1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe functions being used.
17 JH/03 Bug 2269: When presented with a received message having a stupidly large
18 number of DKIM-Signature headers, disable DKIM verification to avoid
19 a resource-consumption attack. The limit is set at twenty.
21 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
22 report of oldest_pass in ${authres } in consequence, and separate out
23 some descriptions of reasons for verification fail.
25 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
26 files in the spool were present and unlocked. A queue-runner could spot
27 them, resulting in a duplicate delivery. Fix that by doing the unlock
28 after the unlink. Investigation by Tim Stewart. Take the opportunity to
29 add more error-checking on spoolfile handling while that code is being
32 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
33 No known attacks, no CVE, this is defensive hardening.
35 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
36 a queue-runner could start a delivery while other operations were ongoing.
37 Cutthrough delivery was a common victim, resulting in duplicate delivery.
38 Found and investigated by Tim Stewart. Fix by using the open message data
39 file handle rather than opening another, and not locally closing it (which
40 releases a lock) for that case, while creating the temporary .eml format
41 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
43 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
44 $sender_verify_failure/$recipient_verify_failure to "random".
46 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
49 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
50 Previously this would segfault.
52 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
55 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
56 like zero, since the resolver should be doing this for us, But we need one
57 as a CNAME but no MX presence gets the CNAME returned; we need to check
58 that doesn't point to an MX to declare it "no MX returned" rather than
59 "error, loop". A new main option is added so the older capability of
60 following some limited number of chain links is maintained.
62 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
64 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
65 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
66 now-deprecated earlier definitions used only specified the range up to TLS
67 1.2 (in the older-version library docs).
69 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
71 JH/15 Rework TLS client-side context management. Stop using a global, and
72 explicitly pass a context around. This enables future use of TLS for
73 connections to service-daemons (eg. malware scanning) while a client smtp
74 connection is using TLS; with cutthrough connections this is quite likely.
76 JH/16 Fix ARC verification to do AS checks in reverse order.
78 JH/17 Support a "tls" option on the ${readsocket } expansion item.
80 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
81 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
82 Previously the "utf8" would be re-prepended for every additional message.
84 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
85 Previously thery were accepted, resulting in issues when attempting to
86 forward messages to a non-supporting MTA.
88 PP/02 Let -n work with printing macros too, not just options.
90 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
91 one parent address was copied, and bogus data was used at delivery-logging
92 time. Either a crash (after delivery) or bogus log data could result.
93 Discovery and analysis by Tim Stewart.
95 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
96 Previously if the string ended mid-character, we did not insert the
97 promised '?' replacement.
99 PP/04 Documentation: current string operators work on bytes, not codepoints.
101 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
102 should pack well giving a smaller memory footprint so better caching and
103 therefore performance. Group the declarations where this can't be done so
104 that the byte-sized flag variables are not interspersed among pointer
105 variables, giving a better chance of good packing by the compiler.
107 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
108 non-null, to avoid issues with sites running BATV. Previously reports were
109 sent with an empty envelope sender so looked like bounces.
111 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
112 The ignore_error flag wasn't being returned from the filter subprocess so
113 was not set for later routers. Investigation and fix by Matthias Kurz.
115 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
116 and a msg:complete for the whole, when a message is manually removed using
117 -Mrm. Developement by Matthias Kurz, hacked on by JH.
119 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
120 a "Gnu special" function, asprintf() in the DB utility binary builds; I
121 hope that is portable enough.
123 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
124 requiring a known-CA anchor certificate; make it now rely entirely on the
125 TLSA as an anchor. Checking the name on the leaf cert against the name
126 on the A-record for the host is still done for TA (but not for EE mode).
128 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
129 included in delivery lines for non-proxied connections, when compiled with
130 SUPPORT_SOCKS and running with proxy logging enabled.
132 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
133 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
134 move the existing event to fire before the normal logging of message
135 failure so that custom logging is bracketed by normal logging.
137 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
138 msg:fail:internal event. Developement by Matthias Kurz.
140 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
141 far too small for todays use of crypto signatures stored there. Go all
142 the way to the max DNS message size of 64kB, even though this might be
143 overmuch for IOT constrained device use.
145 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
146 copy a string over itself. The library routine is documented as not
147 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
149 JH/32 For main options check_spool_space and check_inode_space, where the
150 platform supports 64b integers, support more than the previous 2^31 kB
151 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
152 the previous G, M, k.
154 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
155 $authenticated_fail_id variable on authentication failure. Previously
158 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
159 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
160 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
161 GNUTLS_SEC_PARAM_MEDIUM.
163 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
164 side. Previously we would continue as if no SNI had been received.
166 JH/36 Harden the handling of string-lists. When a list consisted of a sole
167 "<" character, which should be a list-separator specification, we walked
168 off past the nul-terimation.
170 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
171 causes) even when the retry time is not yet met. Previously they were
172 not, meaning that when (say) an account was over-quota and temp-rejecting,
173 and multiple senders' messages were queued, only one sender would get
174 notified on each configured delay_warning cycle.
176 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
178 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
179 cipher-suites, an error can be left on the stack even for a succeeding
180 accept; this results in impossible error messages when a later operation
183 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they return error
184 codes indicating retry. Under TLS1.3 this becomes required.
190 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
191 When redis_servers is set to a list of > 1 element, and the Redis servers
192 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
193 case of MOVED into a DEFER case instead, thus moving the query onto the
194 next server in the list. For a cluster of N elements, all N servers must
195 be defined in redis_servers.
197 GF/02 Catch and remove uninitialized value warning in exiqsumm
198 Check for existence of @ARGV before looking at $ARGV[0]
200 JH/01 Replace the store_release() internal interface with store_newblock(),
201 which internalises the check required to safely use the old one, plus
202 the allocate and data copy operations duplicated in both (!) of the
203 extant use locations.
205 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
206 modifier. This matches the restriction on the commandline.
208 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
209 Previously only the last row was returned.
211 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
212 we assumed that tags in the header were well-formed, and parsed the
213 element content after inspecting only the first char of the tag.
214 Assumptions at that stage could crash the receive process on malformed
217 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
218 While running the DKIM ACL we operate on the Permanent memory pool so that
219 variables created with "set" persist to the DATA ACL. Also (at any time)
220 DNS lookups that fail create cache records using the Permanent pool. But
221 expansions release any allocations made on the current pool - so a dnsdb
222 lookup expansion done in the DKIM ACL releases the memory used for the
223 DNS negative-cache, and bad things result. Solution is to switch to the
224 Main pool for expansions.
225 While we're in that code, add checks on the DNS cache during store_reset,
226 active in the testsuite.
227 Problem spotted, and debugging aided, by Wolfgang Breyha.
229 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
230 When none of the hosts presented to a transport match an already-open
231 connection, close it and proceed with the list. Previously we would
232 queue the message. Spotted by Lena with Yahoo, probably involving
235 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
236 Previously a spurious "250 OK id=" response was appended to the proper
239 JH/08 The "support for" informational output now, which built with Content
240 Scanning support, has a line for the malware scanner interfaces compiled
241 in. Interface can be individually included or not at build time.
243 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
244 by the template makefile "src/EDITME". The "STREAM" support for an older
245 ClamAV interface method is removed.
247 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
248 rows affected is given instead).
250 JH/11 The runtime Berkeley DB library version is now additionally output by
251 "exim -d -bV". Previously only the compile-time version was shown.
253 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
254 SMTP connection. Previously, when one had more receipients than the
255 first, an abortive onward connection was made. Move to full support for
256 multiple onward connections in sequence, handling cutthrough connection
257 for all multi-message initiating connections.
259 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
260 routers. Previously, a multi-recipient message would fail to match the
261 onward-connection opened for the first recipient, and cause its closure.
263 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
264 a timeout on read on a GnuTLS initiating connection, resulting in the
265 initiating connection being dropped. This mattered most when the callout
266 was marked defer_ok. Fix to keep the two timeout-detection methods
269 JH/15 Relax results from ACL control request to enable cutthrough, in
270 unsupported situations, from error to silently (except under debug)
271 ignoring. This covers use with PRDR, frozen messages, queue-only and
274 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
276 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
277 metadata, resulting in a crash in free().
279 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
280 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
281 Broken also in d185889f4, with init system revamp.
283 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
284 Previously we abruptly closed the connection after reading a malware-
285 found indication; now we go on to read the "scan ok" response line,
288 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
289 ACL. Previously, a crash would result.
291 JH/19 Speed up macro lookups during configuration file read, by skipping non-
292 macro text after a replacement (previously it was only once per line) and
293 by skipping builtin macros when searching for an uppercase lead character.
295 JH/20 DANE support moved from Experimental to mainline. The Makefile control
296 for the build is renamed.
298 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
299 was allocated for every new TLS startup, meaning one per message. Fix
300 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
302 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
303 reported the original. Fix to report (as far as possible) the ACL
304 result replacing the original.
306 JH/23 Fix memory leak during multi-message connections using STARTTLS under
307 OpenSSL. Certificate information is loaded for every new TLS startup,
308 and the resources needed to be freed.
310 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
312 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
315 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
316 DATA response info to the (existing) per-recipient response info for
317 the "C=" log element. It can have useful tracking info from the
318 destination system. Patch from Simon Arlott.
320 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
321 length value. Previously this would segfault.
323 HS/02 Support Avast multiline protoocol, this allows passing flags to
324 newer versions of the scanner.
326 JH/28 Ensure that variables possibly set during message acceptance are marked
327 dead before release of memory in the daemon loop. This stops complaints
328 about them when the debug_store option is enabled. Discovered specifically
329 for sender_rate_period, but applies to a whole set of variables.
330 Do the same for the queue-runner and queue-list loops, for variables set
331 from spool message files. Do the same for the SMTP per-message loop, for
332 certain variables indirectly set in ACL operations.
334 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
335 as a multi-recipient message from a mailinglist manager). The coding had
336 an arbitrary cutoff number of characters while checking for more input;
337 enforced by writing a NUL into the buffer. This corrupted long / fast
338 input. The problem was exposed more widely when more pipelineing of SMTP
339 responses was introduced, and one Exim system was feeding another.
340 The symptom is log complaints of SMTP syntax error (NUL chars) on the
341 receiving system, and refused recipients seen by the sending system
342 (propating to people being dropped from mailing lists).
343 Discovered and pinpointed by David Carter.
345 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
346 replaced by the ${authresults } expansion.
348 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
350 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
351 allows proper process termination in container environments.
353 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
354 Previously the "final dot" had a newline after it; ensure it is CR,LF.
356 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
357 and "err_perm", deprecated since 4.83 when the RFC-defined words
358 "temperror" and "permerror" were introduced.
360 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
361 transport-filters or DKIM-signing. The restriction was lost in the
362 consolidation of verify-callout and delivery SMTP handling.
363 Extend the restriction to also cover ARC-signing.
365 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
366 in defer=pass mode supply a 450 to the initiator. Previously the message
369 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
370 tls_require_ciphers is used as before.
372 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
373 "pass_unscanned". Only tmpfails from the scanner are written to
374 the paniclog, as they may require admin intervention (permission
375 denied, license issues). Other scanner errors (like decompression
376 bombs) do not cause a paniclog entry.
378 JH/36 Fix reinitialisation of DKIM logging variable between messages.
379 Previously it was possible to log spurious information in receive log
382 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
383 triggered odd behaviour from Outlook Express clients.
385 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
388 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
389 since the IETF WG has not yet settled on that versus the original
390 "bare" representation.
392 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
393 Previously the millisecond value corrupted the output.
394 Fix also for syslog_pid=no and log_selector +pid, for which the pid
395 corrupted the output.
401 JH/01 Rework error string handling in TLS interface so that the caller in
402 more cases is responsible for logging. This permits library-sourced
403 string to be attached to addresses during delivery, and collapses
404 pairs of long lines into single ones.
406 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
407 during configuration. Wildcards are allowed and expanded.
409 JH/02 Rework error string handling in DKIM to pass more info back to callers.
410 This permits better logging.
412 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
413 do not close it down and have the child transport start it up again on
414 the passed-on TCP connection. Instead, proxy the child (and any
415 subsequent ones) for TLS via a unix-domain socket channel. Logging is
416 affected: the continued delivery log lines do not have any DNSSEC, TLS
417 Certificate or OCSP information. TLS cipher information is still logged.
419 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
420 identical IP addresses on different listening ports. Will also affect
423 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
424 add noisy ifdef guards to special-case this sillyness.
425 Patch from Bernd Kuhls.
427 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
428 than 255 are no longer allowed.
430 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
431 Disable the session-cache too, which might reduce our load. Since we
432 currrectly use a new context for every connection, both as server and
433 client, there is no benefit for these.
434 GnuTLS appears to not support tickets server-side by default (we don't
435 call gnutls_session_ticket_enable_server()) but client side is enabled
436 by default on recent versions (3.1.3 +) unless the PFS priority string
439 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
440 <https://reproducible-builds.org/specs/source-date-epoch/>.
442 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
443 the check for any unsuccessful recipients did not notice the limit, and
444 erroneously found still-pending ones.
446 JH/08 Pipeline CHUNKING command and data together, on kernels that support
447 MSG_MORE. Only in-clear (not on TLS connections).
449 JH/09 Avoid using a temporary file during transport using dkim. Unless a
450 transport-filter is involved we can buffer the headers in memory for
451 creating the signature, and read the spool data file once for the
452 signature and again for transmission.
454 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
455 4.77 as the kernel support then wasn't solid, having issues in 64bit
456 mode. Now, it's been long enough. Add support for FreeBSD also.
458 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
459 case where the routing stage had gathered several addresses to send to
460 a host before calling the transport for the first, we previously failed
461 to close down TLS in the old transport process before passing the TCP
462 connection to the new process. The new one sent a STARTTLS command
463 which naturally failed, giving a failed delivery and bloating the retry
464 database. Investigation and fix prototype from Wolfgang Breyha.
466 JH/12 Fix check on SMTP command input synchronisation. Previously there were
467 false-negatives in the check that the sender had not preempted a response
468 or prompt from Exim (running as a server), due to that code's lack of
469 awareness of the SMTP input buffering.
471 PP/04 Add commandline_checks_require_admin option.
472 Exim drops privileges sanely, various checks such as -be aren't a
473 security problem, as long as you trust local users with access to their
474 own account. When invoked by services which pass untrusted data to
475 Exim, this might be an issue. Set this option in main configuration
476 AND make fixes to the calling application, such as using `--` to stop
479 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
480 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
481 responses to those, into a single TLS record each way (this usually means
482 a single packet). As a side issue, smtp_enforce_sync now works on TLS
485 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
486 affects you only if you're dancing at the edge of the param size limits.
487 If you are, and this message makes sense to you, then: raise the
488 configured limit or use OpenSSL 1.1. Nothing we can do for older
491 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
492 cmdline element to get the documented default one. Previously it was
495 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
496 are unsuable for non-message-specific reasons - eg. connection timeout,
497 banner-time rejection.
499 JH/16 Fix logging of delivery remote port, when specified by router, under
502 PP/06 Repair manualroute's ability to take options in any order, even if one
503 is the name of a transport.
506 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
508 JH/17 Change the list-building routines interface to use the expanding-string
509 triplet model, for better allocation and copying behaviour.
511 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
512 Previously it was constructed the first time a possibly-matching string
513 was met in the configuration file input during startup; now it is done
516 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
517 compatible one, to avoid the (poorly documented) possibility of a config
518 file in the working directory redirecting the DB files, possibly correpting
519 some existing file. CVE-2017-10140 assigned for BDB.
521 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
522 cache-hot. Previously, although the result was properly cached, the
523 initial verify call returned a defer.
525 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
526 the main verify for receipient in uncached-mode.
528 JH/22 Retire historical build files to an "unsupported" subdir. These are
529 defined as "ones for which we have no current evidence of testing".
531 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
532 if present. Previously it was ignored.
534 JH/24 Start using specified-initialisers in C structure init coding. This is
535 a C99 feature (it's 2017, so now considered safe).
537 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
538 if was a fixed-sized field and bitmask ops via macros; it is now more
541 PP/07 GitHub PR 56: Apply MariaDB build fix.
542 Patch provided by Jaroslav Škarvada.
544 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
545 during Coverity cleanups [4.87 JH/47]
546 Diagnosis and fix provided by Michael Fischer v. Mollard.
548 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
549 the right size to place the terminating semicolon on its own folded
550 line, the header hash was calculated to an incorrect value thanks to
551 the (relaxed) space the fold became.
553 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
554 and confused the parent.
556 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
557 which could crash as a result. This could lead to undeliverable messages.
559 JH/28 Logging: "next input sent too soon" now shows where input was truncated
562 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
563 matters on fast-turnover and PID-randomising systems, which were getting
564 out-of-order delivery.
566 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
567 a possibly-overlapping copy. The symptom was that "Remote host closed
568 connection in response to HELO" was logged instead of the actual 4xx
571 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
572 Previously only that bufferd was discarded, resulting in SYMTP command
575 JH/32 DKIM: when a message has multiple signatures matching an identity given
576 in dkim_verify_signers, run the dkim acl once for each. Previously only
577 one run was done. Bug 2189.
579 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
580 "panic the current process" to "deliberately defer". The panic log is
581 still written with the problem list name; the mail and reject logs now
582 get a temp-reject line for the message that was being handled, saying
583 something like "domains check lookup or other defer". The SMTP 451
584 message is still "Temporary local problem".
586 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
587 A crafted sequence of BDAT commands could result in in-use memory beeing
588 freed. CVE-2017-16943.
590 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
591 from SMTP input. Previously it was always done; now only done for DATA
592 and not BDAT commands. CVE-2017-16944.
594 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
595 to the message (such as an overlong header line). Previously this was
596 not done and we did not exit BDAT mode. Followon from the previous item
597 though a different problem.
603 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
604 than -2003 did; needs libidn2 in addition to libidn.
606 JH/02 The path option on a pipe transport is now expanded before use.
608 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
609 Patch provided by "Björn", documentation fix added too.
611 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
612 missing a wire-to-host endian conversion.
614 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
615 close after a BDAT command line could be taken as a following command,
616 giving a synch failure. Fix by only checking for synch immediately
617 before acknowledging the chunk.
619 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
620 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
621 macro. Patches provided by Josh Soref.
623 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
624 Previously we did not; the RFC seems ambiguous and VRFY is not listed
625 by IANA as a service extension. However, John Klensin suggests that we
628 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
629 the dkim code may be unix-mode line endings rather than smtp wire-format
630 CRLF, so prepend a CR to any bare LF.
632 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
633 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
635 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
636 the first were themselves being wrongly included in the feed into dkim
637 processing; with most chunk sizes in use this resulted in an incorrect
638 body hash calculated value.
640 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
641 DKIM signature block, for verification. Although advised against by
642 standards it is specifically not ruled illegal.
644 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
646 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
647 missing a body hash (the bh= tag).
649 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
650 It seems that HAProxy sends the Proxy Protocol information in clear and
651 only then does a TLS startup, so do the same.
653 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
654 TCP connections (such as for Spamd) unless the daemon successfully set
655 Fast Open mode on its listening sockets. This fixes breakage seen on
656 too-old kernels or those not configured for Fast Open, at the cost of
657 requiring both directions being enabled for TFO, and TFO never being used
658 by non-daemon-related Exim processes.
660 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
661 endings, at least on the first header line. Try to canonify any that get
662 past that check, despite the cost.
664 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
665 now limited to an arbitrary five deep, while parsing addresses with the
666 strip_excess_angle_brackets option enabled.
668 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
669 instead leave the unprompted TLS handshake in socket buffer for the
670 TLS library to consume.
672 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
674 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
676 JH/16 Drop variables when they go out of scope. Memory management drops a whole
677 region in one operation, for speed, and this leaves assigned pointers
678 dangling. Add checks run only under the testsuite which checks all
679 variables at a store-reset and panics on a dangling pointer; add code
680 explicitly nulling out all the variables discovered. Fixes one known
681 bug: a transport crash, where a dangling pointer for $sending_ip_address
682 originally assigned in a verify callout, is re-used.
684 PP/06 Drop '.' from @INC in various Perl scripts.
686 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
688 PP/08 Reduce a number of compilation warnings under clang; building with
689 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
690 should be warning-free.
692 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
694 HS/01 Fix portability problems introduced by PP/08 for platforms where
695 realloc(NULL) is not equivalent to malloc() [SunOS et al].
697 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
698 chunk. This allows us to accept broken chunked messages. We need a more
699 general solution here.
701 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
702 already-broken messages in the queue.
704 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
706 JH/19 Fix reference counting bug in routing-generated-address tracking.
712 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
713 supports it and a size is available (ie. the sending peer gave us one).
715 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
716 years of being deprecated). The replacements are the ACLs
717 acl_smtp_mime and acl_not_smtp_mime.
719 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
720 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
721 or even an in-clear connection were permitted. Now, if the host lookup
722 was dnssec and dane was requested then the host is only used if the
723 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
724 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
725 if one fails this test.
726 This means that a poorly-configured remote DNS will make it incommunicado;
727 but it protects against a DNS-interception attack on it.
729 JH/04 Bug 1810: make continued-use of an open smtp transport connection
730 non-noisy when a race steals the message being considered.
732 JH/05 If main configuration option tls_certificate is unset, generate a
733 self-signed certificate for inbound TLS connections.
735 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
736 in rewrites and routers.
738 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
739 and logged a warning sing 4.83; now they are a configuration file error.
741 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
742 (lacking @domain). Apply the same qualification processing as RCPT.
744 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
746 JH/10 Support ${sha256:} applied to a string (as well as the previous
749 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
750 a cutthrough deliver is pending, as we always want to make a connection.
751 This also avoids re-routing the message when later placing the cutthrough
752 connection after a verify cache hit.
753 Do not update it with the verify result either.
755 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
756 when routing results in more than one destination address.
758 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
759 signing (which inhibits the cutthrough capability). Previously only
760 the presence of an option was tested; now an expansion evaluating as
761 empty is permissible (obviously it should depend only on data available
762 when the cutthrough connection is made).
764 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
765 the relevant preceding SMTP command did not note the pipelining mode.
767 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
768 Previously they were not counted.
770 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
771 as one having no matching records. Previously we deferred the message
772 that needed the lookup.
774 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
775 distinguished as "(=".
777 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
778 for missing MX records. Previously it only worked for missing A records.
780 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
782 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
783 after the data-go-ahead and data-ack. Patch from Jason Betts.
785 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
786 even for a "none" policy. Patch from Tony Meyer.
788 JH/22 Fix continued use of a connection for further deliveries. If a port was
789 specified by a router, it must also match for the delivery to be
792 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
793 When one of the recipients of a message was unsuitable for the connection
794 (has no matching addresses), we lost track of needing to mark it
795 deferred. As a result mail would be lost.
797 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
799 JH/25 Decoding ACL controls is now done using a binary search; the source code
800 takes up less space and should be simpler to maintain. Merge the ACL
801 condition decode tables also, with similar effect.
803 JH/26 Fix problem with one_time used on a redirect router which returned the
804 parent address unchanged. A retry would see the parent address marked as
805 delivered, so not attempt the (identical) child. As a result mail would
808 JH/27 Fix a possible security hole, wherein a process operating with the Exim
809 UID can gain a root shell. Credit to http://www.halfdog.net/ for
810 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
813 JH/28 Enable {spool,log} filesystem space and inode checks as default.
814 Main config options check_{log,spool}_{inodes,space} are now
815 100 inodes, 10MB unless set otherwise in the configuration.
817 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
818 Previously it only applied to the main-section connection policy
821 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
823 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
824 by me. Added RFC7919 DH primes as an alternative.
826 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
827 are not in the system include path.
829 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
830 GnuTLS, when a session startup failed (eg because the client disconnected)
831 Exim did stdio operations after fclose. This was exposed by a recent
832 change which nulled out the file handle after the fclose.
834 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
835 signed directly by the cert-signing cert, rather than an intermediate
836 OCSP-signing cert. This is the model used by LetsEncrypt.
838 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
840 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
841 an incoming connection.
843 HS/02 Bug 1802: Do not half-close the connection after sending a request
846 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
847 fallback to "prime256v1".
849 JH/34 SECURITY: Use proper copy of DATA command in error message.
850 Could leak key material. Remotely exploitable. CVE-2016-9963.
856 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
857 and 3.4.4 - once the server is enabled to respond to an OCSP request
858 it does even when not requested, resulting in a stapling non-aware
859 client dropping the TLS connection.
861 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
862 support variable-length bit vectors. No functional change.
864 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
865 The I= interface field on outgoing lines is now after the H= remote
866 host field, same as incoming lines. There is a separate
867 outgoing_interface log selector which allows you to disable the
870 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
871 If not running log_selector +smtp_connection the mainlog would be held
872 open indefinitely after a "too many connections" event, including to a
873 deleted file after a log rotate. Leave the per net connection logging
874 leaving it open for efficiency as that will be quickly detected by the
875 check on the next write.
877 HS/01 Bug 1671: Fix post transport crash.
878 Processing the wait-<transport> messages could crash the delivery
879 process if the message IDs didn't exist for some reason. When
880 using 'split_spool_directory=yes' the construction of the spool
881 file name failed already, exposing the same netto behaviour.
883 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
884 mime_regex ACL conditions.
886 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
887 to DSN fail messages (bounces): remote IP, remote greeting, remote response
888 to HELO, local diagnostic string.
890 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
891 log line to debug. Even when configured with a tls authenticator many
892 client connections are expected to not authenticate in this way, so
893 an authenticate fail is not an error.
895 HS/02 Add the Exim version string to the process info. This way exiwhat
896 gives some more detail about the running daemon.
898 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
899 matter for fast-change records such as DNSBLs.
901 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
902 retry record, even if constant. There may be multiple transports with
903 different interface settings and the retry behaviour needs to be kept
906 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
908 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
910 JH/10 Bug 840: fix log_defer_output option of pipe transport
912 JH/11 Bug 830: use same host for all RCPTS of a message, even under
913 hosts_randomize. This matters a lot when combined with mua_wrapper.
915 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
916 ${quote_pgsql:<string>} operator.
918 JH/13 Bug 1708: avoid misaligned access in cached lookup.
920 JH/14 Change header file name for freeradius-client. Relevant if compiling
921 with Radius support; from the Gentoo tree and checked under Fedora.
923 JH/15 Bug 1712: Introduce $prdr_requested flag variable
925 JH/16 Bug 1714: Permit an empty string as expansion result for transport
926 option transport_filter, meaning no filtering.
928 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
930 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
931 defaults to "*" (all hosts). The variable is now available when not built
932 with TLS, default unset, mainly to enable keeping the testsuite sane.
933 If a server certificate is not supplied (via tls_certificate) an error is
934 logged, and clients will find TLS connections fail on startup. Presumably
935 they will retry in-clear.
936 Packagers of Exim are strongly encouraged to create a server certificate
937 at installation time.
939 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
940 with the $config_file variable.
942 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
943 in transport context, after the attempt, and per-recipient. The latter type
944 is per host attempted. The event data is the error message, and the errno
945 information encodes the lookup type (A vs. MX) used for the (first) host,
946 and the trailing two digits of the smtp 4xx response.
948 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
949 to write to mainlog (or rejectlog, paniclog) in the window between file
950 creation and permissions/ownership being changed. Particularly affects
951 installations where exicyclog is run as root, rather than exim user;
952 result is that the running daemon panics and dies.
954 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
956 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
957 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
958 "pri" and "weight". Note that the previous implicit priority given by the
959 list order is no longer honoured.
961 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
964 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
965 by defining SUPPORT_SOCKS.
967 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
968 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
969 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
970 variables are renamed to proxy_{local,external}_{address,port}.
972 JH/27 Move Internationalisation support from Experimental to mainline, enabled
973 for a build by defining SUPPORT_I18N
975 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
976 of the query string, and make ${quote_redis:} do that quoting.
978 JH/29 Move Events support from Experimental to mainline, enabled by default
979 and removable for a build by defining DISABLE_EVENT.
981 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
983 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
984 cached by the daemon.
986 JH/32 Move Redis support from Experimental to mainline, enabled for a build
987 by defining LOOKUP_REDIS. The libhiredis library is required.
989 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
990 keys are given for lookup.
992 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
993 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
994 only supported when built with TLS support. The PolarSSL SHA routines
995 are still used when the TLS library is too old for convenient support.
997 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
998 openssl_options), for security. OpenSSL forces this from version 1.1.0
999 server-side so match that on older versions.
1001 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1002 allocation for $value could be released as the expansion processing
1003 concluded, but leaving the global pointer active for it.
1005 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1006 and to use the domains and local_parts ACL conditions.
1008 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1009 incorrectly not doubled on cutthrough transmission, hence seen as a
1010 body-termination at the receiving system - resulting in truncated mails.
1011 Commonly the sender saw a TCP-level error, and retransmitted the message
1012 via the normal store-and-forward channel. This could result in duplicates
1013 received - but deduplicating mailstores were liable to retain only the
1014 initial truncated version.
1016 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1018 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1020 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1021 we're in there, support oversigning also; bug 1309.
1023 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1025 HS/04 Add support for keep_environment and add_environment options.
1027 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1028 either intentional arithmetic overflow during PRNG, or testing config-
1031 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1032 delivery resulted in actual delivery. Cancel cutthrough before DATA
1035 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1036 rejects a recipient: pass the reject to the originator.
1038 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1039 Many were false-positives and ignorable, but it's worth fixing the
1042 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1043 for the new environment-manipulation done at startup. Move the routines
1044 from being local to tls.c to being global via the os.c file.
1046 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1047 an extract embedded as result-arg for a map, the first arg for extract
1048 is unavailable so we cannot tell if this is a numbered or keyed
1049 extraction. Accept either.
1055 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1058 JH/02 The smtp transport option "multi_domain" is now expanded.
1060 JH/03 The smtp transport now requests PRDR by default, if the server offers
1063 JH/04 Certificate name checking on server certificates, when exim is a client,
1064 is now done by default. The transport option tls_verify_cert_hostnames
1065 can be used to disable this per-host. The build option
1066 EXPERIMENTAL_CERTNAMES is withdrawn.
1068 JH/05 The value of the tls_verify_certificates smtp transport and main options
1069 default to the word "system" to access the system default CA bundle.
1070 For GnuTLS, only version 3.0.20 or later.
1072 JH/06 Verification of the server certificate for a TLS connection is now tried
1073 (but not required) by default. The verification status is now logged by
1074 default, for both outbound TLS and client-certificate supplying inbound
1077 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1080 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1081 Status Notification (bounce) messages are now MIME format per RFC 3464.
1082 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1083 under the control of the dsn_advertise_hosts option, and routers may
1084 have a dsn_lasthop option.
1086 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1087 default, modifiable by a malware= option. The list separator for
1088 the options can now be changed in the usual way. Bug 68.
1090 JH/10 The smtp_receive_timeout main option is now expanded before use.
1092 JH/11 The incoming_interface log option now also enables logging of the
1093 local interface on delivery outgoing connections.
1095 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1096 if the interface and destination host and port all match.
1098 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1101 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1102 Patch from Andrew Lewis.
1104 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1105 now supports optional time-restrictions, weighting, and priority
1106 modifiers per server. Patch originally by <rommer@active.by>.
1108 JH/16 The spamd_address main option now supports a mixed list of local
1109 and remote servers. Remote servers can be IPv6 addresses, and
1110 specify a port-range.
1112 JH/17 Bug 68: The spamd_address main option now supports an optional
1113 timeout value per server.
1115 JH/18 Bug 1581: Router and transport options headers_add/remove can
1116 now have the list separator specified.
1118 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1121 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1124 JH/21 Support for the A6 type of dns record is withdrawn.
1126 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1127 rather than the verbs used.
1129 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1130 from 255 to 1024 chars.
1132 JH/24 Verification callouts now attempt to use TLS by default.
1134 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1135 are generic router options now. The defaults didn't change.
1137 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1138 Original patch from Alexander Shikoff, worked over by JH.
1140 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1141 files if ZCAT_COMMAND is not executable.
1143 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1145 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1147 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1148 Normally benign, it bites when the pair was led to by a CNAME;
1149 modern usage is to not canonicalize the domain to a CNAME target
1150 (and we were inconsistent anyway for A-only vs AAAA+A).
1152 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1154 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1155 when evaluating $sender_host_dnssec.
1157 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1158 $sender_helo_dnssec variable.
1160 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1162 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1164 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1166 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1167 documented as working, but never had. Support all but $spam_report.
1169 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1170 added for tls authenticator.
1172 HS/03 Add perl_taintmode main config option
1178 TL/01 When running the test suite, the README says that variables such as
1179 no_msglog_check are global and can be placed anywhere in a specific
1180 test's script, however it was observed that placement needed to be near
1181 the beginning for it to behave that way. Changed the runtest perl
1182 script to read through the entire script once to detect and set these
1183 variables, reset to the beginning of the script, and then run through
1184 the script parsing/test process like normal.
1186 TL/02 The BSD's have an arc4random API. One of the functions to induce
1187 adding randomness was arc4random_stir(), but it has been removed in
1188 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1189 function when detected.
1191 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1192 cause callback expansion.
1194 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1195 syntax errors in an expansion can be treated as a string instead of
1196 logging or causing an error, due to the internal use of bool_lax
1197 instead of bool when processing it.
1199 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1200 server certificates when making smtp deliveries.
1202 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1204 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1206 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1208 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1209 Merged patch from Sebastian Wiedenroth.
1211 JH/05 Fix results-pipe from transport process. Several recipients, combined
1212 with certificate use, exposed issues where response data items split
1213 over buffer boundaries were not parsed properly. This eventually
1214 resulted in duplicates being sent. This issue only became common enough
1215 to notice due to the introduction of connection certificate information,
1216 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1218 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1219 size buffer was used, resulting in syntax errors when an expansion
1222 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1223 version 3.3.6 or later.
1225 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1226 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1227 both become "event_action", the variables become $event_name, $event_data
1228 and $event_defer_errno. There is a new variable $verify_mode, usable in
1229 routers, transports and related events. The tls:cert event is now also
1230 raised for inbound connections, if the main configuration event_action
1233 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1234 early OCSP support, but no stapling (appears to be less than 1.0.0).
1236 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1237 server certificate names available under the smtp transport option
1238 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1241 JH/10 Time-related extraction expansions from certificates now use the main
1242 option "timezone" setting for output formatting, and are consistent
1243 between OpenSSL and GnuTLS compilations. Bug 1541.
1245 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1246 encoded parameter in the incoming message. Bug 1558.
1248 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1249 include certificate info, eximon was claiming there were spoolfile
1252 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1254 JH/14 Log delivery-related information more consistently, using the sequence
1255 "H=<name> [<ip>]" wherever possible.
1257 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1258 are problematic for Debian distribution, omit them from the release
1261 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1263 JH/16 Fix string representation of time values on 64bit time_t architectures.
1266 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1267 output list separator was used.
1272 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1273 checkers that were complaining about end of non-void function with no
1276 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1277 This was a regression introduced in 4.83 by another bugfix.
1279 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1281 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1282 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1288 TF/01 Correctly close the server side of TLS when forking for delivery.
1290 When a message was received over SMTP with TLS, Exim failed to clear up
1291 the incoming connection properly after forking off the child process to
1292 deliver the message. In some situations the subsequent outgoing
1293 delivery connection happened to have the same fd number as the incoming
1294 connection previously had. Exim would try to use TLS and fail, logging
1295 a "Bad file descriptor" error.
1297 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1298 utilities have not been installed.
1300 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1301 temporary space as the ACL may create new global variables.
1303 TL/01 LDAP support uses per connection or global context settings, depending
1304 upon the detected version of the libraries at build time.
1306 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1307 to extract and use the src ip:port in logging and expansions as if it
1308 were a direct connection from the outside internet. PPv2 support was
1309 updated based on HAProxy spec change in May 2014.
1311 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1313 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1314 Properly escape header and check for NULL return.
1316 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1319 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1321 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1322 characters in header names, implemented as a verify condition.
1323 Contributed by Michael Fischer v. Mollard.
1325 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1326 results permerror and temperror. Previous values are deprecated but
1327 still accepted. In a future release, err_perm and err_temp will be
1328 completely removed, which will be a backward incompatibility if the
1329 ACL tests for either of these two old results. Patch contributed by
1330 user bes-internal on the mailing list.
1332 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1334 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1335 selectors, in both main and reject logs.
1337 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1340 JH/07 Add malware type "sock" for talking to simple daemon.
1342 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1344 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1345 routers/transports under cutthrough routing.
1347 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1348 numbers. Touch up "bool" conditional to keep the same definition.
1350 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1352 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1354 JH/12 Expand items in router/transport headers_add or headers_remove lists
1355 individually rather than the list as a whole. Bug 1452.
1357 Required for reasonable handling of multiple headers_ options when
1358 they may be empty; requires that headers_remove items with embedded
1359 colons must have them doubled (or the list-separator changed).
1361 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1362 view the policy declared in the DMARC record. Currently, $dmarc_status
1363 is a combined value of both the record presence and the result of the
1366 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1368 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1369 dnslookup router and the smtp transport (applying to the forward
1372 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1373 of ldap servers used for a specific lookup. Patch provided by Heiko
1376 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1377 New variable $lookup_dnssec_authenticated for observability.
1379 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1380 Patch submitted by Lars Timman.
1382 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1384 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1385 Requires trusted mode and valid format message id, aborts otherwise.
1386 Patch contributed by Heiko Schlichting.
1388 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1389 certextract with support for various fields. Bug 1358.
1391 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1392 is requested by default, modifiable by smtp transport option
1395 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1396 operate on certificate variables to give certificate fingerprints
1397 Also new ${sha256:cert_variable}.
1399 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1401 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1402 Christian Aistleitner.
1404 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1406 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1407 file. Patch from Wolfgang Breyha.
1409 JH/25 Expand the coverage of the delivery $host and $host_address to
1410 client authenticators run in verify callout. Bug 1476.
1412 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1413 align with daemon_smtp_ports. Bug 72.
1415 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1416 support and error reporting did not work properly.
1418 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1419 and is readable. Patch from Andrew Colin Kissa.
1421 TL/14 Enhance documentation of ${run expansion and how it parses the
1422 commandline after expansion, particularly in the case when an
1423 unquoted variable expansion results in an empty value.
1425 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1427 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1428 by renaming away from it. Addresses GH issue 15, reported by
1431 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1433 TL/15 SECURITY: prevent double expansion in math comparison functions
1434 (can expand unsanitized data). Not remotely exploitable.
1441 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1443 PP/02 Make -n do something, by making it not do something.
1444 When combined with -bP, the name of an option is not output.
1446 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1449 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1450 $sender_host_name and config options to manage this, and basic check
1453 PP/05 DSCP support for outbound connections and control modifier for inbound.
1455 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1456 (Only plugin which currently uses this is kerberos4, which nobody should
1457 be using, but we should make it available and other future plugins might
1458 conceivably use it, even though it would break NAT; stuff *should* be
1459 using channel bindings instead).
1461 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1462 name; added for Sendmail compatibility; requires admin caller.
1463 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1464 just ignore it); requires trusted caller.
1465 Also parse but ignore: -Ac -Am -X<logfile>
1468 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1470 TL/02 Add +smtp_confirmation as a default logging option.
1472 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1473 Patch by Magnus Holmgren from 2007-02-20.
1475 TL/04 Bugzilla 1281 - Spec typo.
1476 Bugzilla 1283 - Spec typo.
1477 Bugzilla 1290 - Spec grammar fixes.
1479 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1481 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1483 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1484 mailing list by by Dmitry Isaikin.
1486 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1488 JH/02 Support "G" suffix to numbers in ${if comparisons.
1490 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1492 NM/01 Bugzilla 1197 - Spec typo
1493 Bugzilla 1196 - Spec examples corrections
1495 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1497 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1498 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1501 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1502 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1504 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1505 "acl {{name}{arg}...}", and optional args on acl condition
1508 JH/05 Permit multiple router/transport headers_add/remove lines.
1510 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1512 JH/07 Avoid using a waiting database for a single-message-only transport.
1513 Performance patch from Paul Fisher. Bugzilla 1262.
1515 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1518 JH/09 Add $headers_added variable, with content from use of ACL modifier
1519 add_header (but not yet added to the message). Bugzilla 199.
1521 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1522 Pulled from Bugzilla 817 by Wolfgang Breyha.
1524 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1526 (nb: this is the same fix as in Exim 4.80.1)
1528 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1531 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1534 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1536 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1537 advertise SMTP AUTH mechanism to us, instead of a generic
1538 protocol violation error. Also, make Exim more robust to bad
1539 data from the Dovecot auth socket.
1541 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1543 When a queue runner is handling a message, Exim first routes the
1544 recipient addresses, during which it prunes them based on the retry
1545 hints database. After that it attempts to deliver the message to
1546 any remaining recipients. It then updates the hints database using
1549 So if a recipient address works intermittently, it can get repeatedly
1550 deferred at routing time. The retry hints record remains fresh so the
1551 address never reaches the final cutoff time.
1553 This is a fairly common occurrence when a user is bumping up against
1554 their storage quota. Exim had some logic in its local delivery code
1555 to deal with this. However it did not apply to per-recipient defers
1556 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1558 This change adds a proper retry rule check during routing so that the
1559 final cutoff time is checked against the message's age. We only do
1560 this check if there is an address retry record and there is not a
1561 domain retry record; this implies that previous attempts to handle
1562 the address had the retry_use_local_parts option turned on. We use
1563 this as an approximation for the destination being like a local
1564 delivery, as in LMTP.
1566 I suspect this new check makes the old local delivery cutoff check
1567 redundant, but I have not verified this so I left the code in place.
1569 TF/02 Correct gecos expansion when From: is a prefix of the username.
1571 Test 0254 submits a message to Exim with the header
1575 When I ran the test suite under the user fanf2, Exim expanded
1576 the header to contain my full name, whereas it should have added
1577 a Resent-Sender: header. It erroneously treats any prefix of the
1578 username as equal to the username.
1580 This change corrects that bug.
1582 GF/01 DCC debug and logging tidyup
1583 Error conditions log to paniclog rather than rejectlog.
1584 Debug lines prefixed by "DCC: " to remove any ambiguity.
1586 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1588 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1589 Bug spotted by Jeremy Harris; was flawed since initial commit.
1590 Would have resulted in OCSP responses post-SNI triggering an Exim
1591 NULL dereference and crash.
1593 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1595 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1596 Bug detection, analysis and fix by Samuel Thibault.
1597 Bugzilla 1331, Debian bug #698092.
1599 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1601 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1602 Server implementation by Todd Lyons, client by JH.
1603 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1604 config variable "prdr_enable" controls whether the server
1605 advertises the facility. If the client requests PRDR a new
1606 acl_data_smtp_prdr ACL is called once for each recipient, after
1607 the body content is received and before the acl_smtp_data ACL.
1608 The client is controlled by both of: a hosts_try_prdr option
1609 on the smtp transport, and the server advertisement.
1610 Default client logging of deliveries and rejections involving
1611 PRDR are flagged with the string "PRDR".
1613 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1614 fclose(). Diagnosis by Todd Lyons.
1616 PP/17 Update configure.default to handle IPv6 localhost better.
1617 Patch by Alain Williams (plus minor tweaks).
1620 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1621 This is now consistent with GnuTLS, and is now documented: the
1622 previous undocumented portable approach to treating the option as
1623 unset was to force an expansion failure. That still works, and
1624 an empty string is now equivalent.
1626 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1627 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1628 not performing validation itself.
1630 PP/20 Added force_command boolean option to pipe transport.
1631 Patch from Nick Koston, of cPanel Inc.
1633 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1636 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1638 PP/21 Fix eximon continuous updating with timestamped log-files.
1639 Broken in a format-string cleanup in 4.80, missed when I repaired the
1640 other false fix of the same issue.
1641 Report and fix from Heiko Schlichting.
1644 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1645 Report from Prashanth Katuri.
1647 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1648 It's SecureTransport, so affects any MacOS clients which use the
1649 system-integrated TLS libraries, including email clients.
1651 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1652 using a MIME ACL for non-SMTP local injection.
1653 Report and assistance in diagnosis by Warren Baker.
1655 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1657 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1659 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1660 last id that failed so it may be referenced in subsequent ACL's.
1662 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1665 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1666 ldap library initialization, allowing self-signed CA's to be
1667 used. Also properly sets require_cert option later in code by
1668 using NULL (global ldap config) instead of ldap handle (per
1669 session). Bug diagnosis and testing by alxgomz.
1671 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1672 the src/util/ subdirectory.
1674 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1675 renamed to Transport Post Delivery Action by Jeremy Harris, as
1678 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1679 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1680 redis_servers = needs to be configured which will be used by the redis
1681 lookup. Patch from Warren Baker, of The Packet Hub.
1683 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1685 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1686 hostname or reverse DNS when processing a host list. Used suggestions
1687 from multiple comments on this bug.
1689 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1691 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1692 Missed a few lines, added it to make the runtest require no keyboard
1695 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1696 contains upper case chars. Make router use caseful_local_part.
1698 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1699 support when GnuTLS has been built with p11-kit.
1705 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1707 This, or similar/improved, will also be change PP/11 of 4.82.
1713 PP/01 Handle short writes when writing local log-files.
1714 In practice, only affects FreeBSD (8 onwards).
1715 Bugzilla 1053, with thanks to Dmitry Isaikin.
1717 NM/01 Bugzilla 949 - Documentation tweak
1719 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1722 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1724 PP/02 Implemented gsasl authenticator.
1726 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1728 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1729 `pkg-config foo` for cflags/libs.
1731 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1732 with rest of GSASL and with heimdal_gssapi.
1734 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1735 `pkg-config foo` for cflags/libs for the TLS implementation.
1737 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1738 properties get this fed in as external SSF. A number of robustness
1739 and debugging improvements to the cyrus_sasl authenticator.
1741 PP/08 cyrus_sasl server now expands the server_realm option.
1743 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1744 Patch by Jeremy Harris.
1746 PP/10 Added dbmjz lookup type.
1748 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1750 PP/12 MAIL args handles TAB as well as SP, for better interop with
1751 non-compliant senders.
1752 Analysis and variant patch by Todd Lyons.
1754 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1755 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1756 Patch from Dirk Mueller <dmueller@suse.com>
1758 PP/13 tls_peerdn now print-escaped for spool files.
1759 Observed some $tls_peerdn in wild which contained \n, which resulted
1760 in spool file corruption.
1762 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1763 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1764 or write after TLS renegotiation, which otherwise led to messages
1767 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1768 as a tracking header (ie: a signed header comes before the signature).
1769 Patch from Wolfgang Breyha.
1771 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1772 comma-sep list; embedded commas doubled.
1774 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1776 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1778 Report and patch from Dmitry Banschikov.
1780 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1781 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1782 failures. We appear to now support TLS1.1+ with Exim.
1784 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1785 lets Exim select keys and certificates based upon TLS SNI from client.
1786 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1787 before an outbound SMTP session. New log_selector, +tls_sni.
1789 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1790 NULL dereference. Report and patch from Alun Jones.
1792 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1793 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1794 Not seeing resolver debug output on NetBSD, but suspect this is a
1795 resolver implementation change.
1797 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1798 Left warnings. Added "eximon gdb" invocation mode.
1800 PP/21 Defaulting "accept_8bitmime" to true, not false.
1802 PP/22 Added -bw for inetd wait mode support.
1804 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1805 locate the relevant includes and libraries. Made this the default.
1807 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1808 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1810 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1811 This may cause build issues on older platforms.
1813 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1814 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1815 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1816 Added SNI support via GnuTLS too.
1817 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1819 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1821 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1822 Applied second patch from Janne, implementing suggestion to default
1823 multiple-strings-in-record handling to match SPF spec.
1825 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1827 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1828 read-only, out of scope).
1829 Patch from Wolfgang Breyha, report from Stuart Northfield.
1831 PP/29 Fix three issues highlighted by clang analyser static analysis.
1832 Only crash-plausible issue would require the Cambridge-specific
1833 iplookup router and a misconfiguration.
1834 Report from Marcin Mirosław.
1836 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1838 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1839 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1840 As part of this, removing so much warning spew let me fix some minor
1841 real issues in debug logging.
1843 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1844 assignment on my part. Fixed.
1846 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1847 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1848 Janne Snabb (who went above and beyond: thank you).
1850 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1851 string otherwise requires a connection and a bunch more work and it's
1852 relatively easy to get wrong. Should also expose TLS library linkage
1855 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1856 64-bit ${eval} (JH/03).
1858 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1859 GNU libc to support some of the 64-bit stuff, should not lead to
1860 conflicts. Defined before os.h is pulled in, so if a given platform
1861 needs to override this, it can.
1863 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1864 protection layer was required, which is not implemented.
1865 Bugzilla 1254, patch from Wolfgang Breyha.
1867 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1868 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1869 tls_dhparam take prime identifiers. Also unbreak combination of
1870 OpenSSL+DH_params+TLSSNI.
1872 PP/39 Disable SSLv2 by default in OpenSSL support.
1878 PP/01 Solaris build fix for Oracle's LDAP libraries.
1879 Bugzilla 1109, patch from Stephen Usher.
1881 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1883 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1886 TF/02 Fix a couple more cases where we did not log the error message
1887 when unlink() failed. See also change 4.74-TF/03.
1889 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1890 lock up or crash if it happened to be inside a call to libc when it
1891 got a SIGUSR1 from exiwhat.
1893 The SIGUSR1 handler appends the current process status to the process
1894 log which is later printed by exiwhat. It used to use the general
1895 purpose logging code to do this, but several functions it calls are
1896 not safe for signals.
1898 The new output code in the SIGUSR1 handler is specific to the process
1899 log, and simple enough that it's easy to inspect for signal safety.
1900 Removing some special cases also simplifies the general logging code.
1901 Removing the spurious timestamps from the process log simplifies
1904 TF/04 Improved ratelimit ACL condition.
1906 The /noupdate option has been deprecated in favour of /readonly which
1907 has clearer semantics. The /leaky, /strict, and /readonly update modes
1908 are mutually exclusive. The update mode is no longer included in the
1909 database key; it just determines when the database is updated. (This
1910 means that when you upgrade Exim will forget old rate measurements.)
1912 Exim now checks that the per_* options are used with an update mode that
1913 makes sense for the current ACL. For example, when Exim is processing a
1914 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1915 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1916 must specify per_mail/readonly. If you omit the update mode it defaults to
1917 /leaky where that makes sense (as before) or /readonly where required.
1919 The /noupdate option is now undocumented but still supported for
1920 backwards compatibility. It is equivalent to /readonly except that in
1921 ACLs where /readonly is required you may specify /leaky/noupdate or
1922 /strict/noupdate which are treated the same as /readonly.
1924 A useful new feature is the /count= option. This is a generalization
1925 of the per_byte option, so that you can measure the throughput of other
1926 aggregate values. For example, the per_byte option is now equivalent
1927 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1929 The per_rcpt option has been generalized using the /count= mechanism
1930 (though it's more complicated than the per_byte equivalence). When it is
1931 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1932 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1933 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1934 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1935 Note that using per_rcpt with a non-readonly update mode in more than
1936 one ACL will cause the recipients to be double-counted. (The per_mail
1937 and per_byte options don't have this problem.)
1939 The handling of very low rates has changed slightly. If the computed rate
1940 is less than the event's count (usually one) then this event is the first
1941 after a long gap. In this case the rate is set to the same as this event's
1942 count, so that the first message of a spam run is counted properly.
1944 The major new feature is a mechanism for counting the rate of unique
1945 events. The new per_addr option counts the number of different
1946 recipients that someone has sent messages to in the last time period. It
1947 behaves like per_rcpt if all the recipient addresses are different, but
1948 duplicate recipient addresses do not increase the measured rate. Like
1949 the /count= option this is a general mechanism, so the per_addr option
1950 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1951 example, measure the rate that a client uses different sender addresses
1952 with the options per_mail/unique=$sender_address. There are further
1953 details in the main documentation.
1955 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1957 TF/06 Removed a few PCRE remnants.
1959 TF/07 Automatically extract Exim's version number from tags in the git
1960 repository when doing development or release builds.
1962 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1963 Bugzilla 879. Patch from Paul Fisher.
1965 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1966 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1969 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1971 PP/05 Variable $av_failed, true if the AV scanner deferred.
1972 Bugzilla 1078. Patch from John Horne.
1974 PP/06 Stop make process more reliably on build failure.
1975 Bugzilla 1087. Patch from Heiko Schlittermann.
1977 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1978 Bugzilla 1089. Patch from Heiko Schlittermann.
1980 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1981 Bugzilla 1131. Patch from Holger Weiß.
1983 PP/09 Handle IPv6 addresses with SPF.
1984 Bugzilla 860. Patch from Wolfgang Breyha.
1986 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1988 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1991 PP/11 match_* no longer expand right-hand-side by default.
1992 New compile-time build option, EXPAND_LISTMATCH_RHS.
1993 New expansion conditions, "inlist", "inlisti".
1995 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1997 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1999 PP/14 fix log_write() format string regression from TF/03.
2000 Bugzilla 1152. Patch from Dmitry Isaikin.
2006 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2008 PP/02 Harmonised TLS library version reporting; only show if debugging.
2009 Layout now matches that introduced for other libraries in 4.74 PP/03.
2011 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2013 PP/04 New "dns_use_edns0" global option.
2015 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2018 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2019 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2021 TK/01 Updated PolarSSL code to 0.14.2.
2022 Bugzilla 1097. Patch from Andreas Metzler.
2024 PP/07 Catch divide-by-zero in ${eval:...}.
2025 Fixes bugzilla 1102.
2027 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2030 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2031 format-string attack -- SECURITY: remote arbitrary code execution.
2033 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2034 time unintentionally subject to list matching rules, letting the header
2035 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2036 arbitrary string expansion). This allowed for information disclosure.
2038 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2039 INT_MIN/-1 -- value coerced to INT_MAX.
2045 NM/01 Workaround for PCRE version dependency in version reporting
2048 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2049 This fixes portability to compilers other than gcc, notably
2050 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2052 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2053 makefiles for portability to HP-UX and POSIX correctness.
2055 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2056 Also via indented variable definition in the Makefile.
2057 (Debugging by Oliver Heesakkers).
2059 PP/02 Restore caching of spamd results with expanded spamd_address.
2060 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2062 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2063 Improves build reliability. Fix from: Frank Elsner
2065 NM/02 Fix wide character breakage in the rfc2047 coding
2066 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2068 NM/03 Allow underscore in dnslist lookups
2069 Fixes bug 1026. Patch from Graeme Fowler
2071 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2072 Code patches from Adam Ciarcinski of NetBSD.
2074 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2077 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2078 is logged, to avoid truncation. Patch from John Horne.
2080 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2081 Patch from Jakob Hirsch.
2083 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2084 SQL string expansion failure details.
2085 Patch from Andrey Oktyabrski.
2087 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2088 Patch from Simon Arlott.
2090 PP/09 New lookups functionality failed to compile on old gcc which rejects
2091 extern declarations in function scope.
2092 Patch from Oliver Fleischmann
2094 PP/10 Use sig_atomic_t for flags set from signal handlers.
2095 Check getgroups() return and improve debugging.
2096 Fixed developed for diagnosis in bug 927 (which turned out to be
2099 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2100 Patch from Mark Zealey.
2102 PP/12 Bugzilla 1056: Improved spamd server selection.
2103 Patch from Mark Zealey.
2105 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2106 Based on patch from Heiko Schlittermann.
2108 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2109 Patch from Uwe Doering, sign-off by Michael Haardt.
2111 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2112 variable declaration deep within a block. Bug and patch from
2115 PP/15 lookups-Makefile IRIX compatibility coercion.
2117 PP/16 Make DISABLE_DKIM build knob functional.
2119 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2120 Patch by Simon Arlott
2122 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2123 variable argument macros. Our copy now differs from upstream.
2129 TF/01 Failure to get a lock on a hints database can have serious
2130 consequences so log it to the panic log.
2132 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2133 controlled using the smtp_confirmation log selector.
2135 TF/03 Include the error message when we fail to unlink a spool file.
2137 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2138 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2139 for maintaining out-of-tree patches for some time.
2141 PP/01 Bugzilla 139: Documentation and portability issues.
2142 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2143 Handle per-OS dynamic-module compilation flags.
2145 PP/02 Let /dev/null have normal permissions.
2146 The 4.73 fixes were a little too stringent and complained about the
2147 permissions on /dev/null. Exempt it from some checks.
2148 Reported by Andreas M. Kirchwitz.
2150 PP/03 Report version information for many libraries, including
2151 Exim version information for dynamically loaded libraries. Created
2152 version.h, now support a version extension string for distributors
2153 who patch heavily. Dynamic module ABI change.
2155 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2156 privilege escalation vulnerability whereby the Exim run-time user
2157 can cause root to append content of the attacker's choosing to
2160 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2163 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2164 If dropping privileges for untrusted macros, we disabled normal logging
2165 on the basis that it would fail; for the Exim run-time user, this is not
2166 the case, and it resulted in successful deliveries going unlogged.
2167 Fixed. Reported by Andreas Metzler.
2173 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2174 only prepend for the Resent-* case. Fixes regression introduced in
2175 Exim 4.70 by NM/22 for Bugzilla 607.
2177 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2178 increasing numbers of administrators be bitten by this.
2180 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2182 PP/03 Bugzilla 994: added openssl_options main configuration option.
2184 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2186 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2188 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2190 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2191 without a peer certificate, leading to a segfault because of an
2192 assumption that peers always have certificates. Be a little more
2193 paranoid. Problem reported by Martin Tscholak.
2195 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2196 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2197 NB: ClamAV planning to remove STREAM in "middle of 2010".
2198 CL also introduces -bmalware, various -d+acl logging additions and
2199 more caution in buffer sizes.
2201 PP/09 Implemented reverse_ip expansion operator.
2203 PP/10 Bugzilla 937: provide a "debug" ACL control.
2205 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2207 PP/12 Bugzilla 973: Implement --version.
2209 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2211 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2213 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2215 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2216 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2217 ignore trailing whitespace.
2219 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2221 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2224 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2225 Notification from Dr Andrew Aitchison.
2227 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2228 ExtendedDetectionInfo response format.
2229 Notification from John Horne.
2231 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2234 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2235 XSL and documented dependency on system catalogs, with examples of how
2238 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2241 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2242 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2243 configuration file which is writeable by the Exim user or group.
2245 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2246 of configuration files to cover files specified with the -C option if
2247 they are going to be used with root privileges, not just the default
2250 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2251 option (effectively making it always true).
2253 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2254 files to be used while preserving root privileges.
2256 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2257 that rogue child processes cannot use them.
2259 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2260 run-time user, instead of root.
2262 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2263 Exim run-time user without dropping privileges.
2265 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2266 result string, instead of calling string_vformat() twice with the same
2269 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2270 for other users. Others should always drop root privileges if they use
2271 -C on the command line, even for a whitelisted configure file.
2273 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2275 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2281 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2282 $data_path, and $header_path variables; fixed documentation bugs and
2285 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2286 exipick to access non-standard spools, including the "frozen" queue
2289 NM/01 Bugzilla 965: Support mysql stored procedures.
2290 Patch from Alain Williams
2292 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2294 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2295 Patch from Andreas Metzler
2297 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2298 Patch from Kirill Miazine
2300 NM/05 Bugzilla 671: Added umask to procmail example.
2302 JJ/03 installed exipick 20100323.0, fixing doc bug
2304 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2305 directory. Notification and patch from Dan Rosenberg.
2307 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2309 TK/02 Improve log output when DKIM signing operation fails.
2311 MH/01 Treat the transport option dkim_domain as a colon separated
2312 list, not as a single string, and sign the message with each element,
2313 omitting multiple occurences of the same signer.
2315 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2316 Bugzilla 985, 986. Patch by Simon Arlott
2318 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2319 Patch by Simon Arlott
2321 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2322 MBX locking. Notification from Dan Rosenberg.
2328 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2330 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2332 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2334 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2336 NM/04 Bugzilla 915: Flag broken perl installation during build.
2342 TK/01 Added patch by Johannes Berg that expands the main option
2343 "spamd_address" if it starts with a dollar sign.
2345 TK/02 Write list of recipients to X-Envelope-Sender header when building
2346 the mbox-format spool file for content scanning (suggested by Jakob
2349 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2350 (http://www.dcc-servers.net/) support via dccifd. Activated by
2351 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2353 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2354 by Mark Daniel Reidel <mr@df.eu>.
2356 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2357 When building exim an external PCRE library is now needed -
2358 PCRE is a system library on the majority of modern systems.
2359 See entry on PCRE_LIBS in EDITME file.
2361 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2362 conversation. Added nologin parameter to request.
2363 Patch contributed by Kirill Miazine.
2365 TF/01 Do not log submission mode rewrites if they do not change the address.
2367 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2369 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2370 log files in place. Contributed by Roberto Lima.
2372 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2374 TF/03 Bugzilla 615: When checking the local_parts router precondition
2375 after a local_part_suffix or local_part_prefix option, Exim now
2376 does not use the address's named list lookup cache, since this
2377 contains cached lookups for the whole local part.
2379 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2380 Robert Millan. Documentation is in experimental-spec.txt.
2382 TF/04 Bugzilla 668: Fix parallel build (make -j).
2384 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2386 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2387 Patch provided by Jan Srzednicki.
2389 TF/05 Leading white space used to be stripped from $spam_report which
2390 wrecked the formatting. Now it is preserved.
2392 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2393 that they are available at delivery time.
2395 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2397 TF/08 TLS error reporting now respects the incoming_interface and
2398 incoming_port log selectors.
2400 TF/09 Produce a more useful error message if an SMTP transport's hosts
2401 setting expands to an empty string.
2403 NM/06 Bugzilla 744: EXPN did not work under TLS.
2404 Patch provided by Phil Pennock.
2406 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2407 Patch provided by Richard Godbee.
2409 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2410 acl_smtp_notquit, added index entry.
2412 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2413 Patch provided by Eugene Bujak.
2415 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2416 accept(). Patch provided by Maxim Dounin.
2418 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2419 Patch provided by Phil Pennock.
2421 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2423 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2424 Patch provided by Brad "anomie" Jorsch.
2426 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2427 Patch provided by Dean Brooks.
2429 TK/05 Add native DKIM support (does not depend on external libraries).
2431 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2432 Patch provided by Graeme Fowler.
2434 NM/16 Bugzilla 851: Documentation example syntax fix.
2436 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2438 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2441 NM/19 Bugzilla 745: TLS version reporting.
2442 Patch provided by Phil Pennock.
2444 NM/20 Bugzilla 167: bool: condition support.
2445 Patch provided by Phil Pennock.
2447 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2448 clients. Patch provided by Phil Pennock.
2450 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2451 Patch provided by Brad "anomie" Jorsch.
2453 NM/23 Bugzilla 687: Fix misparses in eximstats.
2454 Patch provided by Heiko Schlittermann.
2456 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2457 Patch provided by Heiko Schlittermann.
2459 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2460 plus update to original patch.
2462 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2464 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2465 Patch provided by David Brownlee.
2467 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2469 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2471 NM/30 Bugzilla 888: TLS documentation bugfixes.
2473 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2475 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2476 Unlike the original bugzilla I have changed all shell scripts in src tree.
2478 NM/33 Bugzilla 898: Transport filter timeout fix.
2479 Patch by Todd Rinaldo.
2481 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2482 Patch by Serge Demonchaux.
2484 NM/35 Bugzilla 39: Base64 decode bug fixes.
2485 Patch by Jakob Hirsch.
2487 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2489 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2491 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2493 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2499 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2500 ALT-N's libdkim that I have put here:
2501 http://duncanthrax.net/exim-experimental/
2503 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2504 were called 'true' and it seems that C99 defines that as a reserved
2505 keyword to be used with 'bool' variable types. That means you could
2506 not include C99-style headers which use bools without triggering
2507 build errors in sieve.c.
2509 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2510 as mailq or other aliases. Changed the --help handling significantly
2511 to do whats expected. exim_usage() emits usage/help information.
2513 SC/01 Added the -bylocaldomain option to eximstats.
2515 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2517 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2519 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2525 PH/01 Another patch from the Sieve maintainer.
2527 PH/02 When an IPv6 address is converted to a string for single-key lookup
2528 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2529 dots are used instead of colons so that keys in lsearch files need not
2530 contain colons. This was done some time before quoting was made available
2531 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2532 (notated using the quote facility) so as to distinguish them from IPv4
2533 keys. This meant that lookups for IP addresses in host lists did not work
2534 for iplsearch lookups.
2536 This has been fixed by arranging for IPv6 addresses to be expressed with
2537 colons if the lookup type is iplsearch. This is not incompatible, because
2538 previously such lookups could never work.
2540 The situation is now rather anomalous, since one *can* have colons in
2541 ordinary lsearch keys. However, making the change in all cases is
2542 incompatible and would probably break a number of configurations.
2544 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2547 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2548 conversion specification without a maximum field width, thereby enabling
2549 a rogue spamd server to cause a buffer overflow. While nobody in their
2550 right mind would setup Exim to query an untrusted spamd server, an
2551 attacker that gains access to a server running spamd could potentially
2552 exploit this vulnerability to run arbitrary code as the Exim user.
2554 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2555 $primary_hostname instead of what libspf2 thinks the hosts name is.
2557 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2558 a directory entry by the name of the lookup key. Previously, if a
2559 symlink pointed to a non-existing file or a file in a directory that
2560 Exim lacked permissions to read, a lookup for a key matching that
2561 symlink would fail. Now it is enough that a matching directory entry
2562 exists, symlink or not. (Bugzilla 503.)
2564 PH/03 The body_linecount and body_zerocount variables are now exported in the
2567 PH/04 Added the $dnslist_matched variable.
2569 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2570 This means they are set thereafter only if the connection becomes
2573 PH/06 Added the client_condition to authenticators so that some can be skipped
2574 by clients under certain conditions.
2576 PH/07 The error message for a badly-placed control=no_multiline_responses left
2577 "_responses" off the end of the name.
2579 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2581 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2582 (without spaces) instead of just copying the configuration text.
2584 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2586 PH/11 Added $max_received_linelength.
2588 PH/12 Added +ignore_defer and +include_defer to host lists.
2590 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2591 way in which PCRE > 7.0 is built.
2593 PH/14 Implemented queue_only_load_latch.
2595 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2596 MAIL command. The effect was to mangle the value on 64-bit systems.
2598 PH/16 Another patch from the Sieve maintainer.
2600 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2602 PH/18 If a system quota error occurred while trying to create the file for
2603 a maildir delivery, the message "Mailbox is full" was not appended to the
2604 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2605 only to a quota excession during the actual writing of the file.
2607 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2608 characters?) which causes problems in log lines. The DN values are now
2609 passed through string_printing() before being added to log lines.
2611 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2612 and InterBase are left for another time.)
2614 PH/21 Added message_body_newlines option.
2616 PH/22 Guard against possible overflow in moan_check_errorcopy().
2618 PH/23 POSIX allows open() to be a macro; guard against that.
2620 PH/24 If the recipient of an error message contained an @ in the local part
2621 (suitably quoted, of course), incorrect values were put in $domain and
2622 $local_part during the evaluation of errors_copy.
2628 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2629 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2632 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2633 issue a MAIL command.
2635 PH/02 In an ACL statement such as
2637 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2639 if a client was not listed at all, or was listed with a value other than
2640 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2641 the condition was not true (as it should be), so access was not denied.
2642 The bug was that the ! inversion was incorrectly passed on to the second
2643 item. This has been fixed.
2645 PH/03 Added additional dnslists conditions == and =& which are different from
2646 = and & when the dns lookup returns more than one IP address.
2648 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2649 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2651 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2652 FSYNC, which compiles an option called disable_fsync that allows for
2653 bypassing fsync(). The documentation is heavily laced with warnings.
2655 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2657 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2658 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2659 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2660 including adding "make clean"; (3) Added -fPIC when compiling the test
2661 dynamically loaded module, to get rid of a warning.
2663 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2664 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2665 = 0s. The bug is otherwise harmless.
2667 PH/07 There was a bug in the dovecot authenticator such that the value of
2668 $auth1 could be overwritten, and so not correctly preserved, after a
2669 successful authentication. This usually meant that the value preserved by
2670 the server_setid option was incorrect.
2672 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2674 PH/09 Installed PCRE release 7.0.
2676 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2677 run for batched SMTP input. It is now run at the start of every message
2678 in the batch. While fixing this I discovered that the process information
2679 (output by running exiwhat) was not always getting set for -bs and -bS
2680 input. This is fixed, and it now also says "batched" for BSMTP.
2682 PH/11 Added control=no_pipelining.
2684 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2685 patch, slightly modified), and move the expansion of helo_data till after
2686 the connection is made in the smtp transport (so it can use these
2689 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2691 PH/14 Added log_selector = +pid.
2693 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2695 PH/16 Add ${if forany and ${if forall.
2697 PH/17 Added dsn_from option to vary the From: line in DSNs.
2699 PH/18 Flush SMTP output before performing a callout, unless control =
2700 no_callout_flush is set.
2702 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2703 was true (the default) a successful delivery failed to delete the retry
2704 item, thus causing premature timeout of the address. The bug is now
2707 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2709 PH/21 Long custom messages for fakedefer and fakereject are now split up
2710 into multiline responses in the same way that messages for "deny" and
2711 other ACL rejections are.
2713 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2714 with slight modification.
2716 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2717 draft, changing the syntax and factoring some duplicate code".
2719 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2720 for deliveries of the second and subsequent messages over the same SMTP
2723 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2724 ${reduce, with only minor "tidies".
2726 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2728 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2729 expansion side effects.
2731 PH/27 When a message times out after an over-quota error from an Exim-imposed
2732 quota, the bounce message says "mailbox is full". This message was not
2733 being given when it was a system quota that was exceeded. It now should
2736 MH/03 Made $recipients available in local_scan(). local_scan() already has
2737 better access to the recipient list through recipients_list[], but
2738 $recipients can be useful in postmaster-provided expansion strings.
2740 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2741 in the case of a MAIL command with additional options following the
2742 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2743 were accidentally chopped off.
2745 PH/29 SMTP synchronization checks are implemented when a command is read -
2746 there is a check that no more input is waiting when there shouldn't be
2747 any. However, for some commands, a delay in an ACL can mean that it is
2748 some time before the response is written. In this time, more input might
2749 arrive, invalidly. So now there are extra checks after an ACL has run for
2750 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2751 pipelining has not been advertised.
2753 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2755 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2756 returned in the SMTP response when smtp_return_error_details was false.
2757 This has been fixed.
2759 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2760 instead of the C I/O that was originally supplied, because problems were
2761 reported on Solaris.
2763 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2764 Exim which did not show up earlier: it was assuming that a call to
2765 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2766 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2767 was a macro that became an assignment, so it seemed to work. This has
2768 changed to a proper function call with a void return, hence the compile
2769 error. Exim's code has been fixed.
2771 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2774 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2776 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2778 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2779 because it turns out to be too big or there is a timeout) there is no
2780 "Completed" line in the log. When some messages of this type were
2781 selected by exigrep, they were listed as "not completed". Others were
2782 picked up by some special patterns. I have improved the selection
2783 criteria to be more general.
2785 PH/38 The host_find_failed option in the manualroute router can now be set
2786 to "ignore", to completely ignore a host whose IP address cannot be
2787 found. If all hosts are ignored, the behaviour is controlled by the new
2788 host_all_ignored option.
2790 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2791 homing or because of multiple MX records with /mx) generated more than
2792 one IP address, and the following item turned out to be the local host,
2793 all the secondary addresses of the first item were incorrectly removed
2794 from the list, along with the local host and any following hosts (which
2795 is what is supposed to happen).
2797 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2798 whoever called Exim into the -H file. In the case of the daemon it was
2799 behaving confusingly. When first started, it used values for whoever
2800 started the daemon, but after a SIGHUP it used the Exim user (because it
2801 calls itself on a restart). I have changed the code so that it now always
2804 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2805 message are rejected with the same error (e.g. no authentication or bad
2806 sender address), and a DATA command is nevertheless sent (as can happen
2807 with PIPELINING or a stupid MUA), the error message that was given to the
2808 RCPT commands is included in the rejection of the DATA command. This is
2809 intended to be helpful for MUAs that show only the final error to their
2812 PH/42 Another patch from the Sieve maintainer.
2814 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2815 Eximstats - Fixed some broken HTML links and added missing column headers
2817 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2818 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2820 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2822 PH/43 Yet another patch from the Sieve maintainer.
2824 PH/44 I found a way to check for a TCP/IP connection going away before sending
2825 the response to the final '.' that terminates a message, but only in the
2826 case where the client has not sent further data following the '.'
2827 (unfortunately, this is allowed). However, in many cases there won't be
2828 any further data because there won't be any more messages to send. A call
2829 to select() can be used: if it shows that the input is "ready", there is
2830 either input waiting, or the socket has been closed. An attempt to read
2831 the next input character can distinguish the two cases. Previously, Exim
2832 would have sent an OK response which the client would never have see.
2833 This could lead to message repetition. This fix should cure that, at
2834 least in a lot of common cases.
2836 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2837 advertised in response to EHLO.
2843 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2844 fixed by 4.65/MH/01 (is this a record?) are fixed:
2846 (i) An empty string was always treated as zero by the numeric comparison
2847 operators. This behaviour has been restored.
2849 (ii) It is documented that the numeric comparison operators always treat
2850 their arguments as decimal numbers. This was broken in that numbers
2851 starting with 0 were being interpreted as octal.
2853 While fixing these problems I realized that there was another issue that
2854 hadn't been noticed. Values of message_size_limit (both the global option
2855 and the transport option) were treated as octal if they started with 0.
2856 The documentation was vague. These values are now always treated as
2857 decimal, and I will make that clear in the documentation.
2863 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2864 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2867 MH/01 Don't check that the operands of numeric comparison operators are
2868 integers when their expansion is in "skipping" mode (fixes bug
2869 introduced by 4.64-PH/07).
2871 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2872 child addresses, Exim now panics and dies. Previously, because the count
2873 is held in a short int, deliveries were likely to be lost. As such a
2874 large number of recipients for a single message is ridiculous
2875 (performance will be very, very poor), I have chosen to impose a limit
2876 rather than extend the field.
2882 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2883 leftover -K file (the existence of which was triggered by #402).
2884 While we were at it, introduced process PID as part of the -K
2885 filename. This should rule out race conditions when creating
2888 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2889 processing considerably. Previous code took too long for large mails,
2890 triggering a timeout which in turn triggers #401.
2892 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2893 in the DK code in transports.c. sendfile() is not really portable,
2894 hence the _LINUX specificness.
2896 TF/01 In the add_headers option to the mail command in an Exim filter,
2897 there was a bug that Exim would claim a syntax error in any
2898 header after the first one which had an odd number of characters
2901 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2902 callout verification, Exim cached a "reject" for the entire domain. This
2903 is correct for most verifications, but it is not correct for a recipient
2904 verification with use_sender or use_postmaster set, because in that case
2905 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2906 case of MAIL FROM:<> rejection from other early rejections (e.g.
2907 rejection of HELO). When verifying a recipient using a non-null MAIL
2908 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2909 Whatever the result of the callout, the value of the domain cache is
2910 left unchanged (for any other kind of callout, getting as far as trying
2911 RCPT means that the domain itself is ok).
2913 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2916 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2917 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2918 session, the daemon ignores ECONNECT errors and logs others; it now
2919 ignores EPIPE as well.
2921 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2922 (quoted-printable decoding).
2924 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2925 later the small subsequent patch to fix an introduced bug.
2927 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2929 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2931 PH/08 An error is now given if message_size_limit is specified negative.
2933 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2934 to be given (somewhat) arbitrary names.
2936 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2939 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2940 miscellaneous code fixes
2942 PH/10 Added the log_reject_target ACL modifier to specify where to log
2945 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2946 hostname. This is wrong, because it relates to the incoming message (and
2947 probably the interface on which it is arriving) and not to the outgoing
2948 callout (which could be using a different interface). This has been
2949 changed to use the value of the helo_data option from the smtp transport
2950 instead - this is what is used when a message is actually being sent. If
2951 there is no remote transport (possible with a router that sets up host
2952 addresses), $smtp_active_hostname is used.
2954 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2955 tweaks were necessary in order to get it to work (see also 21 below):
2956 (a) The code assumed that strncpy() returns a negative number on buffer
2957 overflow, which isn't the case. Replaced with Exim's string_format()
2959 (b) There were several signed/unsigned issues. I just did the minimum
2960 hacking in of casts. There is scope for a larger refactoring.
2961 (c) The code used strcasecmp() which is not a standard C function.
2962 Replaced with Exim's strcmpic() function.
2963 (d) The code set only $1; it now sets $auth1 as well.
2964 (e) A simple test gave the error "authentication client didn't specify
2965 service in request". It would seem that Dovecot has changed its
2966 interface. Fortunately there's a specification; I followed it and
2967 changed what the client sends and it appears to be working now.
2969 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2972 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2973 address A is aliased to B and C, where B exists and C does not. Without
2974 -v the output is "A verified" because verification stops after a
2975 successful redirection if more than one address is generated. However,
2976 with -v the child addresses are also verified. Exim was outputting "A
2977 failed to verify" and then showing the successful verification for C,
2978 with its parentage. It now outputs "B failed to verify", showing B's
2979 parentage before showing the successful verification of C.
2981 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2982 look up a TXT record in a specific list after matching in a combined
2985 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2986 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2987 they consult the DNS. I had assumed they would set it the way they
2988 wanted; and indeed my experiments on Linux seem to show that in some
2989 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2990 To be on the safe side, however, I have now made the interface to
2991 host_find_byname() similar to host_find_bydns(), with an argument
2992 containing the DNS resolver options. The host_find_byname() function now
2993 sets these options at its start, just as host_find_bydns() does. The smtp
2994 transport options dns_qualify_single and dns_search_parents are passed to
2995 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2996 of host_find_byname() use the default settings of RES_DEFNAMES
2997 (qualify_single) but not RES_DNSRCH (search_parents).
2999 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3000 spool_read_header() do less string testing, by means of a preliminary
3001 switch on the second character of optional "-foo" lines. (This is
3002 overdue, caused by the large number of possibilities that now exist.
3003 Originally there were few.) While I was there, I also converted the
3004 str(n)cmp tests so they don't re-test the leading "-" and the first
3005 character, in the hope this might squeeze out yet more improvement.
3007 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3008 flag allowing group syntax was set by the header_syntax check but not
3009 turned off, possible causing trouble later; (2) The flag was not being
3010 set at all for the header_verify test, causing "group"-style headers to
3011 be rejected. I have now set it in this case, and also caused header_
3012 verify to ignore an empty address taken from a group. While doing this, I
3013 came across some other cases where the code for allowing group syntax
3014 while scanning a header line wasn't quite right (mostly, not resetting
3015 the flag correctly in the right place). These bugs could have caused
3016 trouble for malformed header lines. I hope it is now all correct.
3018 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3019 with the "reply" argument non-NULL. The code, however (which originally
3020 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3021 but it didn't always do it. This confused somebody who was copying the
3022 code for some other use. I have removed all the tests.
3024 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3025 feature that was used to support insecure browsers during the U.S. crypto
3026 embargo. It requires special client support, and Exim is probably the
3027 only MTA that supported it -- and would never use it because real RSA is
3028 always available. This code has been removed, because it had the bad
3029 effect of slowing Exim down by computing (never used) parameters for the
3030 RSA_EXPORT functionality.
3032 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3033 authenticator to fail if there's a tab character in the incoming data
3034 (there should never be unless someone is messing about, as it's supposed
3035 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3036 if the connection is using TLS or if the remote IP is the same as the
3037 local IP, and the "valid-client-cert option" if a client certificate has
3040 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3041 authenticators. This can be used for authorization after authentication
3042 succeeds. (In the case of plaintext, it servers for both authentication
3045 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3046 if any retry times were supplied.
3048 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3049 connection, where there is no client IP address to check. In this
3050 situation, the verify now always succeeds.
3052 PH/25 Applied John Jetmore's -Mset patch.
3054 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3056 PH/27 In a string expansion for a processed (not raw) header when multiple
3057 headers of the same name were present, leading whitespace was being
3058 removed from all of them, but trailing whitespace was being removed only
3059 from the last one. Now trailing whitespace is removed from each header
3060 before concatenation. Completely empty headers in a concatenation (as
3061 before) are ignored.
3063 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3064 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3066 PH/29 [Removed. This was a change that I later backed out, and forgot to
3067 correct the ChangeLog entry (that I had efficiently created) before
3068 committing the later change.]
3070 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3071 address errors (4xx response to RCPT) over the same connection as other
3072 messages routed to the same hosts. Such deliveries are always "forced",
3073 so retry times are not inspected. This resulted in far too many retries
3074 for the affected addresses. The effect occurred only when there were more
3075 hosts than the hosts_max_try setting in the smtp transport when it had
3076 the 4xx errors. Those hosts that it had tried were not added to the list
3077 of hosts for which the message was waiting, so if all were tried, there
3078 was no problem. Two fixes have been applied:
3080 (i) If there are any address or message errors in an SMTP delivery, none
3081 of the hosts (tried or untried) are now added to the list of hosts
3082 for which the message is waiting, so the message should not be a
3083 candidate for sending over the same connection that was used for a
3084 successful delivery of some other message. This seems entirely
3085 reasonable: after all the message is NOT "waiting for some host".
3086 This is so "obvious" that I'm not sure why it wasn't done
3087 previously. Hope I haven't missed anything, but it can't do any
3088 harm, as the worst effect is to miss an optimization.
3090 (ii) If, despite (i), such a delivery is accidentally attempted, the
3091 routing retry time is respected, so at least it doesn't keep
3092 hammering the server.
3094 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3095 in ${readsocket because some servers need this prod.
3097 PH/32 Added some extra debug output when updating a wait-xxx database.
3099 PH/33 The hint "could be header name not terminated by colon", which has been
3100 given for certain expansion errors for a long time, was not being given
3101 for the ${if def:h_colon_omitted{... case.
3103 PH/34 The spec says: "With one important exception, whenever a domain list is
3104 being scanned, $domain contains the subject domain." There was at least
3105 one case where this was not true.
3107 PH/35 The error "getsockname() failed: connection reset by peer" was being
3108 written to the panic log as well as the main log, but it isn't really
3109 panic-worthy as it just means the connection died rather early on. I have
3110 removed the panic log writing for the ECONNRESET error when getsockname()
3113 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3114 runs only) independently of the message's sender address. This meant
3115 that, if the 4xx error was in fact related to the sender, a different
3116 message to the same recipient with a different sender could confuse
3117 things. In particular, this can happen when sending to a greylisting
3118 server, but other circumstances could also provoke similar problems.
3119 I have changed the default so that the retry time for these errors is now
3120 based a combination of the sender and recipient addresses. This change
3121 can be overridden by setting address_retry_include_sender=false in the
3124 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3125 remote server are returned as part of bounce messages. This was not
3126 happening for LMTP over a pipe (the lmtp transport), but now it is the
3127 same for both kinds of LMTP.
3129 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3130 in header lines that were in fact CNAMEs. This is no longer the case.
3132 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3133 and queue runs started by the daemon processed all messages. This has
3134 been fixed so that -R and -S can now usefully be given with -q<time>.
3136 PH/40 Import PCRE release 6.7 (fixes some bugs).
3138 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3140 PH/42 Give an error if -q is specified more than once.
3142 PH/43 Renamed the variables $interface_address and $interface_port as
3143 $received_ip_address and $received_port, to make it clear that these
3144 values apply to message reception, and not to the outgoing interface when
3145 a message is delivered. (The old names remain recognized, of course.)
3147 PH/44 There was no timeout on the connect() call when using a Unix domain
3148 socket in the ${readsocket expansion. There now is.
3150 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3151 be meaningful with "accept".
3153 SC/01 Eximstats V1.43
3154 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3156 SC/02 Eximstats V1.44
3157 Use a glob alias rather than an array ref in the generated
3158 parser. This improves both readability and performance.
3160 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3161 Collect SpamAssassin and rejection statistics.
3162 Don't display local sender or destination tables unless
3163 there is data to show.
3164 Added average volumes into the top table text output.
3166 SC/04 Eximstats V1.46
3167 Collect data on the number of addresses (recipients)
3168 as well as the number of messages.
3170 SC/05 Eximstats V1.47
3171 Added 'Message too big' to the list of mail rejection
3172 reasons (thanks to Marco Gaiarin).
3174 SC/06 Eximstats V1.48
3175 Mainlog lines which have GMT offsets and are too short to
3176 have a flag are now skipped.
3178 SC/07 Eximstats V1.49 (Alain Williams)
3179 Added the -emptyok flag.
3181 SC/08 Eximstats V1.50
3182 Fixes for obtaining the IP address from reject messages.
3184 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3185 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3186 whitespace changes from 4.64-PH/27
3188 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3191 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3192 are found, allow negative numbers in numeric criteria)
3194 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3196 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3197 to match changes made in 4.64-PH/43
3199 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3201 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3202 "domain" is actually an IP address, and force a failure. This locks out
3203 those revolvers/nameservers that support "A-for-A" lookups, in
3204 contravention of the specifications.
3206 PH/48 When a host name was looked up from an IP address, and the subsequent
3207 forward lookup of the name timed out, the host name was left in
3208 $sender_host_name, contrary to the specification.
3210 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3211 restricted to single-key lookups, Exim was not diagnosing an error if
3212 * or *@ was used with a query-style lookup.
3214 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3216 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3217 long ago, but noone interested enough thought of it. Let's just say that
3218 the "1.1" means that there are some new functions that weren't there at
3219 some point in the past.
3221 PH/51 Error processing for expansion failure of helo_data from an smtp
3222 transport during callout processing was broken.
3224 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3225 tested/used via the -bh/-bhc/-bs options.
3227 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3228 bug, fixed in subsequent PCRE releases).
3230 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3231 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3233 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3239 SC/01 Use a glob alias rather than an array ref in eximstats generated
3240 parser. This improves both readability and performance.
3242 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3243 Don't display local sender or destination tables in eximstats unless
3244 there is data to show.
3245 Added average volumes into the eximstats top table text output.
3247 SC/03 Collect data on the number of addresses (recipients) as well
3248 as the number of messages in eximstats.
3250 TF/01 Correct an error in the documentation for the redirect router. Exim
3251 does (usually) call initgroups() when daemonizing.
3253 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3254 with consistent privilege compared to when running as a daemon.
3256 TF/03 Note in the spec that $authenticated_id is not set for local
3257 submissions from trusted users.
3259 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3260 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3262 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3263 by adding some example configuration directives to the default
3264 configuration file. A little bit of work is required to uncomment the
3265 directives and define how usernames and passwords are checked, but
3266 there is now a framework to start from.
3268 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3269 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3270 without this. I don't know how relevant this is to other LDAP libraries.
3272 PH/02 Add the verb name to the "unknown ACL verb" error.
3274 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3276 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3278 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3279 directory not expanded when it should be if an expanded home directory
3280 was set for the address (which is overridden by the transport).
3282 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3285 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3286 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3287 because it is too late at that time, and has no effect.
3289 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3290 security issue with \' (bugzilla #107). I could not use the
3291 PQescapeStringConn() function, because it needs a PGconn value as one of
3294 PH/08 When testing addresses using -bt, indicate those final addresses that
3295 are duplicates that would not cause an additional delivery. At least one
3296 person was confused, thinking that -bt output corresponded to deliveries.
3297 (Suppressing duplicates isn't a good idea as you lose the information
3298 about possibly different redirections that led to the duplicates.)
3300 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3301 systems where poll() doesn't work, in particular OS X.
3303 PH/10 Added more information to debugging output for retry time not reached.
3305 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3306 operations in malware.c.
3308 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3311 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3312 syslog_duplication set false, log lines that would normally be written
3313 both the the main log and to the reject log were not written to syslog at
3316 PH/14 In the default configuration, change the use of "message" in ACL warn
3317 statements to "add_header".
3319 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3320 not followed by a command (e.g. "seen endif").
3322 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3323 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3326 PH/17 Added extra conditions to the default value of delay_warning_condition
3330 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3331 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3332 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3335 The Auto-Submitted: and various List- headers are standardised, whereas I
3336 don't think Precedence: ever was.
3338 PH/18 Refactored debugging code in route_finduser() to show more information,
3339 in particular, the error code if getpwnam() issues one.
3341 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3342 This is apparently needed in addition to the PH/07 change above to avoid
3343 any possible encoding problems.
3345 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3346 but not after initializing Perl.
3348 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3349 output them only if debugging. By default they are written stderr,
3350 apparently, which is not desirable.
3352 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3355 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3358 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3360 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3361 authenticated or an ident call has been made. Suppress the default
3362 values for $authenticated_id and $authenticated_sender (but permit -oMai
3363 and -oMas) when testing with -bh.
3365 PH/24 Re-jigged the order of the tests in the default configuration so that the
3366 tests for valid domains and recipients precede the DNS black list and CSA
3367 tests, on the grounds that those ones are more expensive.
3369 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3370 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3371 This bug exists in every version of Exim that I still have, right back to
3374 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3375 However, an attempt to turn on case-sensitivity in a regex key by
3376 including (?-i) didn't work because the subject string was already
3377 lowercased, and the effects were non-intuitive. It turns out that a
3378 one-line patch can be used to allow (?-i) to work as expected.
3384 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3385 other effects) broke the use of negated acl sub-conditions.
3387 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3390 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3391 "Deny" causes Exim to reject the incoming connection with a 554 error.
3392 Unfortunately, if there is a major crisis, such as a disk failure,
3393 tcp-wrappers gives "deny", whereas what one would like would be some
3394 kind of temporary error. A kludge has been added to help with this.
3395 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3396 554 error is used if errno is still zero or contains ENOENT (which occurs
3397 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3400 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3402 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3403 errors. Otherwise a message that provokes a temporary error (when other
3404 messages do not) can cause a whole host to time out.
3406 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3407 addresses were routed directly to files or pipes from a redirect router.
3408 File deliveries just didn't batch; pipe deliveries might have suffered
3411 PH/06 A failure to get a lock for a hints database would erroneously always say
3412 "Failed to get write lock", even when it was really a read lock.
3414 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3415 of 0600. This has been changed to use the value of the lockfile_mode
3416 option (which defaults to 0600).
3418 PH/08 Applied small patch from the Sieve maintainer.
3420 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3421 folder from quota calculations, a direct delivery into this folder messed
3422 up the contents of the maildirsize file. This was because the regex was
3423 used only to exclude .Trash (or whatever) when the size of the mailbox
3424 was calculated. There was no check that a delivery was happening into an
3425 excluded directory. This bug has been fixed by ignoring all quota
3426 processing for deliveries into excluded directories.
3428 PH/10 Added the maildirfolder_create_regex option to appendfile.
3434 PH/01 The code for finding all the local interface addresses on a FreeBSD
3435 system running IPv6 was broken. This may well have applied to all BSD
3436 systems, as well as to others that have similar system calls. The broken
3437 code found IPv4 interfaces correctly, but gave incorrect values for the
3438 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3439 that it would not match correctly against @[] and not recognize the IPv6
3442 PH/02 The ipliteral router was not recognizing addresses of the form user@
3443 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3445 PH/03 Added disable_ipv6.
3447 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3448 decoded form, because it is most often used to construct To: headers
3449 lines in autoreplies, and the decoded form may well be syntactically
3450 invalid. However, $reply_address has leading white space removed, and all
3451 newlines turned into spaces so that the autoreply transport does not
3454 PH/05 If group was specified without a user on a router, and no group or user
3455 was specified on a transport, the group from the router was ignored.
3457 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3458 for visible compile-time settings that can be used to change these
3459 numbers, for those that want even more. Backwards compatibility with old
3460 spool files has been maintained. However, going back to a previous Exim
3461 release will lost any variables that are in spool files.
3463 PH/07 Two small changes when running in the test harness: increase delay when
3464 passing a TCP/IP connection to a new process, in case the original
3465 process has to generate a bounce, and remove special handling of
3466 127.0.0.2 (sic), which is no longer necessary.
3468 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3469 be the same on different OS.
3471 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3474 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3475 whether --show-vars was specified or not
3477 JJ/02 exipick: Added support for new ACL variable spool format introduced
3480 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3481 syntactically invalid From: or Reply-to: line, and a filter used this to
3482 generate an autoreply, and therefore failed to obtain an address for the
3483 autoreply, Exim could try to deliver to a non-existent relative file
3484 name, causing unrelated and misleading errors. What now happens is that
3485 it logs this as a hard delivery error, but does not attempt to create a
3488 PH/11 The exinext utility has a -C option for testing purposes, but although
3489 the given file was scanned by exinext itself; it wasn't being passed on
3490 when Exim was called.
3492 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3493 an end-of-file indication when reading a command response.
3495 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3496 compiled. In many other places in Exim, IPv6 addresses are always
3497 recognized, so I have changed this. It also means that IPv4 domain
3498 literals of the form [IPV4:n.n.n.n] are now always recognized.
3500 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3501 used if the router is not running as root, for example, when verifying at
3502 ACL time, or when using -bh. The debugging output from this situation was
3503 non-existent - all you got was a failure to exec. I have made two
3506 (a) Failures to set uid/gid, the current directory, or a process leader
3507 in a subprocess such as that created by queryprogram now generate
3508 suitable debugging output when -d is set.
3510 (b) The queryprogram router detects when it is not running as root,
3511 outputs suitable debugging information if -d is set, and then runs
3512 the subprocess without attempting to change uid/gid.
3514 PH/15 Minor change to Makefile for building test_host (undocumented testing
3517 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3518 additional section of a DNS packet that returns MX or SRV records.
3519 Instead, it always explicitly searches for A/AAAA records. This avoids
3520 major problems that occur when a DNS server includes only records of one
3521 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3522 fixed another bug: if SRV records were looked up and the corresponding
3523 address records were *not* found in the additional section, the port
3524 values from the SRV records were lost.
3526 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3527 using the correct key (the original address) when searching the retry
3528 rules in order to find which one to use for generating the retry hint.
3530 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3531 adding the default one. Similarly, if it contains a Reply-To: header, the
3532 errors_reply_to option, if set, is not used.
3534 PH/19 When calculating a retry time, Exim used to measure the "time since
3535 failure" by looking at the "first failed" field in the retry record. Now
3536 it does not use this if it is later than than the arrival time of the
3537 message. Instead it uses the arrival time. This makes for better
3538 behaviour in cases where some deliveries succeed, thus re-setting the
3539 "first failed" field. An example is a quota failure for a huge message
3540 when small messages continue to be delivered. Without this change, the
3541 "time since failure" will always be short, possible causing more frequent
3542 delivery attempts for the huge message than are intended.
3543 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3545 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3546 $1, $2, $3) because the numerical variables can be reset during some
3547 expansion items (e.g. "match"), thereby losing the authentication data.
3549 PH/21 Make -bV show the size of off_t variables so that the test suite can
3550 decide whether to run tests for quotas > 2G.
3552 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3553 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3554 filecount value is greater than 2G or if a quota value is greater than 2G
3555 on a system where the size of off_t is not greater than 4, a panic error
3558 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3559 never match. The debug and -bh output now contains an explicit error
3560 message indicating a malformed IPv4 address or mask.
3562 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3563 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3564 PH/23 above applies.
3566 PH/25 Do not write to syslog when running in the test harness. The only
3567 occasion when this arises is a failure to open the main or panic logs
3568 (for which there is an explicit test).
3570 PH/26 Added the /no_tell option to "control=freeze".
3572 PH/27 If a host name lookup failed very early in a connection, for example, if
3573 the IP address matched host_lookup and the reverse lookup yielded a name
3574 that did not have a forward lookup, an error message of the form "no IP
3575 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3576 could be logged. Now it outputs the IP address instead of "NULL".
3578 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3579 allows the sender and the authenticated sender to be set when
3580 submitting a message from within Exim. Since child_open_exim() is
3581 documented for local_scan(), the new function should be too.
3583 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3584 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3585 results in an empty string is now treated as unset.
3587 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3589 PH/31 Added sender_verify_fail logging option.
3591 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3592 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3593 tidied the source and removed it altogether.
3595 PH/33 When a queue run was abandoned because the load average was too high, a
3596 log line was always written; now it is written only if the queue_run log
3597 selector is set. In addition, the log line for abandonment now contains
3598 information about the queue run such as the pid. This is always present
3599 in "start" and "stop" lines but was omitted from the "abandon" line.
3601 PH/34 Omit spaces between a header name and the colon in the error message that
3602 is given when verify = headers_syntax fails (if there are lots of them,
3603 the message gets confusing).
3605 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3606 names, as there are now some PTR records that contain slashes. This check
3607 is only to protect against broken name servers that fall over on strange
3608 characters, so the fact that it applies to all lookups doesn't matter.
3610 PH/36 Now that the new test suite is complete, we can remove some of the
3611 special code in Exim that was needed for the old test suite. For example,
3612 sorting DNS records because real resolvers return them in an arbitrary
3613 order. The new test suite's fake resolver always returns records in the
3616 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3617 bounces) except when queue_only is set, to avoid logging races between
3618 the different processes.
3620 PH/38 Panic-die if .include specifies a non-absolute path.
3622 PH/39 A tweak to the "H" retry rule from its user.
3624 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3625 a label. They prevented compilation on older perls.
3627 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3628 a warning to be raised on newish perls.
3630 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3631 on queue. Changes to match documented behaviour of showing count of
3632 messages matching specified criteria.
3634 PH/40 Changed the default ident timeout from 30s to 5s.
3636 PH/41 Added support for the use of login_cap features, on those BSD systems
3637 that have them, for controlling the resources used by pipe deliveries.
3639 PH/42 The content-scanning code uses fopen() to create files in which to put
3640 message data. Previously it was not paying any attention to the mode of
3641 the files. Exim runs with umask(0) because the rest of the code creates
3642 files with open(), and sets the required mode explicitly. Thus, these
3643 files were ending up world-writeable. This was not a big issue, because,
3644 being within the spool directory, they were not world-accessible. I have
3645 created a function called modefopen, which takes an additional mode
3646 argument. It sets umask(777), creates the file, chmods it to the required
3647 mode, then resets the umask. All the relevant calls to fopen() in the
3648 content scanning code have been changed to use this function.
3650 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3651 to 24 hours. This avoids potential overflow problems when processing G
3652 and H retry rules. I suspect nobody ever tinkers with this value.
3654 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3656 PH/45 When the plaintext authenticator is running as a client, the server's
3657 challenges are checked to ensure they are valid base64 strings. By
3658 default, the authentication attempt is cancelled if an invalid string is
3659 received. Setting client_ignore_invalid_base64 true ignores these errors.
3660 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3661 they are received. Thus, the responses can be made to depend on the
3662 challenges. If an invalid string is ignored, an empty string is placed in
3665 PH/46 Messages that are created by the autoreply transport now contains a
3666 References: header, in accordance with RFCs 2822 and 3834.
3668 PH/47 Added authenticated_sender_force to the smtp transport.
3670 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3672 PH/49 Installed latest patch from the Sieve maintainer.
3674 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3675 was also set, the appendfile transport was unnecessarily scanning a
3676 directory of message files (e.g. for maildir delivery) to find the count
3677 of files (along with the size), even though it did not need this
3678 information. It now does the scan only if it needs to find either the
3679 size of the count of files.
3681 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3683 PH/52 Two bugs concerned with error handling when the smtp transport is
3686 (i) Exim was not creating retry information for temporary errors given
3687 for individual recipients after the DATA command when the smtp transport
3688 was used in LMTP mode. This meant that they could be retried too
3689 frequently, and not timed out correctly.
3691 (ii) Exim was setting the flag that allows error details to be returned
3692 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3693 recipients that were returned after the DATA command.
3695 PH/53 This is related to PH/52, but is more general: for any failing address,
3696 when detailed error information was permitted to be returned to the
3697 sender, but the error was temporary, then after the final timeout, only
3698 "retry timeout exceeded" was returned. Now it returns the full error as
3699 well as "retry timeout exceeded".
3701 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3702 do this, and (what is worse) MTAs that accept it.
3704 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3705 will now be deprecated.
3707 PH/56 New os.c-cygwin from the Cygwin maintainer.
3709 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3710 formats (previously only available in exim formats via -bpr, -bpru,
3711 and -bpra. Now also available in native and exiqgrep formats)
3713 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3714 with very large, slow to parse queues
3716 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3718 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3720 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3721 responses to authentication challenges, though it was showing the
3722 challenges; (ii) I've removed the CR characters from the debug output for
3725 PH/58 Allow for the insertion of a newline as well as a space when a string
3726 is turned into more than one encoded-word during RFC 2047 encoding. The
3727 Sieve code now uses this.
3729 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3730 data_4xx, lost_connection, tls_required.
3732 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3733 message was being sent as an SMTP response.
3735 PH/61 Add -l and -k options to exicyclog.
3737 PH/62 When verifying, if an address was redirected to one new address, so that
3738 verification continued, and the new address failed or deferred after
3739 having set something in $address_data, the value of $address_data was not
3740 passed back to the ACL. This was different to the case when no
3741 redirection occurred. The value is now passed back in both cases.
3743 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3744 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3745 use login_cap.h, so on its own it isn't the distinguishing feature. The
3746 new name refers directly to the setclassresources() function.
3748 PH/65 Added configuration files for NetBSD3.
3750 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3752 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3753 is preferred over IPv4.
3755 PH/68 The bounce_return_message and bounce_return_body options were not being
3756 honoured for bounces generated during the reception of non-SMTP messages.
3757 In particular, this applied to messages rejected by the ACL. This bug has
3758 been fixed. However, if bounce_return_message is true and bounce_return_
3759 body is false, the headers that are returned for a non-SMTP message
3760 include only those that have been read before the error was detected.
3761 (In the case of an ACL rejection, they have all been read.)
3763 PH/69 The HTML version of the specification is now built in a directory called
3764 spec_html instead of spec.html, because the latter looks like a path with
3765 a MIME-type, and this confuses some software.
3767 PH/70 Catch two compiler warnings in sieve.c.
3769 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3770 function verify_get_ident() calls ip_connect() to connect a socket, but
3771 if the "connect()" function timed out, ip_connect() used to close the
3772 socket. However, verify_get_ident() also closes the socket later, and in
3773 between Exim writes to the log, which may get opened at this point. When
3774 the socket was closed in ip_connect(), the log could get the same file
3775 descriptor number as the socket. This naturally causes chaos. The fix is
3776 not to close the socket in ip_connect(); the socket should be closed by
3777 the function that creates it. There was only one place in the code where
3778 this was missing, in the iplookup router, which I don't think anybody now
3779 uses, but I've fixed it anyway.
3781 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3782 well as to direct DNS lookups. Otherwise the handling of names in host
3783 lists is inconsistent and therefore confusing.
3789 PH/01 Two changes to the default runtime configuration:
3791 (1) Move the checks for relay_from_hosts and authenticated clients from
3792 after to before the (commented out) DNS black list checks.
3794 (2) Add control=submission to the relay_from_hosts and authenticated
3795 clients checks, on the grounds that messages accepted by these
3796 statements are most likely to be submissions.
3798 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3800 (1) Generate an error if the third argument for the ${prvs expansion is
3803 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3806 (3) Reset the variables that are obtained from the first argument of
3807 ${prvscheck and used in the second argument before leaving the code,
3808 because their memory is reclaimed, so using them afterwards may do
3811 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3812 one (it's much easier than Tom thought :-).
3814 (5) Because of (4), we can now allow for the use of $prvscheck_result
3815 inside the third argument.
3817 PH/03 For some reason, the default setting of PATH when running a command from
3818 a pipe transport was just "/usr/bin". I have changed it to
3821 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3822 anything to be listed in the output from -bV.
3824 PH/05 When a filter generated an autoreply, the entire To: header line was
3825 quoted in the delivery log line, like this:
3827 => >A.N.Other <ano@some.domain> <original@ddress> ...
3829 This has been changed so that it extracts the operative address. There
3830 may be more than one such address. If so, they are comma-separated, like
3833 => >ano@some.domain,ona@other.domain <original@ddress> ...
3835 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3836 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3837 not being looked up in the rDNS to get a host name, Exim was showing the
3838 IP address twice in Received: lines, even though the IP addresses were
3839 identical. For example:
3841 Received: from [1.2.3.4] (helo=[1.2.3.4])
3843 However, if the real host name was known, it was omitting the HELO data
3844 if it matched the actual IP address. This has been tidied up so that it
3845 doesn't show the same IP address twice.
3847 PH/07 When both +timestamp and +memory debugging was on, the value given by
3848 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3849 called by the debug printing, thereby overwriting the timestamp buffer.
3850 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3852 PH/08 When the original message was included in an autoreply transport, it
3853 always said "this is a copy of the message, including all the headers",
3854 even if body_only or headers_only was set. It now gives an appropriate
3857 PH/09 Applied a patch from the Sieve maintainer which:
3859 o fixes some comments
3860 o adds the (disabled) notify extension core
3861 o adds some debug output for the result of if/elsif tests
3862 o points to the current vacation draft in the documentation
3863 and documents the missing references header update
3867 o fixes a bug in processing the envelope test (when testing
3868 multiple envelope elements, the last element determined the
3871 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3872 Electronic Mail") by including:
3874 Auto-submitted: auto-generated
3876 in the messages that it generates (bounce messages and others, such as
3877 warnings). In the case of bounce messages for non-SMTP messages, there was
3878 also a typo: it was using "Auto_submitted" (underscore instead of
3879 hyphen). Since every message generated by Exim is necessarily in response
3880 to another message, thes have all been changed to:
3882 Auto-Submitted: auto-replied
3884 in accordance with these statements in the RFC:
3886 The auto-replied keyword:
3888 - SHOULD be used on messages sent in direct response to another
3889 message by an automatic process,
3891 - MUST NOT be used on manually-generated messages,
3893 - MAY be used on Delivery Status Notifications (DSNs) and Message
3894 Disposition Notifications (MDNs),
3896 - MUST NOT be used on messages generated by automatic or periodic
3897 processes, except for messages which are automatic responses to
3900 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3901 to the default Received: header definition.
3903 PH/12 Added log selector acl_warn_skipped (default on).
3905 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3906 variables because (a) they are in the wrong storage pool and (b) even if
3907 they were copied, it wouldn't work properly because of the caching.
3909 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3910 checking when decoding. Apparently there are clients that generate
3911 overlong encoded strings. Why am I not surprised?
3913 PH/15 If the first argument of "${if match_address" was not empty, but did not
3914 contain an "@" character, Exim crashed. Now it writes a panic log message
3915 and treats the condition as false.
3917 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3919 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3920 extension "envelope-auth". The code is finished and in agreement with
3921 other implementations, but there is no documentation so far and in fact,
3922 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3923 not changing the active code.
3925 Print executed "if" and "elsif" statements when debugging is used. This
3926 helps a great deal to understand what a filter does.
3928 Document more things not specified clearly in RFC3028. I had all this
3929 sorted out, when out of a sudden new issues came to my mind. Oops."
3931 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3934 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3935 canonical form (as documented). However, after a host name lookup from
3936 the IP address, check_host() was doing a simple string comparison with
3937 addresses acquired from the DNS when checking that the found name did
3938 have the original IP as one of its addresses. Since any found IPv6
3939 addresses are likely to be in abbreviated form, the comparison could
3940 fail. Luckily, there already exists a function for doing the comparison
3941 by converting both addresses to binary, so now that is used instead of
3942 the text comparison.
3944 PH/20 There was another similar case to PH/19, when a complete host name was
3945 given in a host list; looking up its IP address could give an abbreviated
3946 form, whereas the current host's name might or might not be abbreviated.
3947 The same fix has been applied.
3953 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3954 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3957 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3958 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3960 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3962 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3963 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3964 Also, for "Auto-Submitted", treat anything other than "no" as
3965 non-personal, in accordance with RFC 3834. (Previously it treated
3966 anything starting "auto-" as non-personal.)
3968 TF/01 The control=submission/name=... option had a problem with syntax
3969 errors if the name included a slash character. The /name= option
3970 now slurps the rest of the string, so it can include any characters
3971 but it must come last in the list of options (after /sender_retain
3974 PH/05 Some modifications to the interface to the fake nameserver for the new
3982 TK/01 Added the "success_on_redirect" address verification option. See
3983 NewStuff for rationale and an example.
3985 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3987 PH/02 Patch to exigrep to allow it to work on syslog lines.
3989 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3990 fread() to skip over the body file's header line, because in Cygwin the
3991 header line is locked and is inaccessible.
3993 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3994 co-exist for some time) to make it clear that it is the Exim ID that is
3995 referenced, not the Message-ID: header line.
3997 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3998 string_format() function, because snprintf() does not exist on all
4001 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4002 ${stat: expansion item.
4004 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4005 protocol synchronization error", to keep the pedants happy.
4007 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4008 well as for IRIX systems, when gcc is being used. See the host.c source
4011 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4013 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4016 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4017 transport and to the smtp transport in LMTP mode.
4019 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4021 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4022 run a filter in a subprocess. This could lead to confusion in subsequent
4023 lookups in the parent process. There should also be a search_tidyup() at
4024 the end of the subprocess.
4026 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4027 only if the host matched helo_try_verify_hosts, which caused the
4028 verification to occur when the EHLO/HELO command was issued. The ACL just
4029 tested the remembered result. Now, if a previous verification attempt has
4030 not happened, "verify = helo" does it there and then.
4032 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4034 TK/03 Fix log output including CR from clamd.
4036 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4037 exist provoked a memory error which could cause a segfault.
4039 PH/15 Installed PCRE 6.2
4041 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4043 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4044 of the problem. Specifically, suggested +O2 rather than +O1 for the
4047 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4049 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4050 record, the port was not correctly logged when the outgoing_port log
4051 selector was set (it logged the transort's default port).
4053 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4054 fallback_hosts, and "hosts" in the smtp transport.
4056 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4057 host errors such as "Connection refused".
4059 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4060 authentication with radiusclient 0.4.9:
4062 - Error returned from rc_read_config was caught wrongly
4063 - Username/password not passed on to radius server due to wrong length.
4065 The presumption is that some radiusclient API changes for 4.51/PH/17
4066 were not taken care of correctly. The code is still untested by me (my
4067 Linux distribution still has 0.3.2 of radiusclient), but it was
4068 contributed by a Radius user.
4070 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4071 expanding the "port" option of the smtp transport.
4073 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4074 while reading a MIME header. Thanks to Tom Hughes for a patch.
4076 PH/24 Include config.h inside local_scan.h so that configuration settings are
4079 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4080 that in an ACL for RCPT (for example), you can examine exactly what was
4083 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4084 commands, but it was not correctly comparing the address with the actual
4085 client host address. Thus, it would show the EHLO address in Received:
4086 header lines when this was not necessary.
4088 PH/27 Added the % operator to ${eval:}.
4090 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4091 it should be ignoring failures (because with -C, for example, it has lost
4092 privilege). It wasn't ignoring creation failures other than "already
4095 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4098 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4099 because an input line was too long, either on its own, or by virtue of
4100 too many continuations, the temporary file was not being removed, and the
4101 return code was incorrect.
4103 PH/31 Missing "BOOL" in function definition in filtertest.c.
4105 PH/32 Applied Sieve patches from the maintainer.
4107 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4109 PH/33 Added "verify = not_blind".
4111 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4112 Local/Makefile (with some defaults set). These are used in built scripts
4113 such as exicyclog, but they have never been used in the exim_install
4114 script (though there are many overriding facilities there). I have
4115 arranged that the exim_install script now takes note of these two
4118 PH/35 Installed configuration files for Dragonfly.
4120 PH/36 When a locally submitted message by a trusted user did not contain a
4121 From: header, and the sender address was obtained from -f or from an SMTP
4122 MAIL command, and the trusted user did not use -F to supply a sender
4123 name, $originator_name was incorrectly used when constructing a From:
4124 header. Furthermore, $originator_name was used for submission mode
4125 messages from external hosts without From: headers in a similar way,
4126 which is clearly wrong.
4128 PH/37 Added control=suppress_local_fixups.
4130 PH/38 When log_selector = +received_sender was set, and the addition of the
4131 sender made the log line's construction buffer exactly full, or one byte
4132 less than full, an overflow happened when the terminating "\n" was
4135 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4136 when the result of a list match is failure because a DNS lookup failed.
4138 PH/40 RM_COMMAND is now used in the building process.
4140 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4141 the "build-* directories that it finds.
4143 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4144 address in a domain literal was a prefix of an interface address.
4146 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4147 when verifying a sender address, unless rewrite_headers is false.
4149 PH/44 Wrote a long comment about why errors_to addresses are verified as
4150 recipients, not senders.
4152 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4153 the ratelimit ACL was added.
4155 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4157 PH/46 Added extra information about PostgreSQL errors to the error string.
4159 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4160 avoiding the need to install special zones in a real server. This is
4161 backwards compatible; if it can't find the fake resolver, it drops back.
4162 Thus, both old and new test suites can be run.
4164 TF/02 Added util/ratelimit.pl
4166 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4167 clock is set back in time.
4169 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4170 Candler <B.Candler@pobox.com>.
4172 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4173 OK for addresses that are the result of redirections.
4175 PH/48 A number of further additions for the benefit of the new test suite,
4176 including a fake gethostbyname() that interfaces to the fake DNS resolver
4179 TF/06 The fix for widen_domains has also been applied to qualify_single and
4180 search_parents which are the other dnslookup options that can cause
4183 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4186 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4188 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4189 service" if it encountered an MX record with an empty target hostname.
4190 The message is now "an MX or SRV record indicated no SMTP service".
4192 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4193 if the DNS of the sending site is misconfigured. This is quite a
4194 common situation. This change restores the behaviour of treating a
4195 helo verification defer as a failure.
4197 PH/51 If self=fail was set on a router, the bounce message did not include the
4198 actual error message.
4204 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4206 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4207 command itself ended in error, the underlying message about the transport
4208 filter timeout was being overwritten with the pipe command error. Now the
4209 underlying error message should be appended to the second error message.
4211 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4213 PH/02 Reduce the amount of output that "make" produces by default. Full output
4214 can still be requested.
4216 PH/03 The warning log line about a condition test deferring for a "warn" verb
4217 was being output only once per connection, rather than after each
4218 occurrence (because it was using the same function as for successful
4219 "warn" verbs). This seems wrong, so I have changed it.
4221 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4222 it should not have, which might have caused a crash in the right
4223 circumstances, but probably never did.
4225 PH/04 Installed a modified version of Tony Finch's patch to make submission
4226 mode fix the return path as well as the Sender: header line, and to
4227 add a /name= option so that you can make the user's friendly name appear
4230 TF/03 Added the control = fakedefer ACL modifier.
4232 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4233 Mark Lowes for thorough testing.
4235 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4237 TK/03 Merged latest SRS patch from Miles Wilton.
4239 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4240 with the definition in sysexits.h (which is #included earlier).
4241 Fortunately, Exim does not actually use EX_OK. The code used to try to
4242 preserve the sysexits.h value, by assuming that macro definitions were
4243 scanned for macro replacements. I have been disabused of this notion,
4244 so now the code just undefines EX_OK before #including unistd.h.
4246 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4247 in the smtp transport. When a block could not be written in a single
4248 write() function, the timeout was being re-applied to each part-write.
4249 This seems wrong - if the receiver was accepting one byte at a time it
4250 would take for ever. The timeout is now adjusted when this happens. It
4251 doesn't have to be particularly precise.
4253 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4254 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4256 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4257 without a domain if the check to <postmaster@domain> fails.
4259 SC/01 Eximstats: added -xls and the ability to specify output files
4260 (patch written by Frank Heydlauf).
4262 SC/02 Eximstats: use FileHandles for outputting results.
4264 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4266 SC/04 Eximstats: fixed display of large numbers with -nvr option
4268 SC/05 Eximstats: fixed merging of reports with empty tables.
4270 SC/06 Eximstats: added the -include_original_destination flag
4272 SC/07 Eximstats: removed tabs and trailing whitespace.
4274 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4276 TK/06 MBOX spool code: Add real "From " MBOX separator line
4277 so the .eml file is really in mbox format (even though
4278 most programs do not really care). Patch from Alex Miller.
4280 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4281 The latter is generated from $received_to and is only set if the
4282 message has one envelope recipient. SA can use these headers,
4283 obviously out-of-the-box. Patch from Alex Miller.
4285 PH/08 The ${def test on a variable was returning false if the variable's
4286 value was "0", contrary to what the specification has always said!
4287 The result should be true unless the variable is empty.
4289 PH/09 The syntax error of a character other than { following "${if
4290 def:variable_name" (after optional whitespace) was not being diagnosed.
4291 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4292 accidental colon was present, for example, could give incorrect results.
4294 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4295 result is used (not including appendfile, where other changes are about
4298 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4299 This involved changing a lot of size variables from int to off_t. It
4300 should work with maildirs and everything.
4302 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4303 spamd dying while we are connected to it.
4305 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4308 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4309 function for BDB 4.3.
4311 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4313 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4314 the sources for PCRE have been re-arranged and more files are now
4317 PH/15 The code I had for printing potentially long long variables in PH/11
4318 above was not the best (it lost precision). The length of off_t variables
4319 is now inspected at build time, and an appropriate printing format (%ld
4320 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4321 to be "long long int" or "long int". This is needed for the internal
4322 formatting function string_vformat().
4324 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4325 the configuration file to be ":syslog", then the script "guesses" where
4326 the logs files are, rather than using the compiled in default. In our
4327 case the guess is not the same as the compiled default, so the script
4328 suddenly stopped working when I started to use syslog. The patch checks
4329 to see if log_file_path is "". If so, it attempts to read it from exim
4330 with no configuration file to get the compiled in version, before it
4331 falls back to the previous guessing code."
4333 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4334 implementing BATV in an Exim configuration. See NewStuff for the gory
4337 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4338 Makefile that are specific to HP-UX.
4340 PH/18 If the "use_postmaster" option was set for a recipient callout together
4341 with the "random" option, the postmaster address was used as the MAIL
4342 FROM address for the random test, but not for the subsequent recipient
4343 test. It is now used for both.
4345 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4346 patch removes a few documentation additions to RFC 3028, because the
4347 latest draft now contains them. It adds the new en;ascii-case comparator
4348 and a new error check for 8bit text in MIME parts. Comparator and
4349 require names are now matched exactly. I enabled the subaddress
4350 extension, but it is not well tested yet (read: it works for me)."
4352 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4353 rework some of the code of TK/09 above to avoid the hardwired use of
4354 "%lld" and "long long". Replaced the call to snprintf() with a call to
4357 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4358 records point to non-existent hosts", "retry timeout exceeded", and
4359 "retry time not reached for any host after a long failure period".
4361 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4362 experimental DomainKeys support:
4364 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4365 (2) On an error such as an illegally used "control", the wrong name for
4366 the control was given.
4368 These problems did NOT occur unless DomainKeys support was compiled.
4370 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4372 PH/24 Added ${if match_ip condition.
4374 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4375 databases so that it will be absolutely obvious if a crash occurs in the
4376 DB library. This is a regular occurrence (often caused by mis-matched
4379 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4380 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4381 picked up on a user's system that detects such things. There doesn't seem
4382 to be a gcc warning option for this - only an attribute that has to be
4383 put on the function's prototype. It seems that in Fedora Core 4 they have
4384 set this on a number of new functions. No doubt there will be more in due
4387 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4388 specify a transport. However, if an address that was verified by such a
4389 router was the subject of a callout, Exim crashed because it tried to
4390 read the rcpt_include_affixes from the non-existent transport. Now it
4391 just assumes that the setting of that option is false. This bug was
4392 introduced by 4.51/PH/31.
4394 PH/28 Changed -d+all to exclude +memory, because that information is very
4395 rarely of interest, but it makes the output a lot bigger. People tend to
4396 do -d+all out of habit.
4398 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4399 code in os-type was giving problems when libc.so lives in lib64, like on
4402 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4403 aren't the modern standard, and it seems that some systems' include files
4404 don't always have them. Exim was already checking for some of the newer
4405 ones like T_AAAA, and defining it itself. I've added checks for all the
4406 record types that Exim uses.
4408 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4409 not automatically generating a new one, as it is supposed to. This
4410 prevented TLS from working. If the file did exist, but contained invalid
4411 data, a new version was generated, as expected. It was only the case of a
4412 non-existent file that was broken.
4414 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4415 with a change in libdomainkeys > 0.64.
4417 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4418 from DNS. If the selector record carries the flag, it now has
4419 precedence over the domain-wide flag.
4421 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4423 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4424 the use of a port name that isn't defined in /etc/services) occurred, the
4425 message was deferred as in a normal delivery, and thus remained on the
4426 spool, instead of being failed because of the mua_wrapper setting. This
4427 is now fixed, and I tidied up some of the mua_wrapper messages at the
4430 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4431 the messages in a hash of arrays rather than using individual hashes.
4432 This is a bit cleaner and results in dramatic memory savings, albeit
4433 at a slight CPU cost.
4435 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4436 as requested by Marc Sherman.
4438 SC/10 Eximstats: added histograms for user specified patterns as requested
4441 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4443 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4444 fopen() in the content-scanning modules that did not already have it.
4450 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4451 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4453 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4455 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4457 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4458 to test Sieve filters that use "vacation".
4460 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4461 that changes the way the GnuTLS parameters are stored in the cache file.
4462 The new format can be generated externally. For backward compatibility,
4463 if the data in the cache doesn't make sense, Exim assumes it has read an
4464 old-format file, and it generates new data and writes a new file. This
4465 means that you can't go back to an older release without removing the
4468 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4469 work if there are any delivery delays because "one_time" forces the
4470 parent to be marked "delivered", so its unseen clone is never tried
4471 again. For this reason, Exim now forbids the simultaneous setting of
4474 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4475 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4476 entry if you want to know the details. The fix, however, neglected to
4477 consider the case when local delivery batching is involved. The test for
4478 "previously delivered" was not happening when checking to see if an
4479 address could be batched with a previous (undelivered) one; under
4480 certain circumstances this could lead to multiple deliveries to the same
4483 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4484 in its include files, and this causes problems building Exim.
4486 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4487 header_syntax) but Exim was just ignoring anything given after a slash.
4488 In particular, this caused confusion with an attempt to use "verify =
4489 reverse_host_lookup/defer_ok". An error is now given when options are
4490 supplied for verify items that do not have them. (Maybe reverse_host_
4491 lookup should have a defer_ok option, but that's a different point.)
4493 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4494 defined by RFC 821) to 2048, because there were problems with some AUTH
4495 commands, and RFC 1869 says the size should be increased for extended
4496 SMTP commands that take arguments.
4498 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4501 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4502 "unknown" error; now it says that the functionality isn't in the binary.
4504 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4505 an address' error message when a string expansion fails (syntax or
4506 whatever). Otherwise the password may appear in the log. Following change
4507 PH/42 below, there is no longer a chance of it appearing in a bounce
4510 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4512 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4513 of its addresses was ever tried. (Bugzilla bug #2.)
4515 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4516 the result incorrectly in the debug output. (It correctly added a newline
4517 to what was transported.)
4519 TF/01 Added $received_time.
4521 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4522 commented out examples of how to interface to a virus scanner and to
4523 SpamAssassin. Also added commented examples of av_scanner and
4524 spamd_address settings.
4526 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4527 and controls are allowed in which ACLs. There were a couple of minor
4528 errors. Some of the entries in the conditions table (which is a table of
4529 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4530 negation of where the condition IS allowed.
4532 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4534 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4535 header file does not have a version number, so I've had to invent a new
4536 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4537 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4538 radiusclient), but it was contributed by a Radius user.
4540 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4541 files or directories, for OpenSSL.
4543 PH/19 When an Exim process that is running as root has to create an Exim log
4544 file, it does so in a subprocess that runs as exim:exim so as to get the
4545 ownership right at creation (otherwise, other Exim processes might see
4546 the file with the wrong ownership). There was no test for failure of this
4547 fork() call, which would lead to the process getting stuck as it waited
4548 for a non-existent subprocess. Forks do occasionally fail when resources
4549 run out. I reviewed all the other calls to fork(); they all seem to check
4552 PH/20 When checking for unexpected SMTP input at connect time (before writing
4553 the banner), Exim was not dealing correctly with a non-positive return
4554 from the read() function. If the client had disconnected by this time,
4555 the result was a log entry for a synchronization error with an empty
4556 string after "input=" when read() returned zero. If read() returned -1
4557 (an event I could not check), uninitialized data bytes were printed.
4558 There were reports of junk text (parts of files, etc) appearing after
4561 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4563 PH/22 Added support for macro redefinition, and (re)definition in between
4564 driver and ACL definitions.
4566 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4567 forgetting to use the resulting value; it was using the unexpanded value.
4569 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4570 hadn't been configured. The fix is from Juergen Kreileder, who
4571 understands it better than I do:
4573 "Here's what I see happening with three configured cyrus_sasl
4574 authenticators configured (plain, login, cram-md5):
4576 On startup auth_cyrus_sasl_init() gets called for each of these.
4577 This means three calls to sasl_listmech() without a specified mech_list.
4578 => SASL tests which mechs of all available mechs actually work
4579 => three warnings about OTP not working
4580 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4582 With the patch, sasl_listmech() also gets called three times. But now
4583 SASL's mech_list option is set to the server_mech specified in the the
4584 authenticator. Or in other words, the answer from sasl_listmech()
4585 gets limited to just the mech you're testing for (which is different
4587 => the return list contains just 'plain' or 'login', 'cram-md5' or
4588 nothing depending on the value of ob->server_mech.
4590 I've just tested the patch: Authentication still works fine,
4591 unavailable mechs specified in the exim configuration are still
4592 caught, and the auth.log warnings about OTP are gone."
4594 PH/25 When debugging is enabled, the contents of the command line are added
4595 to the debugging output, even when log_selector=+arguments is not
4598 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4599 answer is "GNU", and only if the return is "GNU/something" is the answer
4602 PH/27 $acl_verify_message is now set immediately after the failure of a
4603 verification in an ACL, and so is available in subsequent modifiers. In
4604 particular, the message can be preserved by coding like this:
4606 warn !verify = sender
4607 set acl_m0 = $acl_verify_message
4609 Previously, $acl_verify_message was set only while expanding "message"
4610 and "log_message" when a very denied access.
4612 PH/28 Modified OS/os.c-Linux with
4614 -#ifndef OS_LOAD_AVERAGE
4615 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4617 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4618 nomenclature these days.)
4620 PH/29 Installed patch from the Sieve maintainer that adds the options
4621 sieve_useraddress and sieve_subaddress to the redirect router.
4623 PH/30 In these circumstances:
4624 . Two addresses routed to the same list of hosts;
4625 . First host does not offer TLS;
4626 . First host accepts first address;
4627 . First host gives temporary error to second address;
4628 . Second host offers TLS and a TLS session is established;
4629 . Second host accepts second address.
4630 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4631 and peerdn, if requested) that were in fact used only for the second
4634 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4635 attention to any local part prefix or suffix that was matched by the
4636 router that accepted the address. It now behaves in the same way as it
4637 does for delivery: the affixes are removed from the local part unless
4638 rcpt_include_affixes is set on the transport.
4640 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4641 timeout during the DATA phase of an incoming message.
4643 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4644 applied a patch sanctioned by the Sieve maintainer.
4646 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4647 the uid or gid is negative. A case of a negative gid caused this to be
4648 noticed. The fix allows for either to be negative.
4650 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4651 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4654 PH/36 The change PH/12 above was broken. Fixed it.
4656 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4657 the grounds that routing the same address twice would always produce the
4658 same answer. This might have been true once, but it is certainly no
4659 longer true now. Routing a child address may depend on the previous
4660 routing that produced that child. Some complicated redirection strategies
4661 went wrong when messages had multiple recipients, and made Exim's
4662 behaviour dependent on the order in which the addresses were given.
4664 I have moved the duplicate checking until after the routing is complete.
4665 Exim scans the addresses that are assigned to local and remote
4666 transports, and removes any duplicates. This means that more work will be
4667 done, as duplicates will always all be routed, but duplicates are
4668 presumably rare, so I don't expect this is of any significance.
4670 For deliveries to pipes, files, and autoreplies, the duplicate checking
4671 still happens during the routing process, since they are not going to be
4674 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4675 It corrects a timeout issue with spamd. This is Ian's comment: "The
4676 background is that sometimes spamd either never reads data from a
4677 connection it has accepted, or it never writes response data. The exiscan
4678 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4679 blindly assumes that writes won't block so it may never time out."
4681 PH/39 Allow G after quota size as well as K and M.
4683 PH/40 The value set for $authenticated_id in an authenticator may not contain
4684 binary zeroes or newlines because the value is written to log lines and
4685 to spool files. There was no check on this. Now the value is run through
4686 the string_printing() function so that such characters are converted to
4687 printable escape sequences.
4689 PH/41 $message_linecount is a new variable that contains the total number of
4690 lines in the message. Compare $body_linecount, which is the count for the
4693 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4694 bounce and delay warning messages, except in certain special cases, which
4697 (a) An SMTP error message from a remote host;
4698 (b) A message specified in a :fail: redirection;
4699 (c) A message specified in a "fail" command in a system filter;
4700 (d) A message specified in a FAIL return from the queryprogram router;
4701 (e) A message specified by the cannot_route_message router option.
4703 In these cases only, Exim does include the error details in bounce and
4704 warning messages. There are also a few cases where bland messages such
4705 as "unrouteable address" or "local delivery error" are given.
4707 PH/43 $value is now also set for the "else" part of a ${run expansion.
4709 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4710 being worked on, but at least Exim now implements the latest version to
4713 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4714 process to complete was treated as a delivery failure, a timeout while
4715 writing the message to the pipe was logged, but erroneously treated as a
4716 successful delivery. Such timeouts include transport filter timeouts. For
4717 consistency with the overall process timeout, these timeouts are now
4718 treated as errors, giving rise to delivery failures by default. However,
4719 there is now a new Boolean option for the pipe transport called
4720 timeout_defer, which, if set TRUE, converts the failures into defers for
4721 both kinds of timeout. A transport filter timeout is now identified in
4724 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4725 systems where "make" and "gmake" are different, calling "gmake" at top
4726 level broke things. I've arranged for the value of $(MAKE) to be passed
4727 from the Makefile to this script so that it can call the same version of
4731 A note about Exim versions 4.44 and 4.50
4732 ----------------------------------------
4734 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4735 changes of various kinds. As a consequence, a big documentation update was
4736 needed. This delayed the release for rather longer than seemed good, especially
4737 in the light of a couple of (minor) security issues. Therefore, the changes
4738 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4739 release. So 4.44 and 4.50 are in effect two different branches that both start
4742 I have left the 4.50 change log unchanged; it contains all the changes since
4743 4.43. The change log for 4.44 is below; many of its items are identical to
4744 those for 4.50. This seems to be the most sensible way to preserve the
4745 historical information.
4751 1. Minor wording change to the doc/README.SIEVE file.
4753 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4754 computation of the current number of files was incorrect.
4756 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4757 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4760 4. Give more explanation in the error message when the command for a transport
4761 filter fails to execute.
4763 5. There are several places where Exim runs a non-Exim command in a
4764 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4765 was being done only for the command run by the queryprogram router. It is
4766 now done for all such subprocesses. The other cases are: ${run, transport
4767 filters, and the commands run by the lmtp and pipe transports.
4769 6. Added CONFIGURE_GROUP build-time option.
4771 7. Some older OS have a limit of 256 on the maximum number of file
4772 descriptors. Exim was using setrlimit() to set 1000 as a large value
4773 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4774 systems. I've change it so that if it can't get 1000, it tries for 256.
4776 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4777 was an oversight, and furthermore, ever since the addition of extra
4778 controls (e.g. 4.43/32), the checks on when to allow different forms of
4779 "control" were broken. There should now be diagnostics for all cases when a
4780 control that does not make sense is encountered.
4782 9. Added the /retain_sender option to "control=submission".
4784 10. $recipients is now available in the predata ACL (oversight).
4786 11. Tidy the search cache before the fork to do a delivery from a message
4787 received from the command line. Otherwise the child will trigger a lookup
4788 failure and thereby defer the delivery if it tries to use (for example) a
4789 cached ldap connection that the parent has called unbind on.
4791 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4792 of $address_data from the recipient verification was clobbered by the
4793 sender verification.
4795 13. The value of address_data from a sender verification is now available in
4796 $sender_address_data in subsequent conditions in the ACL statement.
4798 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4800 15. Added a new option "connect=<time>" to callout options, to set a different
4803 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4804 was its contents. (It was OK if the option was not defined at all.)
4806 17. A "Completed" log line is now written for messages that are removed from
4807 the spool by the -Mrm option.
4809 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4810 information about exactly what failed.
4812 19. Added -dd to debug only the daemon process.
4814 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4815 handles timeouts, both on the server side and network timeouts. Renamed the
4816 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4818 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4819 It is now set to "smtps".
4821 22. $host_address is now set to the target address during the checking of
4822 ignore_target_hosts.
4824 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4825 being passed; this would have caused $sender_host_name to have been used if
4826 matching the list had actually called for a host name (not very likely,
4827 since this list is usually IP addresses). A host name is now passed as
4830 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4831 code that specifies a non-restarting handler (typically sigaction() in
4832 modern systems) in an attempt to fix a rare and obscure crash bug.
4834 25. Narrowed the window for a race in the daemon that could cause it to ignore
4835 SIGCHLD signals. This is not a major problem, because they are used only to
4836 wake it up if nothing else does.
4838 26. A malformed maildirsize file could cause Exim to calculate negative values
4839 for the mailbox size or file count. Odd effects could occur as a result.
4840 The maildirsize information is now recalculated if the size or filecount
4843 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4844 support for a long time. Removed HAVE_SYS_VFS_H.
4846 28. Installed the latest version of exipick from John Jetmore.
4848 29. In an address list, if the pattern was not a regular expression, an empty
4849 subject address (from a bounce message) matched only if the pattern was an
4850 empty string. Non-empty patterns were not even tested. This was the wrong
4851 because it is perfectly reasonable to use an empty address as part of a
4852 database query. An empty address is now tested by patterns that are
4853 lookups. However, all the other forms of pattern expect the subject to
4854 contain a local part and a domain, and therefore, for them, an empty
4855 address still always fails if the pattern is not itself empty.
4857 30. Exim went into a mad DNS loop when attempting to do a callout where the
4858 host was specified on an smtp transport, and looking it up yielded more
4859 than one IP address.
4861 31. Re-factored the code for checking spool and log partition space into a
4862 function that finds that data and another that does the check. The former
4863 is then used to implement four new variables: $spool_space, $log_space,
4864 $spool_inodes, and $log_inodes.
4866 32. The RFC2047 encoding function was originally intended for short strings
4867 such as real names; it was not keeping to the 75-character limit for
4868 encoded words that the RFC imposes. It now respects the limit, and
4869 generates multiple encoded words if necessary. To be on the safe side, I
4870 have increased the buffer size for the ${rfc2047: expansion operator from
4873 33. It is now permitted to omit both strings after an "if" condition; if the
4874 condition is true, the result is "true". As before, when the second string
4875 is omitted, a false condition yields an empty string. This makes it less
4876 cumbersome to write custom ACL and router conditions.
4878 34. Failure to deliver a bounce message always caused it to be frozen, even if
4879 there was an errors_to setting on the router. The errors_to setting is now
4882 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4883 canonical form (fully expanded) before being placed in
4884 $sender_host_address.
4886 36. The table in the code that translates DNS record types into text (T_A to
4887 "A" for instance) was missing entries for NS and CNAME. It is just possible
4888 that this could have caused confusion if both these types were looked up
4889 for the same domain, because the text type is used as part of Exim's
4890 per-process caching. But the chance of anyone hitting this buglet seems
4893 37. The dnsdb lookup has been extended in a number of ways.
4895 (1) There is a new type, "zns", which walks up the domain tree until it
4896 finds some nameserver records. It should be used with care.
4898 (2) There is a new type, "mxh", which is like "mx" except that it returns
4899 just the host names, not the priorities.
4901 (3) It is now possible to give a list of domains (or IP addresses) to be
4902 looked up. The behaviour when one of the lookups defers can be
4903 controlled by a keyword.
4905 (4) It is now possible to specify the separator character for use when
4906 multiple records are returned.
4908 38. The dnslists ACL condition has been extended: it is now possible to supply
4909 a list of IP addresses and/or domains to be looked up in a particular DNS
4912 39. Added log_selector=+queue_time_overall.
4914 40. When running the queue in the test harness, wait just a tad after forking a
4915 delivery process, to get repeatability of debugging output.
4917 41. Include certificate and key file names in error message when GnuTLS fails
4918 to set them up, because the GnuTLS error message doesn't include the name
4919 of the failing file when there is a problem reading it.
4921 42. Allow both -bf and -bF in the same test run.
4923 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4925 44. The "Exiscan patch" is now merged into the mainline Exim source.
4927 45. Sometimes the final signoff response after QUIT could fail to get
4928 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4929 < 0 before doing a fflush(). This bug looks as though it goes back to the
4930 introduction of TLS in release 3.20, but "sometimes" must have been rare
4931 because the tests only now provoked it.
4933 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4934 (this can affect the format of dates).
4936 47. exim_tidydb, when checking for the continued existence of a message for
4937 which it has found a message-specific retry record, was not finding
4938 messages that were in split spool directories. Consequently, it was
4939 deleting retry records that should have stayed in existence.
4941 48. Steve fixed some bugs in eximstats.
4943 49. The SPA authentication driver was not abandoning authentication and moving
4944 on to the next authenticator when an expansion was forced to fail,
4945 contradicting the general specification for all authenticators. Instead it
4946 was generating a temporary error. It now behaves as specified.
4948 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4949 (the order specifies the preference for clients). The order is now AES256,
4950 AES128, 3DES, ARCFOUR128.
4952 51. Small patch to Sieve code - explicitly set From: when generating an
4955 52. Exim crashed if a remote delivery caused a very long error message to be
4956 recorded - for instance if somebody sent an entire SpamAssassin report back
4957 as a large number of 550 error lines. This bug was coincidentally fixed by
4958 increasing the size of one of Exim's internal buffers (big_buffer) that
4959 happened as part of the Exiscan merge. However, to be on the safe side, I
4960 have made the code more robust (and fixed the comments that describe what
4963 53. Now that there can be additional text after "Completed" in log lines (if
4964 the queue_time_overall log selector is set), a one-byte patch to exigrep
4965 was needed to allow it to recognize "Completed" as not the last thing in
4968 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4969 patch that reportedly fixes this has been added. I am not expert enough to
4970 create a test for it. This is what the patch creator wrote:
4972 "I found a little strange behaviour of ldap code when working with
4973 Windows 2003 AD Domain, where users was placed in more than one
4974 Organization Units. When I tried to give exim partial DN, the exit code
4975 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4976 But simultaneously result of request was absolutely normal ldap result,
4977 so I produce this patch..."
4979 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4980 REFERENCE, so I have modified the code to exclude the patch when that macro
4983 55. Some experimental protocols are using DNS PTR records for new purposes. The
4984 keys for these records are domain names, not reversed IP addresses. The
4985 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4986 leaves it alone. Component reversal etc. now happens only for IP addresses.
4989 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4991 57. Double the size of the debug message buffer (to 2048) so that more of very
4992 long debug lines gets shown.
4994 58. The exicyclog utility now does better if the number of log files to keep
4995 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4997 59. Two changes related to the smtp_active_hostname option:
4999 (1) $smtp_active_hostname is now available as a variable.
5000 (2) The default for smtp_banner uses $smtp_active_hostname instead
5001 of $primary_hostname.
5003 60. The host_aton() function is supposed to be passed a string that is known
5004 to be a valid IP address. However, in the case of IPv6 addresses, it was
5005 not checking this. This is a hostage to fortune. Exim now panics and dies
5006 if the condition is not met. A case was found where this could be provoked
5007 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5008 components; fortuitously, this particular loophole had already been fixed
5009 by change 4.50/55 above.
5011 If there are any other similar loopholes, the new check in host_aton()
5012 itself should stop them being exploited. The report I received stated that
5013 data on the command line could provoke the exploit when Exim was running as
5014 exim, but did not say which command line option was involved. All I could
5015 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5016 running as the user.
5019 61. There was a buffer overflow vulnerability in the SPA authentication code
5020 (which came originally from the Samba project). I have added a test to the
5021 spa_base64_to_bits() function which I hope fixes it.
5024 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5025 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5027 63. The daemon start-up calls getloadavg() while still root for those OS that
5028 need the first call to be done as root, but it missed one case: when
5029 deliver_queue_load_max is set with deliver_drop_privilege. This is
5030 necessary for the benefit of the queue runner, because there is no re-exec
5031 when deliver_drop_privilege is set.
5033 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5034 This has been fixed.
5036 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5037 was in use, was not putting the data itself into the right store pool;
5038 consequently, it could be overwritten for a subsequent message in the same
5039 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5042 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5044 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5045 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5046 the function were treating the return as a boolean value, which happened to
5047 work because 0=false and not-0=true, but is not correct code.
5049 68. The host_aton() function was not handling scoped IPv6 addresses (those
5050 with, for example, "%eth0" on the end) correctly.
5052 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5053 negated items (that is, ~something) in unsigned ints. Some compilers
5054 apparently mutter when there is no cast.
5056 70. If an address verification called from an ACL failed, and did not produce a
5057 user-specific message (i.e. there was only a "system" message), nothing was
5058 put in $acl_verify_message. In this situation, it now puts the system
5061 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5062 session; change 4.31/43 added the unwanted input to the log line - except
5063 that it did not do this in the start of session case. It now does.
5065 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5066 This is wrong and can cause the other end to generate a synchronization
5067 error if it is another Exim or anything else that does the synchronization
5068 check. A QUIT command is no longer sent after a timeout.
5070 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5071 during host lookups.
5073 74. The defer_ok option of callout verification was not working if it was used
5074 when verifying addresses in header lines, that is, for this case:
5076 verify = header_sender/callout=defer_ok
5078 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5079 those file descriptors could be used for SMTP connections. If anything
5080 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5081 could be sent to the SMTP client, causing chaos. The daemon now opens
5082 stdin, stdout, and stderr to /dev/null when it puts itself into the
5085 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5086 log by default. The user can override this with suitable Perl magic.
5088 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5089 the log message when discard triggers, was not working for the DATA ACL or
5090 for the non-SMTP ACL.
5092 78. Error message wording change in sieve.c.
5094 79. If smtp_accept_max_per_host was set, the number of connections could be
5095 restricted to fewer than expected, because the daemon was trying to set up
5096 a new connection before checking whether the processes handling previous
5097 connections had finished. The check for completed processes is now done
5098 earlier. On busy systems, this bug wouldn't be noticed because something
5099 else would have woken the daemon, and it would have reaped the completed
5102 80. If a message was submitted locally by a user whose login name contained one
5103 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5104 It caused a spool format error. I have fixed the spool reading code. A
5105 related problem was that the "from" clause in the Received: line became
5106 illegal because of the space(s). It is now covered by ${quote_local_part.
5108 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5111 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5113 83. Patch from Sieve maintainer to fix unterminated string problem in
5114 "vacation" handling.
5116 84. Some minor changes to the Linux configuration files to help with other
5117 OS variants using glibc.
5119 85. One more patch for Sieve to update vacation handling to latest spec.
5122 ----------------------------------------------------
5123 See the note above about the 4.44 and 4.50 releases.
5124 ----------------------------------------------------
5130 1. Change 4.43/35 introduced a bug that caused file counts to be
5131 incorrectly computed when quota_filecount was set in an appendfile
5134 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5135 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5138 3. Give more explanation in the error message when the command for a transport
5139 filter fails to execute.
5141 4. There are several places where Exim runs a non-Exim command in a
5142 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5143 was being done only for the command run by the queryprogram router. It is
5144 now done for all such subprocesses. The other cases are: ${run, transport
5145 filters, and the commands run by the lmtp and pipe transports.
5147 5. Some older OS have a limit of 256 on the maximum number of file
5148 descriptors. Exim was using setrlimit() to set 1000 as a large value
5149 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5150 systems. I've change it so that if it can't get 1000, it tries for 256.
5152 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5153 was an oversight, and furthermore, ever since the addition of extra
5154 controls (e.g. 4.43/32), the checks on when to allow different forms of
5155 "control" were broken. There should now be diagnostics for all cases when a
5156 control that does not make sense is encountered.
5158 7. $recipients is now available in the predata ACL (oversight).
5160 8. Tidy the search cache before the fork to do a delivery from a message
5161 received from the command line. Otherwise the child will trigger a lookup
5162 failure and thereby defer the delivery if it tries to use (for example) a
5163 cached ldap connection that the parent has called unbind on.
5165 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5166 of $address_data from the recipient verification was clobbered by the
5167 sender verification.
5169 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5170 was its contents. (It was OK if the option was not defined at all.)
5172 11. A "Completed" log line is now written for messages that are removed from
5173 the spool by the -Mrm option.
5175 12. $host_address is now set to the target address during the checking of
5176 ignore_target_hosts.
5178 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5179 being passed; this would have caused $sender_host_name to have been used if
5180 matching the list had actually called for a host name (not very likely,
5181 since this list is usually IP addresses). A host name is now passed as
5184 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5185 code that specifies a non-restarting handler (typically sigaction() in
5186 modern systems) in an attempt to fix a rare and obscure crash bug.
5188 15. Narrowed the window for a race in the daemon that could cause it to ignore
5189 SIGCHLD signals. This is not a major problem, because they are used only to
5190 wake it up if nothing else does.
5192 16. A malformed maildirsize file could cause Exim to calculate negative values
5193 for the mailbox size or file count. Odd effects could occur as a result.
5194 The maildirsize information is now recalculated if the size or filecount
5197 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5198 support for a long time. Removed HAVE_SYS_VFS_H.
5200 18. Updated exipick to current release from John Jetmore.
5202 19. Allow an empty sender to be matched against a lookup in an address list.
5203 Previously the only cases considered were a regular expression, or an
5206 20. Exim went into a mad DNS lookup loop when doing a callout where the
5207 host was specified on the transport, if the DNS lookup yielded more than
5210 21. The RFC2047 encoding function was originally intended for short strings
5211 such as real names; it was not keeping to the 75-character limit for
5212 encoded words that the RFC imposes. It now respects the limit, and
5213 generates multiple encoded words if necessary. To be on the safe side, I
5214 have increased the buffer size for the ${rfc2047: expansion operator from
5217 22. Failure to deliver a bounce message always caused it to be frozen, even if
5218 there was an errors_to setting on the router. The errors_to setting is now
5221 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5222 canonical form (fully expanded) before being placed in
5223 $sender_host_address.
5225 24. Updated eximstats to version 1.33
5227 25. Include certificate and key file names in error message when GnuTLS fails
5228 to set them up, because the GnuTLS error message doesn't include the name
5229 of the failing file when there is a problem reading it.
5231 26. Expand error message when OpenSSL has problems setting up cert/key files.
5234 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5235 (this can affect the format of dates).
5237 28. exim_tidydb, when checking for the continued existence of a message for
5238 which it has found a message-specific retry record, was not finding
5239 messages that were in split spool directories. Consequently, it was
5240 deleting retry records that should have stayed in existence.
5242 29. eximstats updated to version 1.35
5243 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5244 1.35 - bugfix such that pie charts by volume are generated correctly
5246 30. The SPA authentication driver was not abandoning authentication and moving
5247 on to the next authenticator when an expansion was forced to fail,
5248 contradicting the general specification for all authenticators. Instead it
5249 was generating a temporary error. It now behaves as specified.
5251 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5252 (the order specifies the preference for clients). The order is now AES256,
5253 AES128, 3DES, ARCFOUR128.
5255 31. Small patch to Sieve code - explicitly set From: when generating an
5258 32. Exim crashed if a remote delivery caused a very long error message to be
5259 recorded - for instance if somebody sent an entire SpamAssassin report back
5260 as a large number of 550 error lines. This bug was coincidentally fixed by
5261 increasing the size of one of Exim's internal buffers (big_buffer) that
5262 happened as part of the Exiscan merge. However, to be on the safe side, I
5263 have made the code more robust (and fixed the comments that describe what
5266 33. Some experimental protocols are using DNS PTR records for new purposes. The
5267 keys for these records are domain names, not reversed IP addresses. The
5268 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5269 leaves it alone. Component reversal etc. now happens only for IP addresses.
5272 34. The host_aton() function is supposed to be passed a string that is known
5273 to be a valid IP address. However, in the case of IPv6 addresses, it was
5274 not checking this. This is a hostage to fortune. Exim now panics and dies
5275 if the condition is not met. A case was found where this could be provoked
5276 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5277 components; fortuitously, this particular loophole had already been fixed
5278 by change 4.50/55 or 4.44/33 above.
5280 If there are any other similar loopholes, the new check in host_aton()
5281 itself should stop them being exploited. The report I received stated that
5282 data on the command line could provoke the exploit when Exim was running as
5283 exim, but did not say which command line option was involved. All I could
5284 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5285 running as the user.
5288 35. There was a buffer overflow vulnerability in the SPA authentication code
5289 (which came originally from the Samba project). I have added a test to the
5290 spa_base64_to_bits() function which I hope fixes it.
5293 36. The daemon start-up calls getloadavg() while still root for those OS that
5294 need the first call to be done as root, but it missed one case: when
5295 deliver_queue_load_max is set with deliver_drop_privilege. This is
5296 necessary for the benefit of the queue runner, because there is no re-exec
5297 when deliver_drop_privilege is set.
5299 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5300 was in use, was not putting the data itself into the right store pool;
5301 consequently, it could be overwritten for a subsequent message in the same
5302 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5305 38. Sometimes the final signoff response after QUIT could fail to get
5306 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5307 < 0 before doing a fflush(). This bug looks as though it goes back to the
5308 introduction of TLS in release 3.20, but "sometimes" must have been rare
5309 because the tests only now provoked it.
5315 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5316 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5317 Now it returns an integer. A number of calls were still expecting a T/F
5318 return. Fortuitously, in all cases, the tests worked in OK situations,
5319 which is the norm. However, things would have gone wrong on any write
5320 failures on the smtp file descriptor. This function is used when sending
5321 messages over SMTP and also when doing verify callouts.
5323 2. When Exim is called to do synchronous delivery of a locally submitted
5324 message (the -odf or -odi options), it no longer closes stderr before doing
5327 3. Implemented the mua_wrapper option.
5329 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5331 5. Implemented the functions header_remove(), header_testname(),
5332 header_add_at_position(), and receive_remove_recipient(), and exported them
5335 6. If an ACL "warn" statement specified the addition of headers, Exim already
5336 inserted X-ACL-Warn: at the start if there was no header name. However, it
5337 was not making this test for the second and subsequent header lines if
5338 there were newlines in the string. This meant that an invalid header could
5339 be inserted if Exim was badly configured.
5341 7. Allow an ACL "warn" statement to add header lines at the start or after all
5342 the Received: headers, as well as at the end.
5344 8. Added the rcpt_4xx retry error code.
5346 9. Added postmaster_mailfrom=xxx to callout verification option.
5348 10. Added mailfrom=xxxx to the callout verification option, for verify=
5351 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5352 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5354 12. Inserted some casts to stop certain compilers warning when using pointer
5355 differences as field lengths or precisions in printf-type calls (mostly
5356 affecting debugging statements).
5358 13. Added optional readline() support for -be (dynamically loaded).
5360 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5361 same clock tick as a message's arrival, so that its received time was the
5362 same as the "first fail" time on the retry record, and that message
5363 remained on the queue past the ultimate address timeout, every queue runner
5364 would try a delivery (because it was past the ultimate address timeout) but
5365 after another failure, the ultimate address timeout, which should have then
5366 bounced the address, did not kick in. This was a "< instead of <=" error;
5367 in most cases the first failure would have been in the next clock tick
5368 after the received time, and all would be well.
5370 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5371 being recognized when the domain list was tested by the match_domain
5372 condition in an expansion string.
5374 16. Added the ${str2b64: operator.
5376 17. Exim was always calling setrlimit() to set a large limit for the number of
5377 processes, without checking whether the existing limit was already
5378 adequate. (It did check for the limit on file descriptors.) Furthermore,
5379 errors from getrlimit() and setrlimit() were being ignored. Now they are
5380 logged to the main and panic logs, but Exim does carry on, to try to do its
5381 job under whatever limits there are.
5383 18. Imported PCRE 5.0.
5385 19. Trivial typo in log message " temporarily refused connection" (the leading
5388 20. If the log selector return_path_on_delivery was set and an address was
5389 redirected to /dev/null, the delivery process crashed because it assumed
5390 that a return path would always be set for a "successful" delivery. In this
5391 case, the whole delivery is bypassed as an optimization, and therefore no
5394 21. Internal re-arrangement: the function for sending a challenge and reading
5395 a response while authentication was assuming a zero-terminated challenge
5396 string. It's now changed to take a pointer and a length, to allow for
5397 binary data in such strings.
5399 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5401 23. Exim was not respecting finduser_retries when seeking the login of the
5402 uid under which it was called; it was always trying 10 times. (The default
5403 setting of finduser_retries is zero.) Also, it was sleeping after the final
5404 failure, which is pointless.
5406 24. Implemented tls_on_connect_ports.
5408 25. Implemented acl_smtp_predata.
5410 26. If the domain in control=submission is set empty, Exim assumes that the
5411 authenticated id is a complete email address when it generates From: or
5412 Sender: header lines.
5414 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5415 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5416 chown and chgrp in /bin and hostname in /usr/bin.
5418 28. Exim was keeping the "process log" file open after each use, just as it
5419 does for the main log. This opens the possibility of it remaining open for
5420 long periods when the USR1 signal hits a daemon. Occasional processlog
5421 errors were reported, that could have been caused by this. Anyway, it seems
5422 much more sensible not to leave this file open at all, so that is what now
5425 29. The long-running daemon process does not normally write to the log once it
5426 has entered its main loop, and it closes the log before doing so. This is
5427 so that log files can straightforwardly be renamed and moved. However,
5428 there are a couple of unusual error situations where the daemon does write
5429 log entries, and I had neglected to close the log afterwards.
5431 30. The text of an SMTP error response that was received during a remote
5432 delivery was being truncated at 512 bytes. This is too short for some of
5433 the long messages that one sometimes sees. I've increased the limit to
5436 31. It is now possible to make retry rules that apply only when a message has a
5437 specific sender, in particular, an empty sender.
5439 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5440 it possible to be selective about when SMTP synchronization is enforced.
5442 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5444 32. Implemented hosts_connection_nolog.
5446 33. Added an ACL for QUIT.
5448 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5451 35. Added mailbox_size and mailbox_filecount to appendfile.
5453 36. Added control = no_multiline_responses to ACLs.
5455 37. There was a bug in the logic of the code that waits for the clock to tick
5456 in the case where the clock went backwards by a substantial amount such
5457 that the microsecond fraction of "now" was more than the microsecond
5458 fraction of "then" (but the whole seconds number was less).
5460 38. Added support for the libradius Radius client library this is found on
5461 FreeBSD (previously only the radiusclient library was supported).
5467 1. When certain lookups returned multiple values in the form name=value, the
5468 quoting of the values was not always being done properly. Specifically:
5469 (a) If the value started with a double quote, but contained no whitespace,
5471 (b) If the value contained whitespace other than a space character (i.e.
5472 tabs or newlines or carriage returns) it was not quoted.
5473 This fix has been applied to the mysql and pgsql lookups by writing a
5474 separate quoting function and calling it from the lookup code. The fix
5475 should probably also be applied to nisplus, ibase and oracle lookups, but
5476 since I cannot test any of those, I have not disturbed their existing code.
5478 2. A hit in the callout cache for a specific address caused a log line with no
5479 reason for rejecting RCPT. Now it says "Previous (cached) callout
5480 verification failure".
5482 3. There was an off-by-one bug in the queryprogram router. An over-long
5483 return line was truncated at 256 instead of 255 characters, thereby
5484 overflowing its buffer with the terminating zero. As well as fixing this, I
5485 have increased the buffer size to 1024 (and made a note to document this).
5487 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5488 when Exim is waiting for an SMTP response from a remote server, Exim
5489 restarts its select() call on the socket, thereby resetting its timeout.
5490 This is not a problem when such interrupts are rare. Somebody set up a cron
5491 job to run exiwhat every 2 minutes, which is less than the normal select()
5492 timeout (5 or 10 minutes). This meant that the select() timeout never
5493 kicked in because it was always reset. I have fixed this by comparing the
5494 time when an interrupt arrives with the time at the start of the first call
5495 to select(). If more time than the timeout has elapsed, the interrupt is
5496 treated as a timeout.
5498 5. Some internal re-factoring in preparation for the addition of Sieve
5499 extensions (by MH). In particular, the "personal" test is moved to a
5500 separate function, and given an option for scanning Cc: and Bcc: (which is
5501 not set for Exim filters).
5503 6. When Exim created an email address using the login of the caller as the
5504 local part (e.g. when creating a From: or Sender: header line), it was not
5505 quoting the local part when it contained special characters such as @.
5507 7. Installed new OpenBSD configuration files.
5509 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5510 try to make them clearer.
5512 9. Callout options, other than the timeout value, were being ignored when
5513 verifying sender addresses in header lines. For example, when using
5515 verify = header_sender/callout=no_cache
5517 the cache was (incorrectly) being used.
5519 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5520 only the Cygwin environment.
5522 11. When return_path_on_delivery was set as a log selector, if different remote
5523 addresses in the same message used different return paths and parallel
5524 remote delivery occurred, the wrong values would sometimes be logged.
5525 (Whenever a remote delivery process finished, the return path value from
5526 the most recently started remote delivery process was logged.)
5528 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5529 lines when AUTH and/or TLS are in use. This is the "received protocol"
5530 field. Exim used to use "asmtp" for authenticated SMTP, without any
5531 indication (in the protocol name) for TLS use. Now it follows the RFC and
5532 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5533 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5534 names appear in log lines as well as in Received: header lines.
5536 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5537 extensions, and comparison tests, and to fix some bugs.
5539 14. Changes to the "personal" filter test:
5541 (1) The test was buggy in that it was just doing the equivalent of
5542 "contains" tests on header lines. For example, if a user's address was
5543 anne@some.where, the "personal" test would incorrectly be true for
5545 To: susanne@some.where
5547 This test is now done by extracting each address from the header in turn,
5548 and checking the entire address. Other tests that are part of "personal"
5549 are now done using regular expressions (for example, to check local parts
5550 of addresses in From: header lines).
5552 (2) The list of non-personal local parts in From: addresses has been
5553 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5554 taken from the Sieve specification recommendations.
5556 (3) If the message contains any header line starting with "List-" it is
5557 treated as non-personal.
5559 (4) The test for "circular" in the Subject: header line has been removed
5560 because it now seems ill-conceived.
5562 15. Minor typos in src/EDITME comments corrected.
5564 16. Installed latest exipick from John Jetmore.
5566 17. If headers_add on a router specified a text string that was too long for
5567 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5568 of string_sprintf() is now avoided.
5570 18. $message_body_size was not set (it was always zero) when running the DATA
5571 ACL and the local_scan() function.
5573 19. For the "mail" command in an Exim filter, no default was being set for
5574 the once_repeat time, causing a random time value to be used if "once" was
5575 specified. (If the value happened to be <= 0, no repeat happened.) The
5576 default is now 0s, meaning "never repeat". The "vacation" command was OK
5577 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5578 (I found it when inspecting the code).
5580 20. There is now an overall timeout for performing a callout verification. It
5581 defaults to 4 times the callout timeout, which applies to individual SMTP
5582 commands during the callout. The overall timeout applies when there is more
5583 than one host that can be tried. The timeout is checked before trying the
5584 next host. This prevents very long delays if there are a large number of
5585 hosts and all are timing out (e.g. when the network connections are timing
5586 out). The value of the overall timeout can be changed by specifying an
5587 additional sub-option for "callout", called "maxwait". For example:
5589 verify = sender/callout=5s,maxwait=20s
5591 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5592 to the end before writing, but this should make it even safer).
5594 22. Exim was forgetting that it had advertised PIPELINING for the second and
5595 subsequent messages on an SMTP connection. It was also not resetting its
5596 memory on STARTTLS and an internal HELO.
5598 23. When Exim logs an SMTP synchronization error within a session, it now
5599 records whether PIPELINING has been advertised or not.
5601 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5602 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5603 rather than long int.
5605 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5607 26. Added the never_mail option to autoreply.
5613 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5614 crash if the getsockname() call failed; this can happen if a connection is
5615 closed very soon after it is established. The problem was simply in the
5616 order in which certain operations were done, causing Exim to try to write
5617 to the SMTP stream before it had set up the file descriptor. The bug has
5618 been fixed by making things happen in the correct order.
5624 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5625 before the connection was closed, thus losing the rejection response.
5627 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5628 some early Solaris releases, but causes trouble in current releases where
5629 socklen_t is defined.
5631 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5634 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5637 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5638 that releasing the top of it at the end releases what was used for sub-
5639 expansions (unless the block got too big). However, discard this block if
5640 the first thing is a variable or header, so that we can use its block when
5641 it is dynamic (useful for very large $message_headers, for example).
5643 6. Lookups now cache *every* query, not just the most recent. A new, separate
5644 store pool is used for this. It can be recovered when all lookup caches are
5645 flushed. Lookups now release memory at the end of their result strings.
5646 This has involved some general refactoring of the lookup sources.
5648 7. Some code has been added to the store_xxx() functions to reduce the amount
5649 of flapping under certain conditions.
5651 8. log_incoming_interface used to affect only the <= reception log lines. Now
5652 it causes the local interface and port to be added to several more SMTP log
5653 lines, for example "SMTP connection from", and rejection lines.
5655 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5657 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5659 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5660 could be overwritten at the end of the current message (or the start of a
5661 new message if it was set in a HELO ACL). The value is now preserved for
5662 the duration of the SMTP connection.
5664 12. If a transport had a headers_rewrite setting, and a matching header line
5665 contained an unqualified address, that address was qualified, even if it
5666 did not match any rewriting rules. The underlying bug was that the values
5667 of the flags that permit the existence of unqualified sender and recipient
5668 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5669 non-local messages, and by -bnq for local messages) were not being
5670 preserved with the message after it was received.
5672 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5673 "next input=" as part of the text comprising the host identity instead of
5674 the correct text. The code was using the same buffer for two different
5675 strings. However, depending on which order the printing function evaluated
5676 its arguments, the bug did not always show up. Under Linux, for example, my
5677 test suite worked just fine.
5679 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5680 doesn't work with some older versions of Perl. It has been changed to "my",
5681 which in any case is probably the better facility to use.
5683 15. A really picky compiler found some instances of statements for creating
5684 error messages that either had too many or two few arguments for the format
5687 16. The size of the buffer for calls to the DNS resolver has been increased
5688 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5689 for addresses that have a lot of PTR records. This alleviates a problem; it
5690 does not fully solve it.
5692 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5693 buffer now truncates the list and logs the incident, which is the same
5694 action as happens when Exim is looking up a host name and its aliases.
5695 Previously in this situation something unpredictable would happen;
5696 sometimes it was "internal error: store_reset failed".
5698 18. If a server dropped the connection unexpectedly when an Exim client was
5699 using GnuTLS and trying to read a response, the client delivery process
5700 crashed while trying to generate an error log message.
5702 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5703 string, for example:
5705 warn message = H1: something\nH2: something
5707 the text was added as a single header line from Exim's point of view
5708 though it ended up OK in the delivered message. However, searching for the
5709 second and subsequent header lines using $h_h2: did not work. This has been
5710 fixed. Similarly, if a system filter added multiple headers in this way,
5711 the routers could not see them.
5713 20. Expanded the error message when iplsearch is called with an invalid key to
5714 suggest using net-iplsearch in a host list.
5716 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5717 ACLs are no longer actually imposed (and a message to that effect is
5720 22. If a "gecos" field in a passwd entry contained escaped characters, in
5721 particular, if it contained a \" sequence, Exim got it wrong when building
5722 a From: or a Sender: header from that name. A second bug also caused
5723 incorrect handling when an unquoted " was present following a character
5724 that needed quoting.
5726 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5727 was not being matched caselessly.
5729 24. Arranged for all hyphens in the exim.8 source to be escaped with
5732 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5733 properly. Recipient callout cache records were still being keyed to include
5734 the sender, even when use_sender was set false. This led to far more
5735 callouts that were necessary. The sender is no longer included in the key
5736 when use_sender is false.
5738 26. Added "control = submission" modifier to ACLs.
5740 27. Added the ${base62d: operator to decode base 62 numbers.
5742 28. dnsdb lookups can now access SRV records.
5744 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5745 the configuration file.
5747 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5748 (-v) mode. This makes the output for a verbose queue run more intelligible.
5750 31. Added a use_postmaster feature to recipient callouts.
5752 32. Added the $body_zerocount variable, containing the number of binary zero
5753 bytes in the message body.
5755 33. The time of last modification of the "new" subdirectory is now used as the
5756 "mailbox time last read" when there is a quota error for a maildir
5759 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5761 35. Added +ignore_unknown as a special item in host lists.
5763 36. Code for decoding IPv6 addresses in host lists is now included, even if
5764 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5765 address was recognized as an IP address, but was then not correctly decoded
5766 into binary, causing unexpected and incorrect effects when compared with
5773 1. Very minor rewording of debugging text in manualroute to say "list of
5774 hosts" instead of "hostlist".
5776 2. If verify=header_syntax was set, and a header line with an unqualified
5777 address (no domain) and a large number of spaces between the end of the
5778 name and the colon was received, the reception process suffered a buffer
5779 overflow, and (when I tested it) crashed. This was caused by some obsolete
5780 code that should have been removed. The fix is to remove it!
5782 3. When running in the test harness, delay a bit after writing a bounce
5783 message to get a bit more predictability in the log output.
5785 4. Added a call to search_tidyup() just before forking a reception process. In
5786 theory, someone could use a lookup in the expansion of smtp_accept_max_
5787 per_host which, without the tidyup, could leave open a database connection.
5789 5. Added the variables $recipient_data and $sender_data which get set from a
5790 lookup success in an ACL "recipients" or "senders" condition, or a router
5791 "senders" option, similar to $domain_data and $local_part_data.
5793 6. Moved the writing of debug_print from before to after the "senders" test
5796 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5797 problems for message scanning, either using a data ACL, or using
5798 local_scan() because the Received: header was not generated till after they
5799 were called (in order to set the time as the time of reception completion).
5800 I have revised the way this works. The header is now generated after the
5801 body is received, but before the ACL or local_scan() are called. After they
5802 are run, the timestamp in the header is updated.
5808 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5809 before starting a queue runner without re-exec. This happened only when
5810 deliver_drop_privilege was set or when the Exim user was set to root. The
5811 effect of the bug was that timeouts during subsequent deliveries caused
5812 crashes instead of being properly handled. The handler is now left at its
5813 default (and expected) setting.
5815 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5816 message, again when deliver_drop_privilege is set or Exim is run as root.
5817 The bug described in (1) was not present in this case, but the tidying up
5818 of the other signals was missing. I have made the two cases consistent.
5820 3. The ignore_target_hosts setting on a manualroute router was being ignored
5821 for hosts that were looked up using the /MX notation.
5823 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5826 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5827 operated on the sender address. After changing the $sender_address to <>
5828 for the sender address verify, Exim was re-instated it as the original
5829 (before rewriting) address, but remembering that it had rewritten it, so it
5830 wasn't rewriting it again. This bug also had the effect of breaking the
5831 sender address verification caching when the sender address was rewritten.
5833 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5834 This has been changed so that if the ip literal address matches
5835 ignore_target_hosts, the router declines.
5837 7. Added expansion conditions match_domain, match_address, and match_local_
5838 part (NOT match_host).
5840 8. The placeholder for the Received: header didn't have a length field set.
5842 9. Added code to Exim itself and to exim_lock to test for a specific race
5843 condition that could lead to file corruption when using MBX delivery. The
5844 issue is with the lockfile that is created in /tmp. If this file is removed
5845 after a process has opened it but before that process has acquired a lock,
5846 there is the potential for a second process to recreate the file and also
5847 acquire a lock. This could lead to two Exim processes writing to the file
5848 at the same time. The added code performs the same test as UW imapd; it
5849 checks after acquiring the lock that its file descriptor still refers to
5850 the same named file.
5852 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5853 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5856 11. Added the smtp_active_hostname option. If used, this will typically be made
5857 to depend on the incoming interface address. Because $interface_address is
5858 not set up until the daemon has forked a reception process, error responses
5859 that can happen earlier (such as "too many connections") no longer contain
5862 12. If an expansion in a condition on a "warn" statement fails because a lookup
5863 defers, the "warn" statement is abandoned, and the next ACL statement is
5864 processed. Previously this caused the whole ACL to be aborted.
5866 13. Added the iplsearch lookup type.
5868 14. Added ident_timeout as a log selector.
5870 15. Added tls_certificate_verified as a log selector.
5872 16. Added a global option tls_require_ciphers (compare the smtp transport
5873 option of the same name). This controls incoming TLS connections.
5875 17. I finally figured out how to make tls_require_ciphers do a similar thing
5876 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5877 before starting the TLS session.
5879 18. Tabs are now shown as \t in -bP output.
5881 19. If the log selector return_path_on_delivery was set, Exim crashed when
5882 bouncing a message because it had too many Received: header lines.
5884 20. If two routers both had headers_remove settings, and the first one included
5885 a superfluous trailing colon, the final name in the first list and the
5886 first name in the second list were incorrectly joined into one item (with a
5887 colon in the middle).
5893 1. Added -C and -D options to the exinext utility, mainly to make it easier
5894 to include in the automated testing, but these could be helpful when
5895 multiple configurations are in use.
5897 2. The exinext utility was not formatting the output nicely when there was
5898 an alternate port involved in the retry record key, nor when there was a
5899 message id as well (for retries that were specific to a specific message
5900 and a specific host). It was also confused by IPv6 addresses, because of
5901 the additional colons they contain. I have fixed the IPv4 problem, and
5902 patched it up to do a reasonable job for IPv6.
5904 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5905 delivery, the log line now contains "pipelined" if PIPELINING was used.
5907 4. An SMTP transport process used to panic and die if the bind() call to set
5908 an explicit outgoing interface failed. This has been changed; it is now
5909 treated in the same way as a connect() failure.
5911 5. A reference to $sender_host_name in the part of a conditional expansion
5912 that was being skipped was still causing a DNS lookup. This no longer
5915 6. The def: expansion condition was not recognizing references to header lines
5916 that used bh_ and bheader_.
5918 7. Added the _cache feature to named lists.
5920 8. The code for checking quota_filecount in the appendfile transport was
5921 allowing one more file than it should have been.
5923 9. For compatibility with Sendmail, the command line option
5931 and sets the incoming protocol and host name (for trusted callers). The
5932 host name and its colon can be omitted when only the protocol is to be set.
5933 Note the Exim already has two private options, -pd and -ps, that refer to
5934 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5935 "s", but I don't think that's a major issue.
5937 10. A number of refactoring changes to the code, none of which should affect
5940 (a) The number of logging options was getting close to filling up the
5941 32-bit word that was used as a bit map. I have split them into two classes:
5942 those that are passed in the argument to log_write(), and those that are
5943 only ever tested independently outside of that function. These are now in
5944 separate 32-bit words, so there is plenty of room for expansion again.
5945 There is no change in the user interface or the logging behaviour.
5947 (b) When building, for example, log lines, the code previously used a
5948 macro that called string_cat() twice, in order to add two strings. This is
5949 not really sufficiently general. Furthermore, there was one instance where
5950 it was actually wrong because one of the argument was used twice, and in
5951 one call a function was used. (As it happened, calling the function twice
5952 did not affect the overall behaviour.) The macro has been replaced by a
5953 function that can join an arbitrary number of extra strings onto a growing
5956 (c) The code for expansion conditions now uses a table and a binary chop
5957 instead of a serial search (which was left over from when there were very
5958 few conditions). Also, it now recognizes conditions like "pam" even when
5959 the relevant support is not compiled in: a suitably worded error message is
5960 given if an attempt is made to use such a condition.
5962 11. Added ${time_interval:xxxxx}.
5964 12. A bug was causing one of the ddress fields not to be passed back correctly
5965 from remote delivery subprocesses. The field in question was not being
5966 subsequently used, so this caused to problems in practice.
5968 13. Added new log selectors queue_time and deliver_time.
5970 14. Might have fixed a bug in maildirsizefile handling that threw up
5971 "unexpected character" debug warnings, and recalculated the data
5972 unnecessarily. In any case, I expanded the warning message to give more
5975 15. Added the message "Restricted characters in address" to the statements in
5976 the default ACL that block characters like @ and % in local parts.
5978 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5979 Three changes have been made:
5981 (a) There was a serious bug; a negative response to MAIL caused the whole
5982 recipient domain to be cached as invalid, thereby blocking all messages
5983 to all local parts at the same domain, from all senders. This bug has
5984 been fixed. The domain is no longer cached after a negative response to
5985 MAIL if the sender used is not empty.
5987 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5990 (c) A new callout option, "use_sender" has been added for people who want
5991 the modified behaviour.
5997 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6000 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6001 indeed breaks things for older releases.
6003 3. Added additional logging to the case where there is a problem reading data
6004 from a filter that is running in a subprocess using a pipe, in order to
6005 try to track down a specific problem.
6007 4. Testing facility fudge: when running in the test harness and attempting
6008 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6009 getting "No route to host". Convert this to a timeout.
6011 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6014 6. Some OS don't have socklen_t but use size_t instead. This affects the
6015 fifth argument of getsockopt() amongst other things. This is now
6016 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6017 can be set for individual OS. I have set it for SunOS5, OSF1, and
6018 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6019 some earlier ones do not.
6021 7. Change 4.30/15 was not doing the test caselessly.
6023 8. The standard form for an IPv6 address literal was being rejected by address
6024 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6025 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6026 this, as well as the form without the "IPv6" on the front (but only when
6027 address literals are enabled, of course).
6029 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6031 10. Exim crashed if a message with an empty sender address specified by -f
6032 encountered a router with an errors_to setting. This could be provoked only
6033 by a command such as
6037 where an empty string was supplied; "<>" did not hit this bug.
6039 11. Installed PCRE release 4.5.
6041 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6042 remained set. It is now erased.
6044 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6045 times from message ids (which are base 36 rather than the normal 62).
6047 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6048 were being counted as actual protocol errors, and logged if the log
6049 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6050 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6051 and DATA following a set of rejected RCPTs do not count as protocol errors.
6052 In other words, Exim assumes they were pipelined, though this may not
6053 actually be the case. Of course, in all cases the client gets an
6054 appropriate error code.
6056 15. If a lookup fails in an ACL condition, a message about the failure may
6057 be available; it is used if testing the ACL cannot continue, because most
6058 such messages specify what the cause of the deferral is. However, some
6059 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6060 that caused an old message to be retained and used if a later statement
6061 caused a defer, replacing the real cause of the deferral.
6063 16. If an IP address had so many PTR records that the DNS lookup buffer
6064 was not large enough to hold them, Exim could crash while trying to process
6065 the truncated data. It now detects and logs this case.
6067 17. Further to 4.21/58, another change has been made: if (and only if) the
6068 first line of a message (the first header line) ends with CRLF, a bare LF
6069 in a subsequent header line has a space inserted after it, so as not to
6070 terminate the header.
6072 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6073 unnecessarily, used atoi() instead of strtol(), and didn't check the
6074 termination when getting file sizes from file names by regex.
6076 19. Completely re-implemented the support for maildirsize files, in the light
6077 of a number of problems with the previous contributed implementation
6078 (4.30/29). In particular:
6080 . If the quota is zero, the maildirsize file is maintained, but no quota is
6083 . If the maildir directory does not exist, it is created before any attempt
6084 to write a maildirsize file.
6086 . The quota value in the file is just a cache; if the quota is changed in
6087 the transport, the new value overrides.
6089 . A regular expression is available for excluding directories from the
6092 20. The autoreply transport checks the characters in options that define the
6093 message's headers; it allows continued headers, but it was checking with
6094 isspace() after an embedded newline instead of explicitly looking for a
6097 21. If all the "regular" hosts to which an address was routed had passed their
6098 expiry times, and had not reached their retry times, the address was
6099 bounced, even if fallback hosts were defined. Now Exim should go on to try
6102 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6103 equivalent code in the SMTP transport. Some hosts send humungous responses
6104 to HELO/EHLO, more than 1024 it seems.
6106 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6107 gives compiler warnings in some environments. I've now done it "properly",
6110 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6111 (because of problems with the built-in one) was declared to return uschar *
6112 instead of char *, causing compiler failure.
6114 25. Fixed a file descriptor leak when processing alias/forward files.
6116 26. Fixed a minor format string issue in dbfn.c.
6118 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6120 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6121 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6124 29. When a sender address is verified, it is cached, to save repeating the test
6125 when there is more than one recipient in a message. However, when the
6126 verification involves a callout, it is possible for different callout
6127 options to be set for different recipients. It is too complicated to keep
6128 track of this in the cache, so now Exim always runs a verification when a
6129 callout is required, relying on the callout cache for the optimization.
6130 The overhead is duplication of the address routing, but this should not be
6133 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6134 to be verified with callout=postmaster, and the main callout worked but the
6135 postmaster check failed, the verification correctly failed. However, if a
6136 subsequent RCPT command asked for sender verification *without* the
6137 postmaster check, incorrect caching caused this verification also to fail,
6140 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6141 it was not caching the DNS options (qualify_single, search_parents) that
6142 were used when the lookup failed. A subsequent lookup with different
6143 options therefore always gave the same answer, though there were cases
6144 where it should not have. (Example: a "domains = !$mx_any" option on a
6145 dnslookup router: the "domains" option is always processed without any
6146 widening, but the router might have qualify_single set.) Now Exim uses the
6147 cached value only when the same options are set.
6149 32. Added John Jetmore's "exipick" utility to the distribution.
6151 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6152 than a timeout (e.g. a certificate is required, and is not provided), an
6153 Exim server now closes the connection immediately. Previously it waited for
6154 the client to close - but if the client is SSL, it seems that they each
6155 wait for each other, leading to a delay before one of them times out.
6157 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6158 maintained 0.8.x compatibility because I don't think many are using it, and
6159 it is clearly obsolete.
6161 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6164 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6165 client certificate was expired. A simple patch fixes this, though I don't
6166 understand the full logic of why the verify callback is called multiple
6169 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6170 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6171 which causes problems with some clients (such as the Certicom SSL Plus
6172 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6173 disables the coutermeasure allowing Eudora to connect."
6175 38. Exim was not checking that a write() to a log file succeeded. This could
6176 lead to Bad Things if a log got too big, in particular if it hit a file
6177 size limit. Exim now panics and dies if it cannot write to a log file, just
6178 as it does if it cannot open a log file.
6180 39. Modified OS/Makefile-Linux so that it now contains
6182 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6184 The two -D definitions ensure that Exim is compiled with large file
6185 support, which makes it possible to handle log files that are bigger than
6188 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6189 instance) a domain was checked against a named list that involved a lookup,
6190 causing $domain_data to be set, then another domain was checked against the
6191 same list, then the first domain was re-checked, the value of $domain_data
6192 after the final check could be wrong. In particular, if the second check
6193 failed, it could be set empty. This bug probably also applied to
6196 41. The strip_trailing_dot option was not being applied to the address given
6197 with the -f command-line option.
6199 42. The code for reading a message's header from the spool was incrementing
6200 $received_count, but never initializing it. This meant that the value was
6201 incorrect (doubled) while delivering a message in the same process in which
6202 it was received. In the most common configuration of Exim, this never
6203 happens - a fresh exec is done - but it can happen when
6204 deliver_drop_privilege is set.
6206 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6207 it now includes up to 150 characters of the unexpected data in the log
6210 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6211 and building data strings. The size of both of these buffers was 10 000
6212 bytes - far larger than anybody would *ever* want, thought I. Needless to
6213 say, somebody hit the limit. I have increased the maximum line length to
6214 20 000 and the maximum data length of concatenated lines to 100 000. I have
6215 also fixed two bugs, because there was no checking on these buffers. Tsk,
6216 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6217 buffer is too small.
6219 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6220 lsearch lookups. Now it does.
6222 46. When parsing a route_list item in a manualroute router, a fixed-length
6223 buffer was used for the list of hosts. I made this 1024 bytes long,
6224 thinking that nobody would ever have a list of hosts that long. Wrong.
6225 Somebody had a whole pile of complicated expansion conditions, and the
6226 string was silently truncated, leading to an expansion error. It turns out
6227 that it is easier to change to an unlimited length (owing to other changes
6228 that have happened since this code was originally written) than to build
6229 structure for giving a limitation error. The length of the item that
6230 expands into the list of hosts is now unlimited.
6232 47. The lsearch lookup could not handle data where the length of text line was
6233 more than 4095 characters. Such lines were truncated, leading to shortened
6234 data being returned. It should now handle lines of any length.
6236 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6237 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6240 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6241 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6242 no longer happens in comment lines. A list of replacements is now placed
6243 at the head of all of the source files, except those whose only change is
6244 to replace PERL_COMMAND in the very first #! line.
6246 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6247 messages on the queue, with a bottom-up merge sort, using code contributed
6248 by Michael Haardt. This should make operations like -bp somewhat faster on
6249 large queues. It won't affect queue runners, except when queue_run_in_order
6252 51. Installed eximstats 1.31 in the distribution.
6254 52. Added support for SRV lookups to the dnslookup router.
6256 53. If an ACL referred to $message_body or $message_body_end, the value was not
6257 reset for any messages that followed in the same SMTP session.
6259 54. The store-handling optimization for building very long strings was not
6260 differentiating between the different store pools. I don't think this
6261 actually made any difference in practice, but I've tidied it.
6263 55. While running the routers to verify a sender address, $sender_address
6264 was still set to the sender address. This is wrong, because when routing to
6265 send a bounce to the sender, it would be empty. Therefore, I have changed
6266 it so that, while verifying a sender address, $sender_address is set to <>.
6267 (There is no change to what happens when verifying a recipient address.)
6269 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6270 target A or AAAA records (if not already returned) without resetting the
6271 qualify_single or search_parents options of the DNS resolver. These are
6272 inappropriate in this case because the targets of MX and SRV records must
6273 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6274 target that, when qualified, matched something in the local domain. These
6275 two options are now turned off when doing these lookups.
6277 57. It seems that at least some releases of Reiserfs (which does not have the
6278 concept of a fixed number of inodes) returns zero and not -1 for the
6279 number of available inodes. This interacted badly with check_spool_inodes,
6280 which assumed that -1 was the "no such thing" setting. What I have done is
6281 to check that the total number of inodes is greater than zero before doing
6282 the test of how many are available.
6284 58. When a "warn" ACL statement has a log_message modifier, the message is
6285 remembered, and not repeated. This is to avoid a lot of repetition when a
6286 message has many recipients that cause the same warning to be written.
6287 However, Exim was preserving the list of already written lines for an
6288 entire SMTP session, which doesn't seem right. The memory is now reset if a
6289 new message is started.
6291 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6292 debugging output unless log_rewrite was also set.
6294 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6295 of (int)(handle) when we know that handle contains (void *)(-1).
6297 61. The Exim daemon panic-logs an error return when it closes the incoming
6298 connection. However "connection reset by peer" seems to be common, and
6299 isn't really an error worthy of noting specially, so that particular error
6302 62. When Exim is trying to find all the local interfaces, it used to panic and
6303 die if the ioctl to get the interface flags failed. However, it seems that
6304 on at least one OS (Solaris 9) it is possible to have an interface that is
6305 included in the list of interfaces, but for which you get a failure error
6306 for this call. This happens when the interface is not "plumbed" into a
6307 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6308 failure of the "get flags" call assumes that the interface is down.
6310 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6311 makes life easier for people who are doing arithmetic on fields extracted
6312 from dates, where you often get leading zeros that should not be
6313 interpreted as octal.
6315 64. Added qualify_domain to the redirect router, to override the global
6318 65. If a pathologically long header line contained very many addresses (the
6319 report of this problem mentioned 10 000) and each of them was rewritten,
6320 Exim could use up a very large amount of memory. (It kept on making new
6321 copies of the header line as it rewrote, and never released the old ones.)
6322 At the expense of a bit more processing, the header rewriting function has
6323 been changed so that it no longer eats memory in this way.
6325 66. The generation of the Received: header has been moved from the time that a
6326 message starts to be received, to the time that it finishes. The timestamp
6327 in the Received: header should now be very close to that of the <= log
6328 line. There are two side-effects of this change:
6330 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6331 logged header lines no longer include the local Received: line, because
6332 it has not yet been created. The same applies to a copy of the message
6333 that is returned to a non-SMTP sender when a message is rejected.
6335 (b) When a filter file is tested using -bf, no additional Received: header
6336 is added to the test message. After some thought, I decided that this
6339 This change does not affect the value of $received_for. It is still set
6340 after address rewriting, but before local_scan() is called.
6342 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6344 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6345 gave an unhelpful panic error message, and a defer error. I have managed to
6346 change this behaviour so that it now rejects any supplied certificate,
6347 which seems right, as the list of acceptable certificates is empty.
6349 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6350 gave an unhelpful defer error. I have not managed to make this reject any
6351 supplied certificates, but the error message it gives is "no certificate
6352 supplied", which is not helpful.
6354 70. exigrep's output now also includes lines that are not associated with any
6355 message, but which match the given pattern. Implemented by a patch from
6356 Martin Sluka, which also tidied up the Perl a bit.
6358 71. Recipient callout verification, like sender verification, was using <> in
6359 the MAIL FROM command. This isn't really the right thing, since the actual
6360 sender may affect whether the remote host accepts the recipient or not. I
6361 have changed it to use the actual sender in the callout; this means that
6362 the cache record is now keyed on a recipient/sender pair, not just the
6363 recipient address. There doesn't seem to be a real danger of callout loops,
6364 since a callout by the remote host to check the sender would use <>.
6365 [SEE ABOVE: changed after hitting problems.]
6367 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6368 temporary errors. However, in the case of such a code being given after
6369 the end of a data transmission (i.e. after ".") Exim was failing to write
6370 a retry record for the message. (Yes, there was some broken host that was
6371 actually sending 8xx at this point.)
6373 73. An unknown lookup type in a host list could cause Exim to panic-die when
6374 the list was checked. (An example that provoked this was putting <; in the
6375 middle of a list instead of at the start.) If this happened during a DATA
6376 ACL check, a -D file could be left lying around. This kind of configuration
6377 error no longer causes Exim to die; instead it causes a defer error. The
6378 incident is still logged to the main and panic logs.
6380 74. Buglet left over from Exim 3 conversion. The message "too many messages
6381 in one connection" was written to the rejectlog but not the mainlog, except
6382 when address rewriting (yes!) was being logged.
6384 75. Added write_rejectlog option.
6386 76. When a system filter was run not as root (that is, when system_filter_user
6387 was set), the values of the $n variables were not being returned to the
6388 main process; thus, they were not subsequently available in the $sn
6391 77. Added +return_path_on_delivery log selector.
6393 78. A connection timeout was being treated differently from recipients deferred
6394 when testing hosts_max_try with a message that was older than the host's
6395 retry timeout. (The host should not be counted, thus allowing all hosts to
6396 be tried at least once before bouncing.) This may have been the cause of an
6397 occasionally reported bug whereby a message would remain on the queue
6398 longer than the retry timeout, but would be bounced if a delivery was
6399 forced. I say "may" because I never totally pinned down the problem;
6400 setting up timeout/retry tests is difficult. See also the next item.
6402 79. The ultimate address timeout was not being applied to errors that involved
6403 a combination of host plus message (for example, a timeout on a MAIL
6404 command). When an address resolved to a number of possible hosts, and they
6405 were not all tried for each delivery (e.g. because of hosts_max_try), a
6406 message could remain on the queue longer than the retry timeout.
6408 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6411 81. Fixed an obscure SMTP outgoing bug which required at least the following
6412 conditions: (a) there was another message waiting for the same server;
6413 (b) the server returned 5xx to all RCPT commands in the first message so
6414 that the message was not completed; (c) the server dropped the connection
6415 or gave a negative response to the RSET that Exim sends to abort the
6416 transaction. The observed case was a dropped connection after DATA that had
6417 been sent in pipelining mode. That is, the server had advertised PIPELINING
6418 but was not implementing it correctly. The effect of the bug was incorrect
6419 behaviour, such as trying another host, and this could lead to a crash.
6425 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6426 and daemon.c were passed as pointers to ints; they should have been
6427 pointers to socklen_t variables (which are typically unsigned ints).
6429 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6432 3. Fixed a really odd bug that affected only the testing scheme; patching a
6433 certain fixed string in the binary changed the value of another string that
6434 happened to be identical to the end of the original first string.
6436 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6437 name", it returns that address as the IP address. On some operating
6438 systems (e.g. Solaris), it also passes back the IP address string as the
6439 "host name". However, on others (e.g. Linux), it passes back an empty
6440 string. Exim wasn't checking for this, and was changing the host name to an
6441 empty string, assuming it had been canonicalized.
6443 5. Although rare, it is permitted to have more than one PTR record for a given
6444 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6445 all the names associated with an address, because they do in Solaris.
6446 However, it seems that they do not in Linux for data that comes from the
6447 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6448 I found this out when I moved to a new Linux workstation and tried to run
6449 the Exim test suite.
6451 To get round this problem I have changed the code so that it now does its
6452 own call to the DNS to look up PTR records when searching for a host name.
6453 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6454 addresses that are only in /etc/hosts are still found.
6456 This behaviour is, however, controlled by an option called host_lookup_
6457 order, which defaults to "bydns:byaddr". If people want to use the other
6458 order, or indeed, just use one or the other means of lookup, they can
6459 specify it in this variable.
6461 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6462 some operating systems, this comes back from gethostbyaddr() as an empty
6463 string, and this is what Exim used to test for. However, it seems that in
6464 other systems, "." is yielded. Exim now tests for this case too.
6466 7. The values of check_spool_space and check_log_space are now held internally
6467 as a number of kilobytes instead of an absolute number of bytes. If a
6468 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6469 kilobyte. This means that much larger values can be stored.
6471 8. Exim monitor: an attempt to get the action menu when not actually pointing
6472 at a message produces an empty menu entitled "No message selected". This
6473 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6474 no entries in it ("Shell widget menu has zero width and/or height"). So I
6475 have added a single, blank menu entry in this case.
6477 9. Added ${quote_local_part.
6479 10. MIME decoding is now applied to the contents of Subject: header lines when
6482 11. Now that a reference to $sender_host_address automatically causes a reverse
6483 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6484 host lookup before query-style lookups in lists that might use this
6485 variable. This has therefore been abolished, and the "net-" prefix is no
6486 longer necessary for query-style lookups.
6488 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6489 have been a typo for LFLAGS, so it has been changed.
6491 13. The install script calls Exim with "-C /dev/null" in order to find the
6492 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6493 to be output. However, since Exim outputs its version number before the
6494 error, it didn't break the script. It just looked ugly. I fixed this by
6495 always allowing "-C /dev/null" if the caller is root.
6497 14. Ignore overlarge ACL variable number when reading spool file - insurance
6498 against a later release with more variables having written the file.
6500 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6501 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6502 this, as well as the form without the "IPv6" on the front.
6504 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6505 OS/Makefile-Darwin file.
6507 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6508 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6510 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6511 variable. However, some OS use ints for inodes. I've added cast to long int
6512 to get rid of the compiler warning.
6514 19. I had forgotten to lock out "/../" in configuration file names when
6515 ALT_CONFIG_PREFIX was set.
6517 20. Routers used for verification do not need to specify transports. However,
6518 if such a router generated a host list, and callout was configured, Exim
6519 crashed, because it could not find a port number from the (non-existent)
6520 transport. It now assumes port 25 in this circumstance.
6522 21. Added the -t option to exigrep.
6524 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6525 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6526 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6527 them is set, LOOKUP_LSEARCH is forced.
6529 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6531 24. Added sender and host information to the "rejected by local_scan()" log
6532 line; previously there was no indication of these.
6534 25. Added .include_if_exists.
6536 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6537 Linux. It turns out that not all file systems support this. Apparently some
6538 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6539 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6540 not supported on the file descriptor, is now ignored when Exim is trying to
6541 sync a directory. This applies only to Linux.
6543 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6545 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6546 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6549 29. Added support for maildirsize files from supplied patch (modified a bit).
6551 30. The use of :fail: followed by an empty string could lead Exim to respond to
6552 sender verification failures with (e.g.):
6554 550 Verification failed for <xxx>
6555 550 Sender verify failed
6557 where the first response line was missing the '-' that indicates it is not
6558 the final line of the response.
6560 31. The loop for finding the name of the user that called Exim had a hardwired
6561 limit of 10; it now uses the value of finduser_retries, which is used for
6562 all other user lookups.
6564 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6567 33. Exim was neglecting to zero errno before one call of strtol() when
6568 expanding a string and expecting an integer value. On some systems this
6569 resulted in spurious "integer overflow" errors. Also, it was casting the
6570 result into an int without checking.
6572 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6573 did not work. The code looks as if it has *never* worked, though it appears
6574 to have been documented since at least release 1.62. I have made it work.
6576 35. The "timeout_DNS" error in retry rules, also documented since at least
6577 1.62, also never worked. As it isn't clear exactly what this means, and
6578 clearly it isn't a major issue, I have abolished the feature by treating it
6579 as "timeout", and writing a warning to the main and panic logs.
6581 36. The display of retry rules for -brt wasn't always showing the error code
6584 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6585 timeout_connect_A, timeout_connect_MX.
6587 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6588 to the empty sender.
6590 39. The daemon was not analysing the content of -oX till after it had closed
6591 stderr and disconnected from the controlling terminal. This meant that any
6592 syntax errors were only noted on the panic log, and the return code from
6593 the command was 0. By re-arranging the code a little, I've made the
6594 decoding happen first, so such errors now appear on stderr, and the return
6595 code is 1. However, the actual setting up of the sockets still happens in
6596 the disconnected process, so errors there are still only recorded on the
6599 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6600 connections (as happens on some IP stacks) was logged at start up time as
6601 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6602 it from "IPv6 and IPv4", which means that two separate sockets are being
6605 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6606 says whether AF_INET or AF_INET6 was passed as an argument.
6608 42. Exiwhat output was messed up when time zones were included in log
6611 43. Exiwhat now gives more information about the daemon's listening ports,
6612 and whether -tls-on-connect was used.
6614 44. The "port" option of the smtp transport is now expanded.
6616 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6617 silently ignored. Now an error message is written to the main and panic
6620 46. There's a new ACL modifier called "logwrite" which writes to a log file
6621 as soon as it is encountered.
6623 47. Added $local_user_uid and $local_user_gid at routing time.
6625 48. Exim crashed when trying to verify a sender address that was being
6628 49. Exim was recognizing only a space character after ".include". It now also
6629 recognizes a tab character.
6631 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6632 extracting the relevant information from the specification. The man page no
6633 longer contains scrambled data for the -d option, and I've added a section
6634 at the front about calling Exim under different names.
6636 51. Added "extra_headers" argument to the "mail" command in filter files.
6638 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6641 53. Installed eximstats 1.29.
6643 54. Added transport_filter_timeout as a generic transport option.
6645 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6646 Cc: header lines. This was required by RFC 822, but it not required by RFC
6649 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6650 incoming messages that did not have them. Now it does so only if the
6651 message originates locally, that is, if there is no associated remote host
6652 address. When Resent- header lines are present, this applies to the Resent-
6653 lines rather than the non-Resent- lines.
6655 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6656 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6658 58. Messages for configuration errors now include the name of the main
6659 configuration file - useful now that there may be more than one file in a
6660 list (.included file names were always shown).
6662 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6663 for those rare installations that do not start the daemon as root or run it
6664 setuid root. I've cut out the call to initgroups() if the daemon is not
6667 60. The Exim user and group can now be bound into the binary as text strings
6668 that are looked up at the start of Exim's processing.
6670 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6672 62. Added $mailstore_basename variable.
6674 63. Installed patch to sieve.c from Michael Haardt.
6676 64. When Exim failed to open the panic log after failing to open the main log,
6677 the original message it was trying to log was written to stderr and debug
6678 output, but if they were not available (the usual case in production), it
6679 was lost. Now it is written to syslog before the two lines that record the
6680 failures to open the logs.
6682 65. Users' Exim filters run in subprocesses under the user's uid. It is
6683 possible for a "deliver" command or an alias in a "personal" command to
6684 provoke an address rewrite. If logging of address rewriting is configured,
6685 this fails because the process is not running as root or exim. There may be
6686 a better way of dealing with this, but for the moment (because 4.30 needs
6687 to be released), I have disabled address rewrite logging when running a
6688 filter in a non-root, non-exim process.
6694 1. The buildconfig auxiliary program wasn't quoting the value set for
6695 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6696 not defined. This bug was masked in 4.22 by the effect that was fixed in
6699 2. Some messages that were rejected after a message id was allocated were
6700 shown as "incomplete" by exigrep. It no longer does this for messages that
6701 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6703 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6704 have allow_domain_literals set, the ID did not get logged in the <= line.
6705 Domain literals are now always recognized in Message-ID: header lines.
6707 4. The first argument for a ${extract expansion item is the key name or field
6708 number. Leading and trailing spaces in this item were not being ignored,
6709 causing some misleading effects.
6711 5. When deliver_drop_privilege was set, single queue runner processes started
6712 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6713 same command in the process it spins off) were not dropping privilege.
6715 6. When the daemon running as "exim" started a queue runner, it always
6716 re-executed Exim in the spun-off process. This is a waste of effort when
6717 deliver_drop_privilege is set. The new process now just calls the
6718 queue-runner function directly.
6724 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6727 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6728 ignored. Though the use of -oP was forcing the writing of a pid file, it
6729 was always written to the default place.
6731 3. If the message "no IP address found for host xxxx" is generated during
6732 incoming verification, it is now followed by identification of the incoming
6733 connection (so you can more easily find what provoked it).
6735 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6737 5. Added some features to "harden" Exim a bit more against certain attacks:
6739 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6740 be put in Local/Makefile. This is like the never_users runtime option,
6741 but it cannot be overridden. The default setting is "root".
6743 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6744 prefix string with which any file named in a -C command line option
6747 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6748 is retained for -C and -D only if the caller of Exim is root. Without
6749 it, the exim user may also use -C and -D and retain privilege.
6751 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6752 command line option is disabled.
6754 6. Macro names set by the -D option must start with an upper case letter, just
6755 like macro names defined in the configuration file.
6757 7. Added "dereference=" facility to LDAP.
6759 8. Two instances of the typo "uknown" in the source files are fixed.
6761 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6762 the Configure-Makefile script screwed up while processing it.
6764 10. Incorporated PCRE 4.4.
6766 11. The SMTP synchronization check was not operating right at the start of an
6767 SMTP session. For example, it could not catch a HELO sent before the client
6768 waited for the greeting. There is now a check for outstanding input at the
6769 point when the greeting is written. Because of the duplex, asynchronous
6770 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6771 way, but not yet received, when the check is performed.
6773 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6774 on TCP/IP sockets, because this apparently causes some broken clients to
6777 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6778 unchanged) from the Cygwin maintainer.
6780 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6781 was supported instead of testing for mbx. Effectively a typo.
6783 15. The spa authenticator server code was not checking that the input it
6784 received was valid base64.
6786 16. The debug output line for the "set" modifier in ACLs was not showing the
6787 name of the variable that was being set.
6789 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6791 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6792 lookup on its own. Something else was needed to trigger the lookup. For
6793 example, a match in host_lookup or the need for a host name in a host list.
6794 Now, if $sender_host_name is referenced and the host name has not yet been
6795 looked up, a lookup is performed. If the lookup fails, the variable remains
6796 empty, and $host_lookup_failed is set to "1".
6798 19. Added "eqi" as a case-independent comparison operator.
6800 20. The saslauthd authentication condition could segfault if neither service
6801 nor realm was specified.
6803 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6804 error message that was logged was misleading, and incoming SMTP
6805 connections were dropped. The message is now more accurate, and temporary
6806 errors are given to SMTP connections.
6808 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6809 (except RSET) with a 421 error, until QUIT is received. However, it was
6810 failing to send a response to QUIT.
6812 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6813 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6814 ACL is now run after the helo_try_verify_hosts code.
6816 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6817 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6818 case-independent, so other case variants are also recognized). Apparently
6819 some systems use these upper case variants.
6821 25. If more than two messages were waiting for the same host, and a transport
6822 filter was specified for the transport, Exim sent two messages over the
6823 same TCP/IP connection, and then failed with "socket operation on non-
6824 socket" when it tried to send the third.
6826 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6828 27. The extern definition of crypt16() in expand.c was not being excluded when
6829 the OS had its own crypt16() function.
6831 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6832 as a preferred synonym for return_size_limit, both as an option and as an
6835 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6837 30. Changed the default configuration ACL to relax the local part checking rule
6838 for addresses that are not in any local domains. For these addresses,
6839 slashes and pipe symbols are allowed within local parts, but the sequence
6840 /../ is explicitly forbidden.
6842 31. SPA server authentication was not clearing the challenge buffer before
6845 32. log_message in a "warn" ACL statement was writing to the reject log as
6846 well as to the main log, which contradicts the documentation and doesn't
6847 seem right (because no rejection is happening). So I have stopped it.
6849 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6850 However, I am unable to do any testing of this.
6852 34. Fixed an infelicity in the appendfile transport. When checking directories
6853 for a mailbox, to see if any needed to be created, it was accidentally
6854 using path names with one or more superfluous leading slashes; tracing
6855 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6857 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6858 contents are added to the log line that is written for every discarded
6859 recipient. (Previously a log_message setting was ignored.)
6861 36. The ${quote: operator now quotes the string if it is empty.
6863 37. The install script runs exim in order to find its version number. If for
6864 some reason other than non-existence or emptiness, which it checks, it
6865 could not run './exim', it was installing it with an empty version number,
6866 i.e. as "exim-". This error state is now caught, and the installation is
6869 38. An argument was missing from the function that creates an error message
6870 when Exim fails to connect to the socket for saslauthd authentication.
6871 This could cause Exim to crash, or give a corrupted message.
6873 39. Added isip, isip4, and isip6 to ${if conditions.
6875 40. The ACL variables $acl_xx are now saved with the message, and can be
6876 accessed later in routers, transports, and filters.
6878 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6879 strings in the file are not string-expanded.
6881 42. If a MAIL command specified a SIZE value that was too large to fit into an
6882 int variable, the check against message_size_limit failed. Such values are
6883 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6884 day this will have to be increased, but I don't think I want to be around
6885 when emails are that large.
6892 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6893 iconv() is not standard in FreeBSD.
6895 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6896 IPv6 enabled. The observed symptom was a segmentation fault on return from
6897 the function os_common_find_running_interfaces() in src/os.c.
6899 3. In the check_special_case() function in daemon.c I had used "errno" as an
6900 argument name, which causes warnings on some systems. This was basically a
6901 typo, since it was named "eno" in the comments!
6903 4. The code that waits for the clock to tick (at a resolution of some fraction
6904 of a second) so as to ensure message-id uniqueness was always waiting for
6905 at least one whole tick, when it could have waited for less. [This is
6906 almost certainly not relevant at current processor speeds, where it is
6907 unlikely to ever wait at all. But we try to future-proof.]
6909 5. The function that sleeps for a time interval that includes fractions of a
6910 second contained a race. It did not block SIGALRM between setting the
6911 timer, and suspending (a couple of lines later). If the interval was short
6912 and the sigsuspend() was delayed until after it had expired, the suspension
6913 never ended. On busy systems this could lead to processes getting stuck for
6916 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6917 process, before it forks any delivery processes. The open lookup caching
6918 mechanism meant that the open file or database connection was passed into
6919 the delivery process. The problem was that delivery processes always tidy
6920 up cached lookup data. This could cause a problem for the next delivery
6921 process started by the queue runner, because the external queue runner
6922 process does not know about the closure. So the next delivery process
6923 still has data in the lookup cache. In the case of a file lookup, there was
6924 no problem because closing a file descriptor in a subprocess doesn't affect
6925 the parent. However, if the lookup was caching a connection to a database,
6926 the connection was closed, and the second delivery process was likely to
6927 see errors such as "PGSQL: query failed: server closed the connection
6928 unexpectedly". The problem has been fixed by closing all cached lookups
6929 in a queue runner before running a delivery process.
6931 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6932 seem to have the "const" qualifier which it has on other OS. I've
6935 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6936 *of the same type* (client or server) with the same public name that an
6937 error should be diagnosed.
6939 9. When Exim looked up a host name for an IP address, but failed to find the
6940 original IP address when looking up the host name (a safety check), it
6941 output the message "<ip address> does not match any IP for NULL", which was
6942 confusing, to say the least. The bug was that the host name should have
6943 appeared instead of "NULL".
6945 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6946 user that is built into the binary, and the -C or -D options is used, root
6947 privilege is dropped before the configuration file is read. In addition,
6948 logging is switched to stderr instead of the normal log files. If the
6949 configuration then re-defines the Exim user, the unprivileged environment
6950 is probably not what is expected, so Exim logs a panic warning message (but
6953 However, if deliver_drop_privilege is set, the unprivileged state may well
6954 be exactly what is intended, so the warning has been cut out in that case,
6955 and Exim is allowed to try to write to its normal log files.
6961 1. smtp_return_error_details was not giving details for temporary sender
6962 or receiver verification errors.
6964 2. Diagnose a configuration error if two authenticators have the same public
6967 3. Exim used not to create the message log file for a message until the first
6968 delivery attempt. This could be confusing when incoming messages were held
6969 for policy or load reasons. The message log file is now created at the time
6970 the message is received, and an initial "Received" line is written to it.
6972 4. The automatically generated man page for command line options had a minor
6973 bug that caused no ill effects; however, a more serious problem was that
6974 the procedure for building the man page automatically didn't always
6975 operate. Consequently, release 4.20 contains an out-of-date version. This
6976 shouldn't happen again.
6978 5. When building Exim with embedded Perl support, the script that builds the
6979 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6980 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6982 6. The freeze_tell option was not being used for messages that were frozen on
6983 arrival, either by an ACL or by local_scan().
6985 7. Added the smtp_incomplete_transaction log selector.
6987 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6988 was accepting AUTH without a new EHLO.
6990 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6991 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6994 10. Logging of TCP/IP connections (when configured) now happens in the main
6995 daemon process instead of the child process, so that the TCP/IP connection
6996 count is more accurate (but it can never be perfect).
6998 11. The use of "drop" in a nested ACL was not being handled correctly in the
6999 outer ACL. Now, if condition failure induced by the nested "drop" causes
7000 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7001 or "require"), the connection is dropped.
7003 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7004 that yield "discard" can now be used with an "accept" or a "discard" verb,
7005 but an error is generated for any others (because I can't see a useful way
7006 to define what should happen).
7008 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7009 are now processed in the same way as lines in the Exim configuration file.
7010 In particular, continuation lines are supported.
7012 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7014 15. Added -ti meaning -t -i.
7016 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7017 domains, and warn by logging if others are found.
7019 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7020 returned from GIFCONF when Exim is trying to find the list of interfaces on
7021 a host. The code in os.c has been modified to copy each ifreq to an aligned
7022 structure in all cases.
7024 Also, in some cases, the returned ifreq's were being copied to a 'struct
7025 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7026 means the last couple of bytes of an IPv6 address could be chopped if the
7027 ifreq contained only a normal sockaddr (14 bytes storage).
7029 18. Named domain lists were not supported in the hosts_treat_as_local option.
7030 An entry such as +xxxx was not recognized, and was treated as a literal
7033 19. Ensure that header lines added by a DATA ACL are included in the reject log
7034 if the ACL subsequently rejects the message.
7036 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7037 MD5 (which is deprecated).
7039 21. When testing a filter file using -bf, Exim was writing a message when it
7040 took the sender from a "From " line in the message, but it was not doing so
7041 when it took $return_path from a Return-Path: header line. It now does.
7043 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7044 with a valid header line field name (a series of printing characters
7045 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7047 23. Changed "disc" in the source to "disk" to conform to the documentation and
7048 the book and for uniformity.
7050 24. Ignore Sendmail's -Ooption=value command line item.
7052 25. When execve() failed while trying to run a command in a pipe transport,
7053 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7054 could be confused with a return value of 69 from the command itself. This
7055 has been changed to 127, the value the shell returns if it is asked to run
7056 a non-existent command. The wording for the related log line suggests a
7057 non-existent command as the problem.
7059 26. If received_header_text expands to an empty string, do not add a Received:
7060 header line to the message. (Well, it adds a token one on the spool, but
7061 marks it "old" so that it doesn't get used or transmitted.)
7063 27. Installed eximstats 1.28 (addition of -nt option).
7065 28. There was no check for failure on the call to getsockname() in the daemon
7066 code. This can fail if there is a shortage of resources on the system, with
7067 ENOMEM, for example. A temporary error is now given on failure.
7069 29. Contrary to the C standard, it seems that in some environments, the
7070 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7071 program. Exim now does this explicitly; it affects the formatting of
7072 timestamps using strftime().
7074 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7075 complaints. I've now initialized all the variables, to avoid this.
7077 32. Header lines added by a system filter were not being "seen" during
7078 transport-time rewrites.
7080 33. The info_callback() function passed to OpenSSL is set up with type void
7081 (*)(SSL *, int, int), as described somewhere. However, when calling the
7082 function (actually a macro) that sets it up, the type void(*)() is
7083 expected. I've put in a cast to prevent warnings from picky compilers.
7085 34. If a DNS black list lookup found a CNAME record, but there were no A
7086 records associated with the domain it pointed at, Exim crashed.
7088 35. If a DNS black list lookup returned more than one A record, Exim ignored
7089 all but the first. It now scans all returned addresses if a particular IP
7090 value is being sought. In this situation, the contents of the
7091 $dnslist_value variable are a list of all the addresses, separated by a
7094 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7095 to accept a host name and all its aliases if the forward lookup for any of
7096 them yielded the IP address of the incoming connection. Now it accepts only
7097 those names whose forward lookup yields the correct IP address. Any other
7098 names are discarded. This closes a loophole whereby a rogue DNS
7099 administrator could create reverse DNS records to break through a
7100 wildcarded host restriction in an ACL.
7102 37. If a user filter or a system filter that ran in a subprocess used any of
7103 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7104 the wrong values were passed to the pipe command ($thisaddress had the
7105 value of $0, $0 had the value of $1, etc). This bug was introduced by
7106 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7108 38. Improved the line breaking for long SMTP error messages from ACLs.
7109 Previously, if there was no break point between 40 and 75 characters, Exim
7110 left the rest of the message alone. Two changes have been made: (a) I've
7111 reduced the minimum length to 35 characters; (b) if it can't find a break
7112 point between 35 and 75 characters, it looks ahead and uses the first one
7113 that it finds. This may give the occasional overlong line, but at least the
7114 remaining text gets split now.
7116 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7117 file descriptors might be low, and that setting 1000 would always raise it.
7118 It turns out that in some environments, the limit is already over 1000 and
7119 that lowering it causes trouble. So now Exim takes care not to decrease it.
7121 40. When delivering a message, the value of $return_path is set to $sender_
7122 address at the start of routing (routers may change the value). By an
7123 oversight, this default was not being set up when an address was tested by
7124 -bt or -bv, which affected the outcome if any router or filter referred to
7127 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7128 log only when "message" or "log_message" are set. However, if one of the
7129 conditions was an address verification, or a call to a nested ACL, the
7130 messages generated by the underlying test were being passed through. This
7131 no longer happens. The underlying message is available in $acl_verify_
7132 message for both "message" and "log_message" expansions, so it can be
7133 passed through if needed.
7135 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7136 new expansion $bh_ to give the encoded byte string without charset
7137 translation. Translation happens only if iconv() is available; HAVE_ICONV
7138 indicates this at build time. HEADERS_CHARSET gives the charset to
7139 translate to; headers_charset can change it in the configuration, and
7140 "headers charset" can change it in an individual filter file.
7142 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7143 that creates RFC 2047 encoded "words" labels them as that charset instead
7144 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7145 expansion operator; (ii) when Exim creates a From: line for a local
7146 message; (iii) when a header line is rewritten to include a "phrase" part.
7148 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7149 buggy, causing it to skip the first lines of messages whose message ID
7150 ended in 'D'. This would not have bitten before Exim release 4.14, because
7151 message IDs were unlikely to end in 'D' before then. The effect was to have
7152 incorrect size information for certain domains.
7154 45. #include "config.h" was missing at the start of the crypt16.c module. This
7155 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7158 46. If there was a timeout during a "random" callout check, Exim treated it as
7159 a failure of the random address, and carried on sending RSET and the real
7160 address. If the delay was just some slowness somewhere, the response to the
7161 original RCPT would be taken as a response to RSET and so on, causing
7162 mayhem of various kinds.
7164 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7165 when I implemented it. It didn't allow for the fact that some option values
7166 may legitimately be negative (e.g. size_addition), and it didn't even do
7167 the right test for positive values.
7169 48. Domain names in DNS records are case-independent. Exim always looks them up
7170 in lower case. Some resolvers return domain names in exactly the case they
7171 appear in the zone file, that is, they may contain uppercase letters. Not
7172 all resolvers do this - some return always lower case. Exim was treating a
7173 change of case by a resolver as a change of domain, similar to a widening
7174 of a domain abbreviation. This triggered its re-routing code and so it was
7175 trying to route what was effectively the same domain again. This normally
7176 caused routing to fail (because the router wouldn't handle the domain
7177 twice). Now Exim checks for this case specially, and just changes the
7178 casing of the domain that it ultimately uses when it transmits the message
7181 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7184 50. If a filter generated a file delivery with a non-absolute name (possible if
7185 no home directory exists for the router), the forbid_file option was not
7188 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7189 the existing equality matching.
7191 52. Exim was using ints instead of ino_t variables in some places where it was
7192 dealing with inode numbers.
7194 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7195 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7196 called TMPDIR, and if it finds it is different, it changes its value.
7198 54. The smtp_printf() function is now made available to local_scan() so
7199 additional output lines can be written before returning. There is also an
7200 smtp_fflush() function to enable the detection of a dropped connection.
7201 The variables smtp_input and smtp_batched_input are exported to
7204 55. Changed the default runtime configuration: the message "Unknown user"
7205 has been removed from the ACL, and instead placed on the localuser router,
7206 using the cannot_route_message feature. This means that any verification
7207 failures that generate their own messages won't get overridden. Similarly,
7208 the "Unrouteable address" message that was in the ACL for unverifiable
7209 relay addresses has also been removed.
7211 56. Added hosts_avoid_esmtp to the smtp transport.
7213 57. The exicyclog script was not checking for the esoteric option
7214 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7215 will work only if exicyclog is run under the appropriate euid.
7217 58. Following a discussion on the list, the rules by which Exim recognises line
7218 endings on incoming messages have been changed. The -dropcr and drop_cr
7219 options are now no-ops, retained only for backwards compatibility. The
7220 following line terminators are recognized: LF CRLF CR. However, special
7221 processing applies to CR:
7223 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7224 nor a local message in the state where . is a terminator.
7226 (ii) If a bare CR is encountered in a header line, an extra space is added
7227 after the line terminator so as not to end the header. The reasoning
7228 behind this is that bare CRs in header lines are most likely either
7229 to be mistakes, or people trying to play silly games.
7231 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7232 was being incorrectly given as 18 bytes larger than it should have been.
7233 This is a VOB (very old bug).
7235 60. This may never have affected anything current, but just in case it has:
7236 When the local host is found other than at the start of a list of hosts,
7237 the local host, those with the same MX, and any that follow, are discarded.
7238 When the list in question was part of a longer list of hosts, the following
7239 hosts (not currently being processed) were also being discarded. This no
7240 longer happens. I'm not sure if this situation could ever has previously
7243 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7246 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7247 header. This is something that is recommended in a new Internet Draft, and
7248 is something that is documented as being done by Sendmail. There are two
7249 possible values. For messages generated by the autoreply transport, Exim
7252 Auto-Submitted: auto-replied
7254 whereas for all other generated messages (e.g. bounces) it adds
7256 Auto-Submitted: auto-generated
7258 63. The "personal" condition in filters now includes a test for the
7259 Auto-Submitted: header. If it contains the string "auto-" the message it
7260 not considered personal.
7262 64. Added rcpt_include_affixes as a generic transport option.
7264 65. Added queue_only_override (default true).
7266 66. Added the syslog_duplication option.
7268 67. If what should have been the first header line of a message consisted of
7269 a space followed by a colon, Exim was mis-interpreting it as a header line.
7270 It isn't of course - it is syntactically invalid and should therefore be
7271 treated as the start of the message body. The misbehaviour could have
7272 caused a number of strange effects, including loss of data in subsequent
7273 header lines, and spool format errors.
7275 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7276 client host had authenticated. This control can now be exercised by an ACL
7277 for more flexibility.
7279 69. By default, callouts do not happen when testing with -bh. There is now a
7280 variant, -bhc, which does actually run the callout code, including
7281 consulting and updating the callout cache.
7283 70. Added support for saslauthd authentication, courtesy of Alexander
7286 71. If statvfs() failed on the spool or log directories while checking their
7287 size for availability, Exim confusingly gave the error "space shortage".
7288 Furthermore, in debugging mode it crashed with a floating point exception.
7289 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7290 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7291 serious problem, Exim now writes to the main and panic logs when this
7292 happens, with details of the failure. It then refuses to accept the
7293 incoming message, giving the message "spool directory problem" or "log
7294 directory problem" with a 421 code for SMTP messages.
7296 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7297 0, 1, and 2 exist, because some OS complain for execs without them (see
7298 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7299 descriptors. However, the code omitted to check that the open succeeded,
7300 causing mysterious errors if for some reason the permissions on /dev/null
7301 got screwed. Now Exim writes a message to the main and panic logs, and
7302 bombs out if it can't open /dev/null.
7304 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7305 interact so that it is all more flexible. It is supposed to remain
7306 backwards compatible. Also added extra_local_interfaces.
7308 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7309 to bomb out with an assertion failure - to the client this appears as a
7310 connection drop. This problem occurs in the part of the code that was taken
7311 from the Samba project. Fortunately, the assertion is in a very simple
7312 function, so I have fixed this by reproducing the function inline in the
7313 one place where it is called, and arranging for authentication to fail
7314 instead of killing the process with assert().
7316 75. The SPA client code was not working when the server requested OEM rather
7317 than Unicode encoding.
7319 76. Added code to make require_files with a specific uid setting more usable in
7320 the case where statting the file as root fails - usually a non-root-mounted
7321 NFS file system. When this happens and the failure is EACCES, Exim now
7322 forks a subprocess and does the per-uid checking as the relevant uid.
7324 77. Added process_log_path.
7326 78. If log_file_path was not explicitly set, a setting of check_log_space or
7327 check_log_inodes was ignored.
7329 79. If a space check for the spool or log partitions fails, the incident is now
7330 logged. Of course, in the latter case the data may get lost...
7332 80. Added the %p formatting code to string_format() so that it can be used to
7333 print addresses in debug_print(). Adjusted all the address printing in the
7334 debugging in store.c to use %p rather than %d.
7336 81. There was a concern that a line of code in smtp_in.c could overflow a
7337 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7338 initially expressed, the concern was not well-founded, because trailing
7339 spaces are removed early. However, if the trailing spaces were followed by
7340 a NULL, they did not get removed, so the overflow was possible. Two fixes
7343 (a) I re-wrote the offending code in a cleaner fashion.
7344 (b) If an incoming SMTP command contains a NULL character, it is rejected
7347 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7348 runs initgroups(), so that if the Exim user is in any additional groups,
7349 they will be used during message reception.
7355 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob