from the recipient verification was clobbered by the
5171 sender verification.
5173 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5174 was its contents. (It was OK if the option was not defined at all.)
5176 11. A "Completed" log line is now written for messages that are removed from
5177 the spool by the -Mrm option.
5179 12. $host_address is now set to the target address during the checking of
5180 ignore_target_hosts.
5182 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5183 being passed; this would have caused $sender_host_name to have been used if
5184 matching the list had actually called for a host name (not very likely,
5185 since this list is usually IP addresses). A host name is now passed as
5188 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5189 code that specifies a non-restarting handler (typically sigaction() in
5190 modern systems) in an attempt to fix a rare and obscure crash bug.
5192 15. Narrowed the window for a race in the daemon that could cause it to ignore
5193 SIGCHLD signals. This is not a major problem, because they are used only to
5194 wake it up if nothing else does.
5196 16. A malformed maildirsize file could cause Exim to calculate negative values
5197 for the mailbox size or file count. Odd effects could occur as a result.
5198 The maildirsize information is now recalculated if the size or filecount
5201 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5202 support for a long time. Removed HAVE_SYS_VFS_H.
5204 18. Updated exipick to current release from John Jetmore.
5206 19. Allow an empty sender to be matched against a lookup in an address list.
5207 Previously the only cases considered were a regular expression, or an
5210 20. Exim went into a mad DNS lookup loop when doing a callout where the
5211 host was specified on the transport, if the DNS lookup yielded more than
5214 21. The RFC2047 encoding function was originally intended for short strings
5215 such as real names; it was not keeping to the 75-character limit for
5216 encoded words that the RFC imposes. It now respects the limit, and
5217 generates multiple encoded words if necessary. To be on the safe side, I
5218 have increased the buffer size for the ${rfc2047: expansion operator from
5221 22. Failure to deliver a bounce message always caused it to be frozen, even if
5222 there was an errors_to setting on the router. The errors_to setting is now
5225 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5226 canonical form (fully expanded) before being placed in
5227 $sender_host_address.
5229 24. Updated eximstats to version 1.33
5231 25. Include certificate and key file names in error message when GnuTLS fails
5232 to set them up, because the GnuTLS error message doesn't include the name
5233 of the failing file when there is a problem reading it.
5235 26. Expand error message when OpenSSL has problems setting up cert/key files.
5238 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5239 (this can affect the format of dates).
5241 28. exim_tidydb, when checking for the continued existence of a message for
5242 which it has found a message-specific retry record, was not finding
5243 messages that were in split spool directories. Consequently, it was
5244 deleting retry records that should have stayed in existence.
5246 29. eximstats updated to version 1.35
5247 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5248 1.35 - bugfix such that pie charts by volume are generated correctly
5250 30. The SPA authentication driver was not abandoning authentication and moving
5251 on to the next authenticator when an expansion was forced to fail,
5252 contradicting the general specification for all authenticators. Instead it
5253 was generating a temporary error. It now behaves as specified.
5255 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5256 (the order specifies the preference for clients). The order is now AES256,
5257 AES128, 3DES, ARCFOUR128.
5259 31. Small patch to Sieve code - explicitly set From: when generating an
5262 32. Exim crashed if a remote delivery caused a very long error message to be
5263 recorded - for instance if somebody sent an entire SpamAssassin report back
5264 as a large number of 550 error lines. This bug was coincidentally fixed by
5265 increasing the size of one of Exim's internal buffers (big_buffer) that
5266 happened as part of the Exiscan merge. However, to be on the safe side, I
5267 have made the code more robust (and fixed the comments that describe what
5270 33. Some experimental protocols are using DNS PTR records for new purposes. The
5271 keys for these records are domain names, not reversed IP addresses. The
5272 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5273 leaves it alone. Component reversal etc. now happens only for IP addresses.
5276 34. The host_aton() function is supposed to be passed a string that is known
5277 to be a valid IP address. However, in the case of IPv6 addresses, it was
5278 not checking this. This is a hostage to fortune. Exim now panics and dies
5279 if the condition is not met. A case was found where this could be provoked
5280 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5281 components; fortuitously, this particular loophole had already been fixed
5282 by change 4.50/55 or 4.44/33 above.
5284 If there are any other similar loopholes, the new check in host_aton()
5285 itself should stop them being exploited. The report I received stated that
5286 data on the command line could provoke the exploit when Exim was running as
5287 exim, but did not say which command line option was involved. All I could
5288 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5289 running as the user.
5292 35. There was a buffer overflow vulnerability in the SPA authentication code
5293 (which came originally from the Samba project). I have added a test to the
5294 spa_base64_to_bits() function which I hope fixes it.
5297 36. The daemon start-up calls getloadavg() while still root for those OS that
5298 need the first call to be done as root, but it missed one case: when
5299 deliver_queue_load_max is set with deliver_drop_privilege. This is
5300 necessary for the benefit of the queue runner, because there is no re-exec
5301 when deliver_drop_privilege is set.
5303 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5304 was in use, was not putting the data itself into the right store pool;
5305 consequently, it could be overwritten for a subsequent message in the same
5306 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5309 38. Sometimes the final signoff response after QUIT could fail to get
5310 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5311 < 0 before doing a fflush(). This bug looks as though it goes back to the
5312 introduction of TLS in release 3.20, but "sometimes" must have been rare
5313 because the tests only now provoked it.
5319 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5320 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5321 Now it returns an integer. A number of calls were still expecting a T/F
5322 return. Fortuitously, in all cases, the tests worked in OK situations,
5323 which is the norm. However, things would have gone wrong on any write
5324 failures on the smtp file descriptor. This function is used when sending
5325 messages over SMTP and also when doing verify callouts.
5327 2. When Exim is called to do synchronous delivery of a locally submitted
5328 message (the -odf or -odi options), it no longer closes stderr before doing
5331 3. Implemented the mua_wrapper option.
5333 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5335 5. Implemented the functions header_remove(), header_testname(),
5336 header_add_at_position(), and receive_remove_recipient(), and exported them
5339 6. If an ACL "warn" statement specified the addition of headers, Exim already
5340 inserted X-ACL-Warn: at the start if there was no header name. However, it
5341 was not making this test for the second and subsequent header lines if
5342 there were newlines in the string. This meant that an invalid header could
5343 be inserted if Exim was badly configured.
5345 7. Allow an ACL "warn" statement to add header lines at the start or after all
5346 the Received: headers, as well as at the end.
5348 8. Added the rcpt_4xx retry error code.
5350 9. Added postmaster_mailfrom=xxx to callout verification option.
5352 10. Added mailfrom=xxxx to the callout verification option, for verify=
5355 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5356 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5358 12. Inserted some casts to stop certain compilers warning when using pointer
5359 differences as field lengths or precisions in printf-type calls (mostly
5360 affecting debugging statements).
5362 13. Added optional readline() support for -be (dynamically loaded).
5364 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5365 same clock tick as a message's arrival, so that its received time was the
5366 same as the "first fail" time on the retry record, and that message
5367 remained on the queue past the ultimate address timeout, every queue runner
5368 would try a delivery (because it was past the ultimate address timeout) but
5369 after another failure, the ultimate address timeout, which should have then
5370 bounced the address, did not kick in. This was a "< instead of <=" error;
5371 in most cases the first failure would have been in the next clock tick
5372 after the received time, and all would be well.
5374 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5375 being recognized when the domain list was tested by the match_domain
5376 condition in an expansion string.
5378 16. Added the ${str2b64: operator.
5380 17. Exim was always calling setrlimit() to set a large limit for the number of
5381 processes, without checking whether the existing limit was already
5382 adequate. (It did check for the limit on file descriptors.) Furthermore,
5383 errors from getrlimit() and setrlimit() were being ignored. Now they are
5384 logged to the main and panic logs, but Exim does carry on, to try to do its
5385 job under whatever limits there are.
5387 18. Imported PCRE 5.0.
5389 19. Trivial typo in log message " temporarily refused connection" (the leading
5392 20. If the log selector return_path_on_delivery was set and an address was
5393 redirected to /dev/null, the delivery process crashed because it assumed
5394 that a return path would always be set for a "successful" delivery. In this
5395 case, the whole delivery is bypassed as an optimization, and therefore no
5398 21. Internal re-arrangement: the function for sending a challenge and reading
5399 a response while authentication was assuming a zero-terminated challenge
5400 string. It's now changed to take a pointer and a length, to allow for
5401 binary data in such strings.
5403 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5405 23. Exim was not respecting finduser_retries when seeking the login of the
5406 uid under which it was called; it was always trying 10 times. (The default
5407 setting of finduser_retries is zero.) Also, it was sleeping after the final
5408 failure, which is pointless.
5410 24. Implemented tls_on_connect_ports.
5412 25. Implemented acl_smtp_predata.
5414 26. If the domain in control=submission is set empty, Exim assumes that the
5415 authenticated id is a complete email address when it generates From: or
5416 Sender: header lines.
5418 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5419 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5420 chown and chgrp in /bin and hostname in /usr/bin.
5422 28. Exim was keeping the "process log" file open after each use, just as it
5423 does for the main log. This opens the possibility of it remaining open for
5424 long periods when the USR1 signal hits a daemon. Occasional processlog
5425 errors were reported, that could have been caused by this. Anyway, it seems
5426 much more sensible not to leave this file open at all, so that is what now
5429 29. The long-running daemon process does not normally write to the log once it
5430 has entered its main loop, and it closes the log before doing so. This is
5431 so that log files can straightforwardly be renamed and moved. However,
5432 there are a couple of unusual error situations where the daemon does write
5433 log entries, and I had neglected to close the log afterwards.
5435 30. The text of an SMTP error response that was received during a remote
5436 delivery was being truncated at 512 bytes. This is too short for some of
5437 the long messages that one sometimes sees. I've increased the limit to
5440 31. It is now possible to make retry rules that apply only when a message has a
5441 specific sender, in particular, an empty sender.
5443 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5444 it possible to be selective about when SMTP synchronization is enforced.
5446 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5448 32. Implemented hosts_connection_nolog.
5450 33. Added an ACL for QUIT.
5452 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5455 35. Added mailbox_size and mailbox_filecount to appendfile.
5457 36. Added control = no_multiline_responses to ACLs.
5459 37. There was a bug in the logic of the code that waits for the clock to tick
5460 in the case where the clock went backwards by a substantial amount such
5461 that the microsecond fraction of "now" was more than the microsecond
5462 fraction of "then" (but the whole seconds number was less).
5464 38. Added support for the libradius Radius client library this is found on
5465 FreeBSD (previously only the radiusclient library was supported).
5471 1. When certain lookups returned multiple values in the form name=value, the
5472 quoting of the values was not always being done properly. Specifically:
5473 (a) If the value started with a double quote, but contained no whitespace,
5475 (b) If the value contained whitespace other than a space character (i.e.
5476 tabs or newlines or carriage returns) it was not quoted.
5477 This fix has been applied to the mysql and pgsql lookups by writing a
5478 separate quoting function and calling it from the lookup code. The fix
5479 should probably also be applied to nisplus, ibase and oracle lookups, but
5480 since I cannot test any of those, I have not disturbed their existing code.
5482 2. A hit in the callout cache for a specific address caused a log line with no
5483 reason for rejecting RCPT. Now it says "Previous (cached) callout
5484 verification failure".
5486 3. There was an off-by-one bug in the queryprogram router. An over-long
5487 return line was truncated at 256 instead of 255 characters, thereby
5488 overflowing its buffer with the terminating zero. As well as fixing this, I
5489 have increased the buffer size to 1024 (and made a note to document this).
5491 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5492 when Exim is waiting for an SMTP response from a remote server, Exim
5493 restarts its select() call on the socket, thereby resetting its timeout.
5494 This is not a problem when such interrupts are rare. Somebody set up a cron
5495 job to run exiwhat every 2 minutes, which is less than the normal select()
5496 timeout (5 or 10 minutes). This meant that the select() timeout never
5497 kicked in because it was always reset. I have fixed this by comparing the
5498 time when an interrupt arrives with the time at the start of the first call
5499 to select(). If more time than the timeout has elapsed, the interrupt is
5500 treated as a timeout.
5502 5. Some internal re-factoring in preparation for the addition of Sieve
5503 extensions (by MH). In particular, the "personal" test is moved to a
5504 separate function, and given an option for scanning Cc: and Bcc: (which is
5505 not set for Exim filters).
5507 6. When Exim created an email address using the login of the caller as the
5508 local part (e.g. when creating a From: or Sender: header line), it was not
5509 quoting the local part when it contained special characters such as @.
5511 7. Installed new OpenBSD configuration files.
5513 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5514 try to make them clearer.
5516 9. Callout options, other than the timeout value, were being ignored when
5517 verifying sender addresses in header lines. For example, when using
5519 verify = header_sender/callout=no_cache
5521 the cache was (incorrectly) being used.
5523 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5524 only the Cygwin environment.
5526 11. When return_path_on_delivery was set as a log selector, if different remote
5527 addresses in the same message used different return paths and parallel
5528 remote delivery occurred, the wrong values would sometimes be logged.
5529 (Whenever a remote delivery process finished, the return path value from
5530 the most recently started remote delivery process was logged.)
5532 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5533 lines when AUTH and/or TLS are in use. This is the "received protocol"
5534 field. Exim used to use "asmtp" for authenticated SMTP, without any
5535 indication (in the protocol name) for TLS use. Now it follows the RFC and
5536 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5537 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5538 names appear in log lines as well as in Received: header lines.
5540 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5541 extensions, and comparison tests, and to fix some bugs.
5543 14. Changes to the "personal" filter test:
5545 (1) The test was buggy in that it was just doing the equivalent of
5546 "contains" tests on header lines. For example, if a user's address was
5547 anne@some.where, the "personal" test would incorrectly be true for
5549 To: susanne@some.where
5551 This test is now done by extracting each address from the header in turn,
5552 and checking the entire address. Other tests that are part of "personal"
5553 are now done using regular expressions (for example, to check local parts
5554 of addresses in From: header lines).
5556 (2) The list of non-personal local parts in From: addresses has been
5557 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5558 taken from the Sieve specification recommendations.
5560 (3) If the message contains any header line starting with "List-" it is
5561 treated as non-personal.
5563 (4) The test for "circular" in the Subject: header line has been removed
5564 because it now seems ill-conceived.
5566 15. Minor typos in src/EDITME comments corrected.
5568 16. Installed latest exipick from John Jetmore.
5570 17. If headers_add on a router specified a text string that was too long for
5571 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5572 of string_sprintf() is now avoided.
5574 18. $message_body_size was not set (it was always zero) when running the DATA
5575 ACL and the local_scan() function.
5577 19. For the "mail" command in an Exim filter, no default was being set for
5578 the once_repeat time, causing a random time value to be used if "once" was
5579 specified. (If the value happened to be <= 0, no repeat happened.) The
5580 default is now 0s, meaning "never repeat". The "vacation" command was OK
5581 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5582 (I found it when inspecting the code).
5584 20. There is now an overall timeout for performing a callout verification. It
5585 defaults to 4 times the callout timeout, which applies to individual SMTP
5586 commands during the callout. The overall timeout applies when there is more
5587 than one host that can be tried. The timeout is checked before trying the
5588 next host. This prevents very long delays if there are a large number of
5589 hosts and all are timing out (e.g. when the network connections are timing
5590 out). The value of the overall timeout can be changed by specifying an
5591 additional sub-option for "callout", called "maxwait". For example:
5593 verify = sender/callout=5s,maxwait=20s
5595 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5596 to the end before writing, but this should make it even safer).
5598 22. Exim was forgetting that it had advertised PIPELINING for the second and
5599 subsequent messages on an SMTP connection. It was also not resetting its
5600 memory on STARTTLS and an internal HELO.
5602 23. When Exim logs an SMTP synchronization error within a session, it now
5603 records whether PIPELINING has been advertised or not.
5605 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5606 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5607 rather than long int.
5609 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5611 26. Added the never_mail option to autoreply.
5617 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5618 crash if the getsockname() call failed; this can happen if a connection is
5619 closed very soon after it is established. The problem was simply in the
5620 order in which certain operations were done, causing Exim to try to write
5621 to the SMTP stream before it had set up the file descriptor. The bug has
5622 been fixed by making things happen in the correct order.
5628 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5629 before the connection was closed, thus losing the rejection response.
5631 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5632 some early Solaris releases, but causes trouble in current releases where
5633 socklen_t is defined.
5635 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5638 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5641 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5642 that releasing the top of it at the end releases what was used for sub-
5643 expansions (unless the block got too big). However, discard this block if
5644 the first thing is a variable or header, so that we can use its block when
5645 it is dynamic (useful for very large $message_headers, for example).
5647 6. Lookups now cache *every* query, not just the most recent. A new, separate
5648 store pool is used for this. It can be recovered when all lookup caches are
5649 flushed. Lookups now release memory at the end of their result strings.
5650 This has involved some general refactoring of the lookup sources.
5652 7. Some code has been added to the store_xxx() functions to reduce the amount
5653 of flapping under certain conditions.
5655 8. log_incoming_interface used to affect only the <= reception log lines. Now
5656 it causes the local interface and port to be added to several more SMTP log
5657 lines, for example "SMTP connection from", and rejection lines.
5659 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5661 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5663 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5664 could be overwritten at the end of the current message (or the start of a
5665 new message if it was set in a HELO ACL). The value is now preserved for
5666 the duration of the SMTP connection.
5668 12. If a transport had a headers_rewrite setting, and a matching header line
5669 contained an unqualified address, that address was qualified, even if it
5670 did not match any rewriting rules. The underlying bug was that the values
5671 of the flags that permit the existence of unqualified sender and recipient
5672 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5673 non-local messages, and by -bnq for local messages) were not being
5674 preserved with the message after it was received.
5676 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5677 "next input=" as part of the text comprising the host identity instead of
5678 the correct text. The code was using the same buffer for two different
5679 strings. However, depending on which order the printing function evaluated
5680 its arguments, the bug did not always show up. Under Linux, for example, my
5681 test suite worked just fine.
5683 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5684 doesn't work with some older versions of Perl. It has been changed to "my",
5685 which in any case is probably the better facility to use.
5687 15. A really picky compiler found some instances of statements for creating
5688 error messages that either had too many or two few arguments for the format
5691 16. The size of the buffer for calls to the DNS resolver has been increased
5692 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5693 for addresses that have a lot of PTR records. This alleviates a problem; it
5694 does not fully solve it.
5696 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5697 buffer now truncates the list and logs the incident, which is the same
5698 action as happens when Exim is looking up a host name and its aliases.
5699 Previously in this situation something unpredictable would happen;
5700 sometimes it was "internal error: store_reset failed".
5702 18. If a server dropped the connection unexpectedly when an Exim client was
5703 using GnuTLS and trying to read a response, the client delivery process
5704 crashed while trying to generate an error log message.
5706 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5707 string, for example:
5709 warn message = H1: something\nH2: something
5711 the text was added as a single header line from Exim's point of view
5712 though it ended up OK in the delivered message. However, searching for the
5713 second and subsequent header lines using $h_h2: did not work. This has been
5714 fixed. Similarly, if a system filter added multiple headers in this way,
5715 the routers could not see them.
5717 20. Expanded the error message when iplsearch is called with an invalid key to
5718 suggest using net-iplsearch in a host list.
5720 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5721 ACLs are no longer actually imposed (and a message to that effect is
5724 22. If a "gecos" field in a passwd entry contained escaped characters, in
5725 particular, if it contained a \" sequence, Exim got it wrong when building
5726 a From: or a Sender: header from that name. A second bug also caused
5727 incorrect handling when an unquoted " was present following a character
5728 that needed quoting.
5730 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5731 was not being matched caselessly.
5733 24. Arranged for all hyphens in the exim.8 source to be escaped with
5736 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5737 properly. Recipient callout cache records were still being keyed to include
5738 the sender, even when use_sender was set false. This led to far more
5739 callouts that were necessary. The sender is no longer included in the key
5740 when use_sender is false.
5742 26. Added "control = submission" modifier to ACLs.
5744 27. Added the ${base62d: operator to decode base 62 numbers.
5746 28. dnsdb lookups can now access SRV records.
5748 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5749 the configuration file.
5751 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5752 (-v) mode. This makes the output for a verbose queue run more intelligible.
5754 31. Added a use_postmaster feature to recipient callouts.
5756 32. Added the $body_zerocount variable, containing the number of binary zero
5757 bytes in the message body.
5759 33. The time of last modification of the "new" subdirectory is now used as the
5760 "mailbox time last read" when there is a quota error for a maildir
5763 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5765 35. Added +ignore_unknown as a special item in host lists.
5767 36. Code for decoding IPv6 addresses in host lists is now included, even if
5768 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5769 address was recognized as an IP address, but was then not correctly decoded
5770 into binary, causing unexpected and incorrect effects when compared with
5777 1. Very minor rewording of debugging text in manualroute to say "list of
5778 hosts" instead of "hostlist".
5780 2. If verify=header_syntax was set, and a header line with an unqualified
5781 address (no domain) and a large number of spaces between the end of the
5782 name and the colon was received, the reception process suffered a buffer
5783 overflow, and (when I tested it) crashed. This was caused by some obsolete
5784 code that should have been removed. The fix is to remove it!
5786 3. When running in the test harness, delay a bit after writing a bounce
5787 message to get a bit more predictability in the log output.
5789 4. Added a call to search_tidyup() just before forking a reception process. In
5790 theory, someone could use a lookup in the expansion of smtp_accept_max_
5791 per_host which, without the tidyup, could leave open a database connection.
5793 5. Added the variables $recipient_data and $sender_data which get set from a
5794 lookup success in an ACL "recipients" or "senders" condition, or a router
5795 "senders" option, similar to $domain_data and $local_part_data.
5797 6. Moved the writing of debug_print from before to after the "senders" test
5800 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5801 problems for message scanning, either using a data ACL, or using
5802 local_scan() because the Received: header was not generated till after they
5803 were called (in order to set the time as the time of reception completion).
5804 I have revised the way this works. The header is now generated after the
5805 body is received, but before the ACL or local_scan() are called. After they
5806 are run, the timestamp in the header is updated.
5812 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5813 before starting a queue runner without re-exec. This happened only when
5814 deliver_drop_privilege was set or when the Exim user was set to root. The
5815 effect of the bug was that timeouts during subsequent deliveries caused
5816 crashes instead of being properly handled. The handler is now left at its
5817 default (and expected) setting.
5819 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5820 message, again when deliver_drop_privilege is set or Exim is run as root.
5821 The bug described in (1) was not present in this case, but the tidying up
5822 of the other signals was missing. I have made the two cases consistent.
5824 3. The ignore_target_hosts setting on a manualroute router was being ignored
5825 for hosts that were looked up using the /MX notation.
5827 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5830 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5831 operated on the sender address. After changing the $sender_address to <>
5832 for the sender address verify, Exim was re-instated it as the original
5833 (before rewriting) address, but remembering that it had rewritten it, so it
5834 wasn't rewriting it again. This bug also had the effect of breaking the
5835 sender address verification caching when the sender address was rewritten.
5837 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5838 This has been changed so that if the ip literal address matches
5839 ignore_target_hosts, the router declines.
5841 7. Added expansion conditions match_domain, match_address, and match_local_
5842 part (NOT match_host).
5844 8. The placeholder for the Received: header didn't have a length field set.
5846 9. Added code to Exim itself and to exim_lock to test for a specific race
5847 condition that could lead to file corruption when using MBX delivery. The
5848 issue is with the lockfile that is created in /tmp. If this file is removed
5849 after a process has opened it but before that process has acquired a lock,
5850 there is the potential for a second process to recreate the file and also
5851 acquire a lock. This could lead to two Exim processes writing to the file
5852 at the same time. The added code performs the same test as UW imapd; it
5853 checks after acquiring the lock that its file descriptor still refers to
5854 the same named file.
5856 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5857 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5860 11. Added the smtp_active_hostname option. If used, this will typically be made
5861 to depend on the incoming interface address. Because $interface_address is
5862 not set up until the daemon has forked a reception process, error responses
5863 that can happen earlier (such as "too many connections") no longer contain
5866 12. If an expansion in a condition on a "warn" statement fails because a lookup
5867 defers, the "warn" statement is abandoned, and the next ACL statement is
5868 processed. Previously this caused the whole ACL to be aborted.
5870 13. Added the iplsearch lookup type.
5872 14. Added ident_timeout as a log selector.
5874 15. Added tls_certificate_verified as a log selector.
5876 16. Added a global option tls_require_ciphers (compare the smtp transport
5877 option of the same name). This controls incoming TLS connections.
5879 17. I finally figured out how to make tls_require_ciphers do a similar thing
5880 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5881 before starting the TLS session.
5883 18. Tabs are now shown as \t in -bP output.
5885 19. If the log selector return_path_on_delivery was set, Exim crashed when
5886 bouncing a message because it had too many Received: header lines.
5888 20. If two routers both had headers_remove settings, and the first one included
5889 a superfluous trailing colon, the final name in the first list and the
5890 first name in the second list were incorrectly joined into one item (with a
5891 colon in the middle).
5897 1. Added -C and -D options to the exinext utility, mainly to make it easier
5898 to include in the automated testing, but these could be helpful when
5899 multiple configurations are in use.
5901 2. The exinext utility was not formatting the output nicely when there was
5902 an alternate port involved in the retry record key, nor when there was a
5903 message id as well (for retries that were specific to a specific message
5904 and a specific host). It was also confused by IPv6 addresses, because of
5905 the additional colons they contain. I have fixed the IPv4 problem, and
5906 patched it up to do a reasonable job for IPv6.
5908 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5909 delivery, the log line now contains "pipelined" if PIPELINING was used.
5911 4. An SMTP transport process used to panic and die if the bind() call to set
5912 an explicit outgoing interface failed. This has been changed; it is now
5913 treated in the same way as a connect() failure.
5915 5. A reference to $sender_host_name in the part of a conditional expansion
5916 that was being skipped was still causing a DNS lookup. This no longer
5919 6. The def: expansion condition was not recognizing references to header lines
5920 that used bh_ and bheader_.
5922 7. Added the _cache feature to named lists.
5924 8. The code for checking quota_filecount in the appendfile transport was
5925 allowing one more file than it should have been.
5927 9. For compatibility with Sendmail, the command line option
5935 and sets the incoming protocol and host name (for trusted callers). The
5936 host name and its colon can be omitted when only the protocol is to be set.
5937 Note the Exim already has two private options, -pd and -ps, that refer to
5938 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5939 "s", but I don't think that's a major issue.
5941 10. A number of refactoring changes to the code, none of which should affect
5944 (a) The number of logging options was getting close to filling up the
5945 32-bit word that was used as a bit map. I have split them into two classes:
5946 those that are passed in the argument to log_write(), and those that are
5947 only ever tested independently outside of that function. These are now in
5948 separate 32-bit words, so there is plenty of room for expansion again.
5949 There is no change in the user interface or the logging behaviour.
5951 (b) When building, for example, log lines, the code previously used a
5952 macro that called string_cat() twice, in order to add two strings. This is
5953 not really sufficiently general. Furthermore, there was one instance where
5954 it was actually wrong because one of the argument was used twice, and in
5955 one call a function was used. (As it happened, calling the function twice
5956 did not affect the overall behaviour.) The macro has been replaced by a
5957 function that can join an arbitrary number of extra strings onto a growing
5960 (c) The code for expansion conditions now uses a table and a binary chop
5961 instead of a serial search (which was left over from when there were very
5962 few conditions). Also, it now recognizes conditions like "pam" even when
5963 the relevant support is not compiled in: a suitably worded error message is
5964 given if an attempt is made to use such a condition.
5966 11. Added ${time_interval:xxxxx}.
5968 12. A bug was causing one of the ddress fields not to be passed back correctly
5969 from remote delivery subprocesses. The field in question was not being
5970 subsequently used, so this caused to problems in practice.
5972 13. Added new log selectors queue_time and deliver_time.
5974 14. Might have fixed a bug in maildirsizefile handling that threw up
5975 "unexpected character" debug warnings, and recalculated the data
5976 unnecessarily. In any case, I expanded the warning message to give more
5979 15. Added the message "Restricted characters in address" to the statements in
5980 the default ACL that block characters like @ and % in local parts.
5982 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5983 Three changes have been made:
5985 (a) There was a serious bug; a negative response to MAIL caused the whole
5986 recipient domain to be cached as invalid, thereby blocking all messages
5987 to all local parts at the same domain, from all senders. This bug has
5988 been fixed. The domain is no longer cached after a negative response to
5989 MAIL if the sender used is not empty.
5991 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5994 (c) A new callout option, "use_sender" has been added for people who want
5995 the modified behaviour.
6001 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6004 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6005 indeed breaks things for older releases.
6007 3. Added additional logging to the case where there is a problem reading data
6008 from a filter that is running in a subprocess using a pipe, in order to
6009 try to track down a specific problem.
6011 4. Testing facility fudge: when running in the test harness and attempting
6012 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6013 getting "No route to host". Convert this to a timeout.
6015 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6018 6. Some OS don't have socklen_t but use size_t instead. This affects the
6019 fifth argument of getsockopt() amongst other things. This is now
6020 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6021 can be set for individual OS. I have set it for SunOS5, OSF1, and
6022 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6023 some earlier ones do not.
6025 7. Change 4.30/15 was not doing the test caselessly.
6027 8. The standard form for an IPv6 address literal was being rejected by address
6028 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6029 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6030 this, as well as the form without the "IPv6" on the front (but only when
6031 address literals are enabled, of course).
6033 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6035 10. Exim crashed if a message with an empty sender address specified by -f
6036 encountered a router with an errors_to setting. This could be provoked only
6037 by a command such as
6041 where an empty string was supplied; "<>" did not hit this bug.
6043 11. Installed PCRE release 4.5.
6045 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6046 remained set. It is now erased.
6048 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6049 times from message ids (which are base 36 rather than the normal 62).
6051 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6052 were being counted as actual protocol errors, and logged if the log
6053 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6054 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6055 and DATA following a set of rejected RCPTs do not count as protocol errors.
6056 In other words, Exim assumes they were pipelined, though this may not
6057 actually be the case. Of course, in all cases the client gets an
6058 appropriate error code.
6060 15. If a lookup fails in an ACL condition, a message about the failure may
6061 be available; it is used if testing the ACL cannot continue, because most
6062 such messages specify what the cause of the deferral is. However, some
6063 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6064 that caused an old message to be retained and used if a later statement
6065 caused a defer, replacing the real cause of the deferral.
6067 16. If an IP address had so many PTR records that the DNS lookup buffer
6068 was not large enough to hold them, Exim could crash while trying to process
6069 the truncated data. It now detects and logs this case.
6071 17. Further to 4.21/58, another change has been made: if (and only if) the
6072 first line of a message (the first header line) ends with CRLF, a bare LF
6073 in a subsequent header line has a space inserted after it, so as not to
6074 terminate the header.
6076 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6077 unnecessarily, used atoi() instead of strtol(), and didn't check the
6078 termination when getting file sizes from file names by regex.
6080 19. Completely re-implemented the support for maildirsize files, in the light
6081 of a number of problems with the previous contributed implementation
6082 (4.30/29). In particular:
6084 . If the quota is zero, the maildirsize file is maintained, but no quota is
6087 . If the maildir directory does not exist, it is created before any attempt
6088 to write a maildirsize file.
6090 . The quota value in the file is just a cache; if the quota is changed in
6091 the transport, the new value overrides.
6093 . A regular expression is available for excluding directories from the
6096 20. The autoreply transport checks the characters in options that define the
6097 message's headers; it allows continued headers, but it was checking with
6098 isspace() after an embedded newline instead of explicitly looking for a
6101 21. If all the "regular" hosts to which an address was routed had passed their
6102 expiry times, and had not reached their retry times, the address was
6103 bounced, even if fallback hosts were defined. Now Exim should go on to try
6106 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6107 equivalent code in the SMTP transport. Some hosts send humungous responses
6108 to HELO/EHLO, more than 1024 it seems.
6110 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6111 gives compiler warnings in some environments. I've now done it "properly",
6114 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6115 (because of problems with the built-in one) was declared to return uschar *
6116 instead of char *, causing compiler failure.
6118 25. Fixed a file descriptor leak when processing alias/forward files.
6120 26. Fixed a minor format string issue in dbfn.c.
6122 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6124 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6125 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6128 29. When a sender address is verified, it is cached, to save repeating the test
6129 when there is more than one recipient in a message. However, when the
6130 verification involves a callout, it is possible for different callout
6131 options to be set for different recipients. It is too complicated to keep
6132 track of this in the cache, so now Exim always runs a verification when a
6133 callout is required, relying on the callout cache for the optimization.
6134 The overhead is duplication of the address routing, but this should not be
6137 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6138 to be verified with callout=postmaster, and the main callout worked but the
6139 postmaster check failed, the verification correctly failed. However, if a
6140 subsequent RCPT command asked for sender verification *without* the
6141 postmaster check, incorrect caching caused this verification also to fail,
6144 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6145 it was not caching the DNS options (qualify_single, search_parents) that
6146 were used when the lookup failed. A subsequent lookup with different
6147 options therefore always gave the same answer, though there were cases
6148 where it should not have. (Example: a "domains = !$mx_any" option on a
6149 dnslookup router: the "domains" option is always processed without any
6150 widening, but the router might have qualify_single set.) Now Exim uses the
6151 cached value only when the same options are set.
6153 32. Added John Jetmore's "exipick" utility to the distribution.
6155 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6156 than a timeout (e.g. a certificate is required, and is not provided), an
6157 Exim server now closes the connection immediately. Previously it waited for
6158 the client to close - but if the client is SSL, it seems that they each
6159 wait for each other, leading to a delay before one of them times out.
6161 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6162 maintained 0.8.x compatibility because I don't think many are using it, and
6163 it is clearly obsolete.
6165 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6168 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6169 client certificate was expired. A simple patch fixes this, though I don't
6170 understand the full logic of why the verify callback is called multiple
6173 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6174 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6175 which causes problems with some clients (such as the Certicom SSL Plus
6176 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6177 disables the coutermeasure allowing Eudora to connect."
6179 38. Exim was not checking that a write() to a log file succeeded. This could
6180 lead to Bad Things if a log got too big, in particular if it hit a file
6181 size limit. Exim now panics and dies if it cannot write to a log file, just
6182 as it does if it cannot open a log file.
6184 39. Modified OS/Makefile-Linux so that it now contains
6186 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6188 The two -D definitions ensure that Exim is compiled with large file
6189 support, which makes it possible to handle log files that are bigger than
6192 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6193 instance) a domain was checked against a named list that involved a lookup,
6194 causing $domain_data to be set, then another domain was checked against the
6195 same list, then the first domain was re-checked, the value of $domain_data
6196 after the final check could be wrong. In particular, if the second check
6197 failed, it could be set empty. This bug probably also applied to
6200 41. The strip_trailing_dot option was not being applied to the address given
6201 with the -f command-line option.
6203 42. The code for reading a message's header from the spool was incrementing
6204 $received_count, but never initializing it. This meant that the value was
6205 incorrect (doubled) while delivering a message in the same process in which
6206 it was received. In the most common configuration of Exim, this never
6207 happens - a fresh exec is done - but it can happen when
6208 deliver_drop_privilege is set.
6210 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6211 it now includes up to 150 characters of the unexpected data in the log
6214 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6215 and building data strings. The size of both of these buffers was 10 000
6216 bytes - far larger than anybody would *ever* want, thought I. Needless to
6217 say, somebody hit the limit. I have increased the maximum line length to
6218 20 000 and the maximum data length of concatenated lines to 100 000. I have
6219 also fixed two bugs, because there was no checking on these buffers. Tsk,
6220 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6221 buffer is too small.
6223 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6224 lsearch lookups. Now it does.
6226 46. When parsing a route_list item in a manualroute router, a fixed-length
6227 buffer was used for the list of hosts. I made this 1024 bytes long,
6228 thinking that nobody would ever have a list of hosts that long. Wrong.
6229 Somebody had a whole pile of complicated expansion conditions, and the
6230 string was silently truncated, leading to an expansion error. It turns out
6231 that it is easier to change to an unlimited length (owing to other changes
6232 that have happened since this code was originally written) than to build
6233 structure for giving a limitation error. The length of the item that
6234 expands into the list of hosts is now unlimited.
6236 47. The lsearch lookup could not handle data where the length of text line was
6237 more than 4095 characters. Such lines were truncated, leading to shortened
6238 data being returned. It should now handle lines of any length.
6240 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6241 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6244 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6245 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6246 no longer happens in comment lines. A list of replacements is now placed
6247 at the head of all of the source files, except those whose only change is
6248 to replace PERL_COMMAND in the very first #! line.
6250 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6251 messages on the queue, with a bottom-up merge sort, using code contributed
6252 by Michael Haardt. This should make operations like -bp somewhat faster on
6253 large queues. It won't affect queue runners, except when queue_run_in_order
6256 51. Installed eximstats 1.31 in the distribution.
6258 52. Added support for SRV lookups to the dnslookup router.
6260 53. If an ACL referred to $message_body or $message_body_end, the value was not
6261 reset for any messages that followed in the same SMTP session.
6263 54. The store-handling optimization for building very long strings was not
6264 differentiating between the different store pools. I don't think this
6265 actually made any difference in practice, but I've tidied it.
6267 55. While running the routers to verify a sender address, $sender_address
6268 was still set to the sender address. This is wrong, because when routing to
6269 send a bounce to the sender, it would be empty. Therefore, I have changed
6270 it so that, while verifying a sender address, $sender_address is set to <>.
6271 (There is no change to what happens when verifying a recipient address.)
6273 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6274 target A or AAAA records (if not already returned) without resetting the
6275 qualify_single or search_parents options of the DNS resolver. These are
6276 inappropriate in this case because the targets of MX and SRV records must
6277 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6278 target that, when qualified, matched something in the local domain. These
6279 two options are now turned off when doing these lookups.
6281 57. It seems that at least some releases of Reiserfs (which does not have the
6282 concept of a fixed number of inodes) returns zero and not -1 for the
6283 number of available inodes. This interacted badly with check_spool_inodes,
6284 which assumed that -1 was the "no such thing" setting. What I have done is
6285 to check that the total number of inodes is greater than zero before doing
6286 the test of how many are available.
6288 58. When a "warn" ACL statement has a log_message modifier, the message is
6289 remembered, and not repeated. This is to avoid a lot of repetition when a
6290 message has many recipients that cause the same warning to be written.
6291 However, Exim was preserving the list of already written lines for an
6292 entire SMTP session, which doesn't seem right. The memory is now reset if a
6293 new message is started.
6295 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6296 debugging output unless log_rewrite was also set.
6298 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6299 of (int)(handle) when we know that handle contains (void *)(-1).
6301 61. The Exim daemon panic-logs an error return when it closes the incoming
6302 connection. However "connection reset by peer" seems to be common, and
6303 isn't really an error worthy of noting specially, so that particular error
6306 62. When Exim is trying to find all the local interfaces, it used to panic and
6307 die if the ioctl to get the interface flags failed. However, it seems that
6308 on at least one OS (Solaris 9) it is possible to have an interface that is
6309 included in the list of interfaces, but for which you get a failure error
6310 for this call. This happens when the interface is not "plumbed" into a
6311 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6312 failure of the "get flags" call assumes that the interface is down.
6314 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6315 makes life easier for people who are doing arithmetic on fields extracted
6316 from dates, where you often get leading zeros that should not be
6317 interpreted as octal.
6319 64. Added qualify_domain to the redirect router, to override the global
6322 65. If a pathologically long header line contained very many addresses (the
6323 report of this problem mentioned 10 000) and each of them was rewritten,
6324 Exim could use up a very large amount of memory. (It kept on making new
6325 copies of the header line as it rewrote, and never released the old ones.)
6326 At the expense of a bit more processing, the header rewriting function has
6327 been changed so that it no longer eats memory in this way.
6329 66. The generation of the Received: header has been moved from the time that a
6330 message starts to be received, to the time that it finishes. The timestamp
6331 in the Received: header should now be very close to that of the <= log
6332 line. There are two side-effects of this change:
6334 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6335 logged header lines no longer include the local Received: line, because
6336 it has not yet been created. The same applies to a copy of the message
6337 that is returned to a non-SMTP sender when a message is rejected.
6339 (b) When a filter file is tested using -bf, no additional Received: header
6340 is added to the test message. After some thought, I decided that this
6343 This change does not affect the value of $received_for. It is still set
6344 after address rewriting, but before local_scan() is called.
6346 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6348 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6349 gave an unhelpful panic error message, and a defer error. I have managed to
6350 change this behaviour so that it now rejects any supplied certificate,
6351 which seems right, as the list of acceptable certificates is empty.
6353 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6354 gave an unhelpful defer error. I have not managed to make this reject any
6355 supplied certificates, but the error message it gives is "no certificate
6356 supplied", which is not helpful.
6358 70. exigrep's output now also includes lines that are not associated with any
6359 message, but which match the given pattern. Implemented by a patch from
6360 Martin Sluka, which also tidied up the Perl a bit.
6362 71. Recipient callout verification, like sender verification, was using <> in
6363 the MAIL FROM command. This isn't really the right thing, since the actual
6364 sender may affect whether the remote host accepts the recipient or not. I
6365 have changed it to use the actual sender in the callout; this means that
6366 the cache record is now keyed on a recipient/sender pair, not just the
6367 recipient address. There doesn't seem to be a real danger of callout loops,
6368 since a callout by the remote host to check the sender would use <>.
6369 [SEE ABOVE: changed after hitting problems.]
6371 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6372 temporary errors. However, in the case of such a code being given after
6373 the end of a data transmission (i.e. after ".") Exim was failing to write
6374 a retry record for the message. (Yes, there was some broken host that was
6375 actually sending 8xx at this point.)
6377 73. An unknown lookup type in a host list could cause Exim to panic-die when
6378 the list was checked. (An example that provoked this was putting <; in the
6379 middle of a list instead of at the start.) If this happened during a DATA
6380 ACL check, a -D file could be left lying around. This kind of configuration
6381 error no longer causes Exim to die; instead it causes a defer error. The
6382 incident is still logged to the main and panic logs.
6384 74. Buglet left over from Exim 3 conversion. The message "too many messages
6385 in one connection" was written to the rejectlog but not the mainlog, except
6386 when address rewriting (yes!) was being logged.
6388 75. Added write_rejectlog option.
6390 76. When a system filter was run not as root (that is, when system_filter_user
6391 was set), the values of the $n variables were not being returned to the
6392 main process; thus, they were not subsequently available in the $sn
6395 77. Added +return_path_on_delivery log selector.
6397 78. A connection timeout was being treated differently from recipients deferred
6398 when testing hosts_max_try with a message that was older than the host's
6399 retry timeout. (The host should not be counted, thus allowing all hosts to
6400 be tried at least once before bouncing.) This may have been the cause of an
6401 occasionally reported bug whereby a message would remain on the queue
6402 longer than the retry timeout, but would be bounced if a delivery was
6403 forced. I say "may" because I never totally pinned down the problem;
6404 setting up timeout/retry tests is difficult. See also the next item.
6406 79. The ultimate address timeout was not being applied to errors that involved
6407 a combination of host plus message (for example, a timeout on a MAIL
6408 command). When an address resolved to a number of possible hosts, and they
6409 were not all tried for each delivery (e.g. because of hosts_max_try), a
6410 message could remain on the queue longer than the retry timeout.
6412 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6415 81. Fixed an obscure SMTP outgoing bug which required at least the following
6416 conditions: (a) there was another message waiting for the same server;
6417 (b) the server returned 5xx to all RCPT commands in the first message so
6418 that the message was not completed; (c) the server dropped the connection
6419 or gave a negative response to the RSET that Exim sends to abort the
6420 transaction. The observed case was a dropped connection after DATA that had
6421 been sent in pipelining mode. That is, the server had advertised PIPELINING
6422 but was not implementing it correctly. The effect of the bug was incorrect
6423 behaviour, such as trying another host, and this could lead to a crash.
6429 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6430 and daemon.c were passed as pointers to ints; they should have been
6431 pointers to socklen_t variables (which are typically unsigned ints).
6433 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6436 3. Fixed a really odd bug that affected only the testing scheme; patching a
6437 certain fixed string in the binary changed the value of another string that
6438 happened to be identical to the end of the original first string.
6440 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6441 name", it returns that address as the IP address. On some operating
6442 systems (e.g. Solaris), it also passes back the IP address string as the
6443 "host name". However, on others (e.g. Linux), it passes back an empty
6444 string. Exim wasn't checking for this, and was changing the host name to an
6445 empty string, assuming it had been canonicalized.
6447 5. Although rare, it is permitted to have more than one PTR record for a given
6448 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6449 all the names associated with an address, because they do in Solaris.
6450 However, it seems that they do not in Linux for data that comes from the
6451 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6452 I found this out when I moved to a new Linux workstation and tried to run
6453 the Exim test suite.
6455 To get round this problem I have changed the code so that it now does its
6456 own call to the DNS to look up PTR records when searching for a host name.
6457 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6458 addresses that are only in /etc/hosts are still found.
6460 This behaviour is, however, controlled by an option called host_lookup_
6461 order, which defaults to "bydns:byaddr". If people want to use the other
6462 order, or indeed, just use one or the other means of lookup, they can
6463 specify it in this variable.
6465 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6466 some operating systems, this comes back from gethostbyaddr() as an empty
6467 string, and this is what Exim used to test for. However, it seems that in
6468 other systems, "." is yielded. Exim now tests for this case too.
6470 7. The values of check_spool_space and check_log_space are now held internally
6471 as a number of kilobytes instead of an absolute number of bytes. If a
6472 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6473 kilobyte. This means that much larger values can be stored.
6475 8. Exim monitor: an attempt to get the action menu when not actually pointing
6476 at a message produces an empty menu entitled "No message selected". This
6477 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6478 no entries in it ("Shell widget menu has zero width and/or height"). So I
6479 have added a single, blank menu entry in this case.
6481 9. Added ${quote_local_part.
6483 10. MIME decoding is now applied to the contents of Subject: header lines when
6486 11. Now that a reference to $sender_host_address automatically causes a reverse
6487 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6488 host lookup before query-style lookups in lists that might use this
6489 variable. This has therefore been abolished, and the "net-" prefix is no
6490 longer necessary for query-style lookups.
6492 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6493 have been a typo for LFLAGS, so it has been changed.
6495 13. The install script calls Exim with "-C /dev/null" in order to find the
6496 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6497 to be output. However, since Exim outputs its version number before the
6498 error, it didn't break the script. It just looked ugly. I fixed this by
6499 always allowing "-C /dev/null" if the caller is root.
6501 14. Ignore overlarge ACL variable number when reading spool file - insurance
6502 against a later release with more variables having written the file.
6504 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6505 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6506 this, as well as the form without the "IPv6" on the front.
6508 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6509 OS/Makefile-Darwin file.
6511 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6512 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6514 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6515 variable. However, some OS use ints for inodes. I've added cast to long int
6516 to get rid of the compiler warning.
6518 19. I had forgotten to lock out "/../" in configuration file names when
6519 ALT_CONFIG_PREFIX was set.
6521 20. Routers used for verification do not need to specify transports. However,
6522 if such a router generated a host list, and callout was configured, Exim
6523 crashed, because it could not find a port number from the (non-existent)
6524 transport. It now assumes port 25 in this circumstance.
6526 21. Added the -t option to exigrep.
6528 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6529 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6530 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6531 them is set, LOOKUP_LSEARCH is forced.
6533 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6535 24. Added sender and host information to the "rejected by local_scan()" log
6536 line; previously there was no indication of these.
6538 25. Added .include_if_exists.
6540 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6541 Linux. It turns out that not all file systems support this. Apparently some
6542 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6543 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6544 not supported on the file descriptor, is now ignored when Exim is trying to
6545 sync a directory. This applies only to Linux.
6547 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6549 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6550 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6553 29. Added support for maildirsize files from supplied patch (modified a bit).
6555 30. The use of :fail: followed by an empty string could lead Exim to respond to
6556 sender verification failures with (e.g.):
6558 550 Verification failed for <xxx>
6559 550 Sender verify failed
6561 where the first response line was missing the '-' that indicates it is not
6562 the final line of the response.
6564 31. The loop for finding the name of the user that called Exim had a hardwired
6565 limit of 10; it now uses the value of finduser_retries, which is used for
6566 all other user lookups.
6568 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6571 33. Exim was neglecting to zero errno before one call of strtol() when
6572 expanding a string and expecting an integer value. On some systems this
6573 resulted in spurious "integer overflow" errors. Also, it was casting the
6574 result into an int without checking.
6576 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6577 did not work. The code looks as if it has *never* worked, though it appears
6578 to have been documented since at least release 1.62. I have made it work.
6580 35. The "timeout_DNS" error in retry rules, also documented since at least
6581 1.62, also never worked. As it isn't clear exactly what this means, and
6582 clearly it isn't a major issue, I have abolished the feature by treating it
6583 as "timeout", and writing a warning to the main and panic logs.
6585 36. The display of retry rules for -brt wasn't always showing the error code
6588 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6589 timeout_connect_A, timeout_connect_MX.
6591 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6592 to the empty sender.
6594 39. The daemon was not analysing the content of -oX till after it had closed
6595 stderr and disconnected from the controlling terminal. This meant that any
6596 syntax errors were only noted on the panic log, and the return code from
6597 the command was 0. By re-arranging the code a little, I've made the
6598 decoding happen first, so such errors now appear on stderr, and the return
6599 code is 1. However, the actual setting up of the sockets still happens in
6600 the disconnected process, so errors there are still only recorded on the
6603 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6604 connections (as happens on some IP stacks) was logged at start up time as
6605 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6606 it from "IPv6 and IPv4", which means that two separate sockets are being
6609 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6610 says whether AF_INET or AF_INET6 was passed as an argument.
6612 42. Exiwhat output was messed up when time zones were included in log
6615 43. Exiwhat now gives more information about the daemon's listening ports,
6616 and whether -tls-on-connect was used.
6618 44. The "port" option of the smtp transport is now expanded.
6620 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6621 silently ignored. Now an error message is written to the main and panic
6624 46. There's a new ACL modifier called "logwrite" which writes to a log file
6625 as soon as it is encountered.
6627 47. Added $local_user_uid and $local_user_gid at routing time.
6629 48. Exim crashed when trying to verify a sender address that was being
6632 49. Exim was recognizing only a space character after ".include". It now also
6633 recognizes a tab character.
6635 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6636 extracting the relevant information from the specification. The man page no
6637 longer contains scrambled data for the -d option, and I've added a section
6638 at the front about calling Exim under different names.
6640 51. Added "extra_headers" argument to the "mail" command in filter files.
6642 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6645 53. Installed eximstats 1.29.
6647 54. Added transport_filter_timeout as a generic transport option.
6649 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6650 Cc: header lines. This was required by RFC 822, but it not required by RFC
6653 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6654 incoming messages that did not have them. Now it does so only if the
6655 message originates locally, that is, if there is no associated remote host
6656 address. When Resent- header lines are present, this applies to the Resent-
6657 lines rather than the non-Resent- lines.
6659 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6660 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6662 58. Messages for configuration errors now include the name of the main
6663 configuration file - useful now that there may be more than one file in a
6664 list (.included file names were always shown).
6666 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6667 for those rare installations that do not start the daemon as root or run it
6668 setuid root. I've cut out the call to initgroups() if the daemon is not
6671 60. The Exim user and group can now be bound into the binary as text strings
6672 that are looked up at the start of Exim's processing.
6674 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6676 62. Added $mailstore_basename variable.
6678 63. Installed patch to sieve.c from Michael Haardt.
6680 64. When Exim failed to open the panic log after failing to open the main log,
6681 the original message it was trying to log was written to stderr and debug
6682 output, but if they were not available (the usual case in production), it
6683 was lost. Now it is written to syslog before the two lines that record the
6684 failures to open the logs.
6686 65. Users' Exim filters run in subprocesses under the user's uid. It is
6687 possible for a "deliver" command or an alias in a "personal" command to
6688 provoke an address rewrite. If logging of address rewriting is configured,
6689 this fails because the process is not running as root or exim. There may be
6690 a better way of dealing with this, but for the moment (because 4.30 needs
6691 to be released), I have disabled address rewrite logging when running a
6692 filter in a non-root, non-exim process.
6698 1. The buildconfig auxiliary program wasn't quoting the value set for
6699 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6700 not defined. This bug was masked in 4.22 by the effect that was fixed in
6703 2. Some messages that were rejected after a message id was allocated were
6704 shown as "incomplete" by exigrep. It no longer does this for messages that
6705 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6707 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6708 have allow_domain_literals set, the ID did not get logged in the <= line.
6709 Domain literals are now always recognized in Message-ID: header lines.
6711 4. The first argument for a ${extract expansion item is the key name or field
6712 number. Leading and trailing spaces in this item were not being ignored,
6713 causing some misleading effects.
6715 5. When deliver_drop_privilege was set, single queue runner processes started
6716 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6717 same command in the process it spins off) were not dropping privilege.
6719 6. When the daemon running as "exim" started a queue runner, it always
6720 re-executed Exim in the spun-off process. This is a waste of effort when
6721 deliver_drop_privilege is set. The new process now just calls the
6722 queue-runner function directly.
6728 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6731 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6732 ignored. Though the use of -oP was forcing the writing of a pid file, it
6733 was always written to the default place.
6735 3. If the message "no IP address found for host xxxx" is generated during
6736 incoming verification, it is now followed by identification of the incoming
6737 connection (so you can more easily find what provoked it).
6739 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6741 5. Added some features to "harden" Exim a bit more against certain attacks:
6743 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6744 be put in Local/Makefile. This is like the never_users runtime option,
6745 but it cannot be overridden. The default setting is "root".
6747 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6748 prefix string with which any file named in a -C command line option
6751 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6752 is retained for -C and -D only if the caller of Exim is root. Without
6753 it, the exim user may also use -C and -D and retain privilege.
6755 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6756 command line option is disabled.
6758 6. Macro names set by the -D option must start with an upper case letter, just
6759 like macro names defined in the configuration file.
6761 7. Added "dereference=" facility to LDAP.
6763 8. Two instances of the typo "uknown" in the source files are fixed.
6765 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6766 the Configure-Makefile script screwed up while processing it.
6768 10. Incorporated PCRE 4.4.
6770 11. The SMTP synchronization check was not operating right at the start of an
6771 SMTP session. For example, it could not catch a HELO sent before the client
6772 waited for the greeting. There is now a check for outstanding input at the
6773 point when the greeting is written. Because of the duplex, asynchronous
6774 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6775 way, but not yet received, when the check is performed.
6777 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6778 on TCP/IP sockets, because this apparently causes some broken clients to
6781 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6782 unchanged) from the Cygwin maintainer.
6784 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6785 was supported instead of testing for mbx. Effectively a typo.
6787 15. The spa authenticator server code was not checking that the input it
6788 received was valid base64.
6790 16. The debug output line for the "set" modifier in ACLs was not showing the
6791 name of the variable that was being set.
6793 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6795 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6796 lookup on its own. Something else was needed to trigger the lookup. For
6797 example, a match in host_lookup or the need for a host name in a host list.
6798 Now, if $sender_host_name is referenced and the host name has not yet been
6799 looked up, a lookup is performed. If the lookup fails, the variable remains
6800 empty, and $host_lookup_failed is set to "1".
6802 19. Added "eqi" as a case-independent comparison operator.
6804 20. The saslauthd authentication condition could segfault if neither service
6805 nor realm was specified.
6807 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6808 error message that was logged was misleading, and incoming SMTP
6809 connections were dropped. The message is now more accurate, and temporary
6810 errors are given to SMTP connections.
6812 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6813 (except RSET) with a 421 error, until QUIT is received. However, it was
6814 failing to send a response to QUIT.
6816 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6817 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6818 ACL is now run after the helo_try_verify_hosts code.
6820 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6821 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6822 case-independent, so other case variants are also recognized). Apparently
6823 some systems use these upper case variants.
6825 25. If more than two messages were waiting for the same host, and a transport
6826 filter was specified for the transport, Exim sent two messages over the
6827 same TCP/IP connection, and then failed with "socket operation on non-
6828 socket" when it tried to send the third.
6830 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6832 27. The extern definition of crypt16() in expand.c was not being excluded when
6833 the OS had its own crypt16() function.
6835 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6836 as a preferred synonym for return_size_limit, both as an option and as an
6839 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6841 30. Changed the default configuration ACL to relax the local part checking rule
6842 for addresses that are not in any local domains. For these addresses,
6843 slashes and pipe symbols are allowed within local parts, but the sequence
6844 /../ is explicitly forbidden.
6846 31. SPA server authentication was not clearing the challenge buffer before
6849 32. log_message in a "warn" ACL statement was writing to the reject log as
6850 well as to the main log, which contradicts the documentation and doesn't
6851 seem right (because no rejection is happening). So I have stopped it.
6853 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6854 However, I am unable to do any testing of this.
6856 34. Fixed an infelicity in the appendfile transport. When checking directories
6857 for a mailbox, to see if any needed to be created, it was accidentally
6858 using path names with one or more superfluous leading slashes; tracing
6859 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6861 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6862 contents are added to the log line that is written for every discarded
6863 recipient. (Previously a log_message setting was ignored.)
6865 36. The ${quote: operator now quotes the string if it is empty.
6867 37. The install script runs exim in order to find its version number. If for
6868 some reason other than non-existence or emptiness, which it checks, it
6869 could not run './exim', it was installing it with an empty version number,
6870 i.e. as "exim-". This error state is now caught, and the installation is
6873 38. An argument was missing from the function that creates an error message
6874 when Exim fails to connect to the socket for saslauthd authentication.
6875 This could cause Exim to crash, or give a corrupted message.
6877 39. Added isip, isip4, and isip6 to ${if conditions.
6879 40. The ACL variables $acl_xx are now saved with the message, and can be
6880 accessed later in routers, transports, and filters.
6882 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6883 strings in the file are not string-expanded.
6885 42. If a MAIL command specified a SIZE value that was too large to fit into an
6886 int variable, the check against message_size_limit failed. Such values are
6887 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6888 day this will have to be increased, but I don't think I want to be around
6889 when emails are that large.
6896 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6897 iconv() is not standard in FreeBSD.
6899 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6900 IPv6 enabled. The observed symptom was a segmentation fault on return from
6901 the function os_common_find_running_interfaces() in src/os.c.
6903 3. In the check_special_case() function in daemon.c I had used "errno" as an
6904 argument name, which causes warnings on some systems. This was basically a
6905 typo, since it was named "eno" in the comments!
6907 4. The code that waits for the clock to tick (at a resolution of some fraction
6908 of a second) so as to ensure message-id uniqueness was always waiting for
6909 at least one whole tick, when it could have waited for less. [This is
6910 almost certainly not relevant at current processor speeds, where it is
6911 unlikely to ever wait at all. But we try to future-proof.]
6913 5. The function that sleeps for a time interval that includes fractions of a
6914 second contained a race. It did not block SIGALRM between setting the
6915 timer, and suspending (a couple of lines later). If the interval was short
6916 and the sigsuspend() was delayed until after it had expired, the suspension
6917 never ended. On busy systems this could lead to processes getting stuck for
6920 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6921 process, before it forks any delivery processes. The open lookup caching
6922 mechanism meant that the open file or database connection was passed into
6923 the delivery process. The problem was that delivery processes always tidy
6924 up cached lookup data. This could cause a problem for the next delivery
6925 process started by the queue runner, because the external queue runner
6926 process does not know about the closure. So the next delivery process
6927 still has data in the lookup cache. In the case of a file lookup, there was
6928 no problem because closing a file descriptor in a subprocess doesn't affect
6929 the parent. However, if the lookup was caching a connection to a database,
6930 the connection was closed, and the second delivery process was likely to
6931 see errors such as "PGSQL: query failed: server closed the connection
6932 unexpectedly". The problem has been fixed by closing all cached lookups
6933 in a queue runner before running a delivery process.
6935 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6936 seem to have the "const" qualifier which it has on other OS. I've
6939 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6940 *of the same type* (client or server) with the same public name that an
6941 error should be diagnosed.
6943 9. When Exim looked up a host name for an IP address, but failed to find the
6944 original IP address when looking up the host name (a safety check), it
6945 output the message "<ip address> does not match any IP for NULL", which was
6946 confusing, to say the least. The bug was that the host name should have
6947 appeared instead of "NULL".
6949 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6950 user that is built into the binary, and the -C or -D options is used, root
6951 privilege is dropped before the configuration file is read. In addition,
6952 logging is switched to stderr instead of the normal log files. If the
6953 configuration then re-defines the Exim user, the unprivileged environment
6954 is probably not what is expected, so Exim logs a panic warning message (but
6957 However, if deliver_drop_privilege is set, the unprivileged state may well
6958 be exactly what is intended, so the warning has been cut out in that case,
6959 and Exim is allowed to try to write to its normal log files.
6965 1. smtp_return_error_details was not giving details for temporary sender
6966 or receiver verification errors.
6968 2. Diagnose a configuration error if two authenticators have the same public
6971 3. Exim used not to create the message log file for a message until the first
6972 delivery attempt. This could be confusing when incoming messages were held
6973 for policy or load reasons. The message log file is now created at the time
6974 the message is received, and an initial "Received" line is written to it.
6976 4. The automatically generated man page for command line options had a minor
6977 bug that caused no ill effects; however, a more serious problem was that
6978 the procedure for building the man page automatically didn't always
6979 operate. Consequently, release 4.20 contains an out-of-date version. This
6980 shouldn't happen again.
6982 5. When building Exim with embedded Perl support, the script that builds the
6983 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6984 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6986 6. The freeze_tell option was not being used for messages that were frozen on
6987 arrival, either by an ACL or by local_scan().
6989 7. Added the smtp_incomplete_transaction log selector.
6991 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6992 was accepting AUTH without a new EHLO.
6994 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6995 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6998 10. Logging of TCP/IP connections (when configured) now happens in the main
6999 daemon process instead of the child process, so that the TCP/IP connection
7000 count is more accurate (but it can never be perfect).
7002 11. The use of "drop" in a nested ACL was not being handled correctly in the
7003 outer ACL. Now, if condition failure induced by the nested "drop" causes
7004 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7005 or "require"), the connection is dropped.
7007 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7008 that yield "discard" can now be used with an "accept" or a "discard" verb,
7009 but an error is generated for any others (because I can't see a useful way
7010 to define what should happen).
7012 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7013 are now processed in the same way as lines in the Exim configuration file.
7014 In particular, continuation lines are supported.
7016 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7018 15. Added -ti meaning -t -i.
7020 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7021 domains, and warn by logging if others are found.
7023 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7024 returned from GIFCONF when Exim is trying to find the list of interfaces on
7025 a host. The code in os.c has been modified to copy each ifreq to an aligned
7026 structure in all cases.
7028 Also, in some cases, the returned ifreq's were being copied to a 'struct
7029 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7030 means the last couple of bytes of an IPv6 address could be chopped if the
7031 ifreq contained only a normal sockaddr (14 bytes storage).
7033 18. Named domain lists were not supported in the hosts_treat_as_local option.
7034 An entry such as +xxxx was not recognized, and was treated as a literal
7037 19. Ensure that header lines added by a DATA ACL are included in the reject log
7038 if the ACL subsequently rejects the message.
7040 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7041 MD5 (which is deprecated).
7043 21. When testing a filter file using -bf, Exim was writing a message when it
7044 took the sender from a "From " line in the message, but it was not doing so
7045 when it took $return_path from a Return-Path: header line. It now does.
7047 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7048 with a valid header line field name (a series of printing characters
7049 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7051 23. Changed "disc" in the source to "disk" to conform to the documentation and
7052 the book and for uniformity.
7054 24. Ignore Sendmail's -Ooption=value command line item.
7056 25. When execve() failed while trying to run a command in a pipe transport,
7057 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7058 could be confused with a return value of 69 from the command itself. This
7059 has been changed to 127, the value the shell returns if it is asked to run
7060 a non-existent command. The wording for the related log line suggests a
7061 non-existent command as the problem.
7063 26. If received_header_text expands to an empty string, do not add a Received:
7064 header line to the message. (Well, it adds a token one on the spool, but
7065 marks it "old" so that it doesn't get used or transmitted.)
7067 27. Installed eximstats 1.28 (addition of -nt option).
7069 28. There was no check for failure on the call to getsockname() in the daemon
7070 code. This can fail if there is a shortage of resources on the system, with
7071 ENOMEM, for example. A temporary error is now given on failure.
7073 29. Contrary to the C standard, it seems that in some environments, the
7074 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7075 program. Exim now does this explicitly; it affects the formatting of
7076 timestamps using strftime().
7078 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7079 complaints. I've now initialized all the variables, to avoid this.
7081 32. Header lines added by a system filter were not being "seen" during
7082 transport-time rewrites.
7084 33. The info_callback() function passed to OpenSSL is set up with type void
7085 (*)(SSL *, int, int), as described somewhere. However, when calling the
7086 function (actually a macro) that sets it up, the type void(*)() is
7087 expected. I've put in a cast to prevent warnings from picky compilers.
7089 34. If a DNS black list lookup found a CNAME record, but there were no A
7090 records associated with the domain it pointed at, Exim crashed.
7092 35. If a DNS black list lookup returned more than one A record, Exim ignored
7093 all but the first. It now scans all returned addresses if a particular IP
7094 value is being sought. In this situation, the contents of the
7095 $dnslist_value variable are a list of all the addresses, separated by a
7098 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7099 to accept a host name and all its aliases if the forward lookup for any of
7100 them yielded the IP address of the incoming connection. Now it accepts only
7101 those names whose forward lookup yields the correct IP address. Any other
7102 names are discarded. This closes a loophole whereby a rogue DNS
7103 administrator could create reverse DNS records to break through a
7104 wildcarded host restriction in an ACL.
7106 37. If a user filter or a system filter that ran in a subprocess used any of
7107 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7108 the wrong values were passed to the pipe command ($thisaddress had the
7109 value of $0, $0 had the value of $1, etc). This bug was introduced by
7110 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7112 38. Improved the line breaking for long SMTP error messages from ACLs.
7113 Previously, if there was no break point between 40 and 75 characters, Exim
7114 left the rest of the message alone. Two changes have been made: (a) I've
7115 reduced the minimum length to 35 characters; (b) if it can't find a break
7116 point between 35 and 75 characters, it looks ahead and uses the first one
7117 that it finds. This may give the occasional overlong line, but at least the
7118 remaining text gets split now.
7120 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7121 file descriptors might be low, and that setting 1000 would always raise it.
7122 It turns out that in some environments, the limit is already over 1000 and
7123 that lowering it causes trouble. So now Exim takes care not to decrease it.
7125 40. When delivering a message, the value of $return_path is set to $sender_
7126 address at the start of routing (routers may change the value). By an
7127 oversight, this default was not being set up when an address was tested by
7128 -bt or -bv, which affected the outcome if any router or filter referred to
7131 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7132 log only when "message" or "log_message" are set. However, if one of the
7133 conditions was an address verification, or a call to a nested ACL, the
7134 messages generated by the underlying test were being passed through. This
7135 no longer happens. The underlying message is available in $acl_verify_
7136 message for both "message" and "log_message" expansions, so it can be
7137 passed through if needed.
7139 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7140 new expansion $bh_ to give the encoded byte string without charset
7141 translation. Translation happens only if iconv() is available; HAVE_ICONV
7142 indicates this at build time. HEADERS_CHARSET gives the charset to
7143 translate to; headers_charset can change it in the configuration, and
7144 "headers charset" can change it in an individual filter file.
7146 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7147 that creates RFC 2047 encoded "words" labels them as that charset instead
7148 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7149 expansion operator; (ii) when Exim creates a From: line for a local
7150 message; (iii) when a header line is rewritten to include a "phrase" part.
7152 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7153 buggy, causing it to skip the first lines of messages whose message ID
7154 ended in 'D'. This would not have bitten before Exim release 4.14, because
7155 message IDs were unlikely to end in 'D' before then. The effect was to have
7156 incorrect size information for certain domains.
7158 45. #include "config.h" was missing at the start of the crypt16.c module. This
7159 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7162 46. If there was a timeout during a "random" callout check, Exim treated it as
7163 a failure of the random address, and carried on sending RSET and the real
7164 address. If the delay was just some slowness somewhere, the response to the
7165 original RCPT would be taken as a response to RSET and so on, causing
7166 mayhem of various kinds.
7168 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7169 when I implemented it. It didn't allow for the fact that some option values
7170 may legitimately be negative (e.g. size_addition), and it didn't even do
7171 the right test for positive values.
7173 48. Domain names in DNS records are case-independent. Exim always looks them up
7174 in lower case. Some resolvers return domain names in exactly the case they
7175 appear in the zone file, that is, they may contain uppercase letters. Not
7176 all resolvers do this - some return always lower case. Exim was treating a
7177 change of case by a resolver as a change of domain, similar to a widening
7178 of a domain abbreviation. This triggered its re-routing code and so it was
7179 trying to route what was effectively the same domain again. This normally
7180 caused routing to fail (because the router wouldn't handle the domain
7181 twice). Now Exim checks for this case specially, and just changes the
7182 casing of the domain that it ultimately uses when it transmits the message
7185 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7188 50. If a filter generated a file delivery with a non-absolute name (possible if
7189 no home directory exists for the router), the forbid_file option was not
7192 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7193 the existing equality matching.
7195 52. Exim was using ints instead of ino_t variables in some places where it was
7196 dealing with inode numbers.
7198 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7199 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7200 called TMPDIR, and if it finds it is different, it changes its value.
7202 54. The smtp_printf() function is now made available to local_scan() so
7203 additional output lines can be written before returning. There is also an
7204 smtp_fflush() function to enable the detection of a dropped connection.
7205 The variables smtp_input and smtp_batched_input are exported to
7208 55. Changed the default runtime configuration: the message "Unknown user"
7209 has been removed from the ACL, and instead placed on the localuser router,
7210 using the cannot_route_message feature. This means that any verification
7211 failures that generate their own messages won't get overridden. Similarly,
7212 the "Unrouteable address" message that was in the ACL for unverifiable
7213 relay addresses has also been removed.
7215 56. Added hosts_avoid_esmtp to the smtp transport.
7217 57. The exicyclog script was not checking for the esoteric option
7218 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7219 will work only if exicyclog is run under the appropriate euid.
7221 58. Following a discussion on the list, the rules by which Exim recognises line
7222 endings on incoming messages have been changed. The -dropcr and drop_cr
7223 options are now no-ops, retained only for backwards compatibility. The
7224 following line terminators are recognized: LF CRLF CR. However, special
7225 processing applies to CR:
7227 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7228 nor a local message in the state where . is a terminator.
7230 (ii) If a bare CR is encountered in a header line, an extra space is added
7231 after the line terminator so as not to end the header. The reasoning
7232 behind this is that bare CRs in header lines are most likely either
7233 to be mistakes, or people trying to play silly games.
7235 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7236 was being incorrectly given as 18 bytes larger than it should have been.
7237 This is a VOB (very old bug).
7239 60. This may never have affected anything current, but just in case it has:
7240 When the local host is found other than at the start of a list of hosts,
7241 the local host, those with the same MX, and any that follow, are discarded.
7242 When the list in question was part of a longer list of hosts, the following
7243 hosts (not currently being processed) were also being discarded. This no
7244 longer happens. I'm not sure if this situation could ever has previously
7247 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7250 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7251 header. This is something that is recommended in a new Internet Draft, and
7252 is something that is documented as being done by Sendmail. There are two
7253 possible values. For messages generated by the autoreply transport, Exim
7256 Auto-Submitted: auto-replied
7258 whereas for all other generated messages (e.g. bounces) it adds
7260 Auto-Submitted: auto-generated
7262 63. The "personal" condition in filters now includes a test for the
7263 Auto-Submitted: header. If it contains the string "auto-" the message it
7264 not considered personal.
7266 64. Added rcpt_include_affixes as a generic transport option.
7268 65. Added queue_only_override (default true).
7270 66. Added the syslog_duplication option.
7272 67. If what should have been the first header line of a message consisted of
7273 a space followed by a colon, Exim was mis-interpreting it as a header line.
7274 It isn't of course - it is syntactically invalid and should therefore be
7275 treated as the start of the message body. The misbehaviour could have
7276 caused a number of strange effects, including loss of data in subsequent
7277 header lines, and spool format errors.
7279 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7280 client host had authenticated. This control can now be exercised by an ACL
7281 for more flexibility.
7283 69. By default, callouts do not happen when testing with -bh. There is now a
7284 variant, -bhc, which does actually run the callout code, including
7285 consulting and updating the callout cache.
7287 70. Added support for saslauthd authentication, courtesy of Alexander
7290 71. If statvfs() failed on the spool or log directories while checking their
7291 size for availability, Exim confusingly gave the error "space shortage".
7292 Furthermore, in debugging mode it crashed with a floating point exception.
7293 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7294 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7295 serious problem, Exim now writes to the main and panic logs when this
7296 happens, with details of the failure. It then refuses to accept the
7297 incoming message, giving the message "spool directory problem" or "log
7298 directory problem" with a 421 code for SMTP messages.
7300 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7301 0, 1, and 2 exist, because some OS complain for execs without them (see
7302 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7303 descriptors. However, the code omitted to check that the open succeeded,
7304 causing mysterious errors if for some reason the permissions on /dev/null
7305 got screwed. Now Exim writes a message to the main and panic logs, and
7306 bombs out if it can't open /dev/null.
7308 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7309 interact so that it is all more flexible. It is supposed to remain
7310 backwards compatible. Also added extra_local_interfaces.
7312 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7313 to bomb out with an assertion failure - to the client this appears as a
7314 connection drop. This problem occurs in the part of the code that was taken
7315 from the Samba project. Fortunately, the assertion is in a very simple
7316 function, so I have fixed this by reproducing the function inline in the
7317 one place where it is called, and arranging for authentication to fail
7318 instead of killing the process with assert().
7320 75. The SPA client code was not working when the server requested OEM rather
7321 than Unicode encoding.
7323 76. Added code to make require_files with a specific uid setting more usable in
7324 the case where statting the file as root fails - usually a non-root-mounted
7325 NFS file system. When this happens and the failure is EACCES, Exim now
7326 forks a subprocess and does the per-uid checking as the relevant uid.
7328 77. Added process_log_path.
7330 78. If log_file_path was not explicitly set, a setting of check_log_space or
7331 check_log_inodes was ignored.
7333 79. If a space check for the spool or log partitions fails, the incident is now
7334 logged. Of course, in the latter case the data may get lost...
7336 80. Added the %p formatting code to string_format() so that it can be used to
7337 print addresses in debug_print(). Adjusted all the address printing in the
7338 debugging in store.c to use %p rather than %d.
7340 81. There was a concern that a line of code in smtp_in.c could overflow a
7341 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7342 initially expressed, the concern was not well-founded, because trailing
7343 spaces are removed early. However, if the trailing spaces were followed by
7344 a NULL, they did not get removed, so the overflow was possible. Two fixes
7347 (a) I re-wrote the offending code in a cleaner fashion.
7348 (b) If an incoming SMTP command contains a NULL character, it is rejected
7351 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7352 runs initgroups(), so that if the Exim user is in any additional groups,
7353 they will be used during message reception.
7359 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob