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
184 return error codes indicating retry. Under TLS1.3 this becomes required.
186 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
187 it only wrote the new authenticators, resulting in a lack of tracking of
188 peer changes of ESMTP extensions until the next cache flush.
190 JH/41 Fix the loop reading a message header line to check for integer overflow,
191 and more-often against header_maxsize. Previously a crafted message could
192 induce a crash of the recive process; now the message is cleanly rejected.
198 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
199 When redis_servers is set to a list of > 1 element, and the Redis servers
200 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
201 case of MOVED into a DEFER case instead, thus moving the query onto the
202 next server in the list. For a cluster of N elements, all N servers must
203 be defined in redis_servers.
205 GF/02 Catch and remove uninitialized value warning in exiqsumm
206 Check for existence of @ARGV before looking at $ARGV[0]
208 JH/01 Replace the store_release() internal interface with store_newblock(),
209 which internalises the check required to safely use the old one, plus
210 the allocate and data copy operations duplicated in both (!) of the
211 extant use locations.
213 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
214 modifier. This matches the restriction on the commandline.
216 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
217 Previously only the last row was returned.
219 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
220 we assumed that tags in the header were well-formed, and parsed the
221 element content after inspecting only the first char of the tag.
222 Assumptions at that stage could crash the receive process on malformed
225 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
226 While running the DKIM ACL we operate on the Permanent memory pool so that
227 variables created with "set" persist to the DATA ACL. Also (at any time)
228 DNS lookups that fail create cache records using the Permanent pool. But
229 expansions release any allocations made on the current pool - so a dnsdb
230 lookup expansion done in the DKIM ACL releases the memory used for the
231 DNS negative-cache, and bad things result. Solution is to switch to the
232 Main pool for expansions.
233 While we're in that code, add checks on the DNS cache during store_reset,
234 active in the testsuite.
235 Problem spotted, and debugging aided, by Wolfgang Breyha.
237 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
238 When none of the hosts presented to a transport match an already-open
239 connection, close it and proceed with the list. Previously we would
240 queue the message. Spotted by Lena with Yahoo, probably involving
243 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
244 Previously a spurious "250 OK id=" response was appended to the proper
247 JH/08 The "support for" informational output now, which built with Content
248 Scanning support, has a line for the malware scanner interfaces compiled
249 in. Interface can be individually included or not at build time.
251 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
252 by the template makefile "src/EDITME". The "STREAM" support for an older
253 ClamAV interface method is removed.
255 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
256 rows affected is given instead).
258 JH/11 The runtime Berkeley DB library version is now additionally output by
259 "exim -d -bV". Previously only the compile-time version was shown.
261 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
262 SMTP connection. Previously, when one had more receipients than the
263 first, an abortive onward connection was made. Move to full support for
264 multiple onward connections in sequence, handling cutthrough connection
265 for all multi-message initiating connections.
267 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
268 routers. Previously, a multi-recipient message would fail to match the
269 onward-connection opened for the first recipient, and cause its closure.
271 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
272 a timeout on read on a GnuTLS initiating connection, resulting in the
273 initiating connection being dropped. This mattered most when the callout
274 was marked defer_ok. Fix to keep the two timeout-detection methods
277 JH/15 Relax results from ACL control request to enable cutthrough, in
278 unsupported situations, from error to silently (except under debug)
279 ignoring. This covers use with PRDR, frozen messages, queue-only and
282 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
284 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
285 metadata, resulting in a crash in free().
287 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
288 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
289 Broken also in d185889f4, with init system revamp.
291 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
292 Previously we abruptly closed the connection after reading a malware-
293 found indication; now we go on to read the "scan ok" response line,
296 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
297 ACL. Previously, a crash would result.
299 JH/19 Speed up macro lookups during configuration file read, by skipping non-
300 macro text after a replacement (previously it was only once per line) and
301 by skipping builtin macros when searching for an uppercase lead character.
303 JH/20 DANE support moved from Experimental to mainline. The Makefile control
304 for the build is renamed.
306 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
307 was allocated for every new TLS startup, meaning one per message. Fix
308 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
310 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
311 reported the original. Fix to report (as far as possible) the ACL
312 result replacing the original.
314 JH/23 Fix memory leak during multi-message connections using STARTTLS under
315 OpenSSL. Certificate information is loaded for every new TLS startup,
316 and the resources needed to be freed.
318 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
320 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
323 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
324 DATA response info to the (existing) per-recipient response info for
325 the "C=" log element. It can have useful tracking info from the
326 destination system. Patch from Simon Arlott.
328 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
329 length value. Previously this would segfault.
331 HS/02 Support Avast multiline protoocol, this allows passing flags to
332 newer versions of the scanner.
334 JH/28 Ensure that variables possibly set during message acceptance are marked
335 dead before release of memory in the daemon loop. This stops complaints
336 about them when the debug_store option is enabled. Discovered specifically
337 for sender_rate_period, but applies to a whole set of variables.
338 Do the same for the queue-runner and queue-list loops, for variables set
339 from spool message files. Do the same for the SMTP per-message loop, for
340 certain variables indirectly set in ACL operations.
342 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
343 as a multi-recipient message from a mailinglist manager). The coding had
344 an arbitrary cutoff number of characters while checking for more input;
345 enforced by writing a NUL into the buffer. This corrupted long / fast
346 input. The problem was exposed more widely when more pipelineing of SMTP
347 responses was introduced, and one Exim system was feeding another.
348 The symptom is log complaints of SMTP syntax error (NUL chars) on the
349 receiving system, and refused recipients seen by the sending system
350 (propating to people being dropped from mailing lists).
351 Discovered and pinpointed by David Carter.
353 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
354 replaced by the ${authresults } expansion.
356 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
358 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
359 allows proper process termination in container environments.
361 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
362 Previously the "final dot" had a newline after it; ensure it is CR,LF.
364 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
365 and "err_perm", deprecated since 4.83 when the RFC-defined words
366 "temperror" and "permerror" were introduced.
368 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
369 transport-filters or DKIM-signing. The restriction was lost in the
370 consolidation of verify-callout and delivery SMTP handling.
371 Extend the restriction to also cover ARC-signing.
373 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
374 in defer=pass mode supply a 450 to the initiator. Previously the message
377 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
378 tls_require_ciphers is used as before.
380 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
381 "pass_unscanned". Only tmpfails from the scanner are written to
382 the paniclog, as they may require admin intervention (permission
383 denied, license issues). Other scanner errors (like decompression
384 bombs) do not cause a paniclog entry.
386 JH/36 Fix reinitialisation of DKIM logging variable between messages.
387 Previously it was possible to log spurious information in receive log
390 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
391 triggered odd behaviour from Outlook Express clients.
393 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
396 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
397 since the IETF WG has not yet settled on that versus the original
398 "bare" representation.
400 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
401 Previously the millisecond value corrupted the output.
402 Fix also for syslog_pid=no and log_selector +pid, for which the pid
403 corrupted the output.
409 JH/01 Rework error string handling in TLS interface so that the caller in
410 more cases is responsible for logging. This permits library-sourced
411 string to be attached to addresses during delivery, and collapses
412 pairs of long lines into single ones.
414 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
415 during configuration. Wildcards are allowed and expanded.
417 JH/02 Rework error string handling in DKIM to pass more info back to callers.
418 This permits better logging.
420 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
421 do not close it down and have the child transport start it up again on
422 the passed-on TCP connection. Instead, proxy the child (and any
423 subsequent ones) for TLS via a unix-domain socket channel. Logging is
424 affected: the continued delivery log lines do not have any DNSSEC, TLS
425 Certificate or OCSP information. TLS cipher information is still logged.
427 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
428 identical IP addresses on different listening ports. Will also affect
431 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
432 add noisy ifdef guards to special-case this sillyness.
433 Patch from Bernd Kuhls.
435 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
436 than 255 are no longer allowed.
438 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
439 Disable the session-cache too, which might reduce our load. Since we
440 currrectly use a new context for every connection, both as server and
441 client, there is no benefit for these.
442 GnuTLS appears to not support tickets server-side by default (we don't
443 call gnutls_session_ticket_enable_server()) but client side is enabled
444 by default on recent versions (3.1.3 +) unless the PFS priority string
447 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
448 <https://reproducible-builds.org/specs/source-date-epoch/>.
450 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
451 the check for any unsuccessful recipients did not notice the limit, and
452 erroneously found still-pending ones.
454 JH/08 Pipeline CHUNKING command and data together, on kernels that support
455 MSG_MORE. Only in-clear (not on TLS connections).
457 JH/09 Avoid using a temporary file during transport using dkim. Unless a
458 transport-filter is involved we can buffer the headers in memory for
459 creating the signature, and read the spool data file once for the
460 signature and again for transmission.
462 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
463 4.77 as the kernel support then wasn't solid, having issues in 64bit
464 mode. Now, it's been long enough. Add support for FreeBSD also.
466 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
467 case where the routing stage had gathered several addresses to send to
468 a host before calling the transport for the first, we previously failed
469 to close down TLS in the old transport process before passing the TCP
470 connection to the new process. The new one sent a STARTTLS command
471 which naturally failed, giving a failed delivery and bloating the retry
472 database. Investigation and fix prototype from Wolfgang Breyha.
474 JH/12 Fix check on SMTP command input synchronisation. Previously there were
475 false-negatives in the check that the sender had not preempted a response
476 or prompt from Exim (running as a server), due to that code's lack of
477 awareness of the SMTP input buffering.
479 PP/04 Add commandline_checks_require_admin option.
480 Exim drops privileges sanely, various checks such as -be aren't a
481 security problem, as long as you trust local users with access to their
482 own account. When invoked by services which pass untrusted data to
483 Exim, this might be an issue. Set this option in main configuration
484 AND make fixes to the calling application, such as using `--` to stop
487 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
488 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
489 responses to those, into a single TLS record each way (this usually means
490 a single packet). As a side issue, smtp_enforce_sync now works on TLS
493 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
494 affects you only if you're dancing at the edge of the param size limits.
495 If you are, and this message makes sense to you, then: raise the
496 configured limit or use OpenSSL 1.1. Nothing we can do for older
499 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
500 cmdline element to get the documented default one. Previously it was
503 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
504 are unsuable for non-message-specific reasons - eg. connection timeout,
505 banner-time rejection.
507 JH/16 Fix logging of delivery remote port, when specified by router, under
510 PP/06 Repair manualroute's ability to take options in any order, even if one
511 is the name of a transport.
514 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
516 JH/17 Change the list-building routines interface to use the expanding-string
517 triplet model, for better allocation and copying behaviour.
519 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
520 Previously it was constructed the first time a possibly-matching string
521 was met in the configuration file input during startup; now it is done
524 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
525 compatible one, to avoid the (poorly documented) possibility of a config
526 file in the working directory redirecting the DB files, possibly correpting
527 some existing file. CVE-2017-10140 assigned for BDB.
529 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
530 cache-hot. Previously, although the result was properly cached, the
531 initial verify call returned a defer.
533 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
534 the main verify for receipient in uncached-mode.
536 JH/22 Retire historical build files to an "unsupported" subdir. These are
537 defined as "ones for which we have no current evidence of testing".
539 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
540 if present. Previously it was ignored.
542 JH/24 Start using specified-initialisers in C structure init coding. This is
543 a C99 feature (it's 2017, so now considered safe).
545 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
546 if was a fixed-sized field and bitmask ops via macros; it is now more
549 PP/07 GitHub PR 56: Apply MariaDB build fix.
550 Patch provided by Jaroslav Škarvada.
552 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
553 during Coverity cleanups [4.87 JH/47]
554 Diagnosis and fix provided by Michael Fischer v. Mollard.
556 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
557 the right size to place the terminating semicolon on its own folded
558 line, the header hash was calculated to an incorrect value thanks to
559 the (relaxed) space the fold became.
561 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
562 and confused the parent.
564 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
565 which could crash as a result. This could lead to undeliverable messages.
567 JH/28 Logging: "next input sent too soon" now shows where input was truncated
570 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
571 matters on fast-turnover and PID-randomising systems, which were getting
572 out-of-order delivery.
574 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
575 a possibly-overlapping copy. The symptom was that "Remote host closed
576 connection in response to HELO" was logged instead of the actual 4xx
579 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
580 Previously only that bufferd was discarded, resulting in SYMTP command
583 JH/32 DKIM: when a message has multiple signatures matching an identity given
584 in dkim_verify_signers, run the dkim acl once for each. Previously only
585 one run was done. Bug 2189.
587 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
588 "panic the current process" to "deliberately defer". The panic log is
589 still written with the problem list name; the mail and reject logs now
590 get a temp-reject line for the message that was being handled, saying
591 something like "domains check lookup or other defer". The SMTP 451
592 message is still "Temporary local problem".
594 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
595 A crafted sequence of BDAT commands could result in in-use memory beeing
596 freed. CVE-2017-16943.
598 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
599 from SMTP input. Previously it was always done; now only done for DATA
600 and not BDAT commands. CVE-2017-16944.
602 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
603 to the message (such as an overlong header line). Previously this was
604 not done and we did not exit BDAT mode. Followon from the previous item
605 though a different problem.
611 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
612 than -2003 did; needs libidn2 in addition to libidn.
614 JH/02 The path option on a pipe transport is now expanded before use.
616 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
617 Patch provided by "Björn", documentation fix added too.
619 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
620 missing a wire-to-host endian conversion.
622 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
623 close after a BDAT command line could be taken as a following command,
624 giving a synch failure. Fix by only checking for synch immediately
625 before acknowledging the chunk.
627 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
628 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
629 macro. Patches provided by Josh Soref.
631 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
632 Previously we did not; the RFC seems ambiguous and VRFY is not listed
633 by IANA as a service extension. However, John Klensin suggests that we
636 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
637 the dkim code may be unix-mode line endings rather than smtp wire-format
638 CRLF, so prepend a CR to any bare LF.
640 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
641 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
643 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
644 the first were themselves being wrongly included in the feed into dkim
645 processing; with most chunk sizes in use this resulted in an incorrect
646 body hash calculated value.
648 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
649 DKIM signature block, for verification. Although advised against by
650 standards it is specifically not ruled illegal.
652 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
654 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
655 missing a body hash (the bh= tag).
657 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
658 It seems that HAProxy sends the Proxy Protocol information in clear and
659 only then does a TLS startup, so do the same.
661 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
662 TCP connections (such as for Spamd) unless the daemon successfully set
663 Fast Open mode on its listening sockets. This fixes breakage seen on
664 too-old kernels or those not configured for Fast Open, at the cost of
665 requiring both directions being enabled for TFO, and TFO never being used
666 by non-daemon-related Exim processes.
668 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
669 endings, at least on the first header line. Try to canonify any that get
670 past that check, despite the cost.
672 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
673 now limited to an arbitrary five deep, while parsing addresses with the
674 strip_excess_angle_brackets option enabled.
676 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
677 instead leave the unprompted TLS handshake in socket buffer for the
678 TLS library to consume.
680 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
682 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
684 JH/16 Drop variables when they go out of scope. Memory management drops a whole
685 region in one operation, for speed, and this leaves assigned pointers
686 dangling. Add checks run only under the testsuite which checks all
687 variables at a store-reset and panics on a dangling pointer; add code
688 explicitly nulling out all the variables discovered. Fixes one known
689 bug: a transport crash, where a dangling pointer for $sending_ip_address
690 originally assigned in a verify callout, is re-used.
692 PP/06 Drop '.' from @INC in various Perl scripts.
694 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
696 PP/08 Reduce a number of compilation warnings under clang; building with
697 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
698 should be warning-free.
700 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
702 HS/01 Fix portability problems introduced by PP/08 for platforms where
703 realloc(NULL) is not equivalent to malloc() [SunOS et al].
705 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
706 chunk. This allows us to accept broken chunked messages. We need a more
707 general solution here.
709 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
710 already-broken messages in the queue.
712 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
714 JH/19 Fix reference counting bug in routing-generated-address tracking.
720 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
721 supports it and a size is available (ie. the sending peer gave us one).
723 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
724 years of being deprecated). The replacements are the ACLs
725 acl_smtp_mime and acl_not_smtp_mime.
727 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
728 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
729 or even an in-clear connection were permitted. Now, if the host lookup
730 was dnssec and dane was requested then the host is only used if the
731 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
732 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
733 if one fails this test.
734 This means that a poorly-configured remote DNS will make it incommunicado;
735 but it protects against a DNS-interception attack on it.
737 JH/04 Bug 1810: make continued-use of an open smtp transport connection
738 non-noisy when a race steals the message being considered.
740 JH/05 If main configuration option tls_certificate is unset, generate a
741 self-signed certificate for inbound TLS connections.
743 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
744 in rewrites and routers.
746 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
747 and logged a warning sing 4.83; now they are a configuration file error.
749 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
750 (lacking @domain). Apply the same qualification processing as RCPT.
752 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
754 JH/10 Support ${sha256:} applied to a string (as well as the previous
757 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
758 a cutthrough deliver is pending, as we always want to make a connection.
759 This also avoids re-routing the message when later placing the cutthrough
760 connection after a verify cache hit.
761 Do not update it with the verify result either.
763 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
764 when routing results in more than one destination address.
766 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
767 signing (which inhibits the cutthrough capability). Previously only
768 the presence of an option was tested; now an expansion evaluating as
769 empty is permissible (obviously it should depend only on data available
770 when the cutthrough connection is made).
772 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
773 the relevant preceding SMTP command did not note the pipelining mode.
775 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
776 Previously they were not counted.
778 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
779 as one having no matching records. Previously we deferred the message
780 that needed the lookup.
782 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
783 distinguished as "(=".
785 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
786 for missing MX records. Previously it only worked for missing A records.
788 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
790 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
791 after the data-go-ahead and data-ack. Patch from Jason Betts.
793 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
794 even for a "none" policy. Patch from Tony Meyer.
796 JH/22 Fix continued use of a connection for further deliveries. If a port was
797 specified by a router, it must also match for the delivery to be
800 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
801 When one of the recipients of a message was unsuitable for the connection
802 (has no matching addresses), we lost track of needing to mark it
803 deferred. As a result mail would be lost.
805 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
807 JH/25 Decoding ACL controls is now done using a binary search; the source code
808 takes up less space and should be simpler to maintain. Merge the ACL
809 condition decode tables also, with similar effect.
811 JH/26 Fix problem with one_time used on a redirect router which returned the
812 parent address unchanged. A retry would see the parent address marked as
813 delivered, so not attempt the (identical) child. As a result mail would
816 JH/27 Fix a possible security hole, wherein a process operating with the Exim
817 UID can gain a root shell. Credit to http://www.halfdog.net/ for
818 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
821 JH/28 Enable {spool,log} filesystem space and inode checks as default.
822 Main config options check_{log,spool}_{inodes,space} are now
823 100 inodes, 10MB unless set otherwise in the configuration.
825 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
826 Previously it only applied to the main-section connection policy
829 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
831 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
832 by me. Added RFC7919 DH primes as an alternative.
834 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
835 are not in the system include path.
837 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
838 GnuTLS, when a session startup failed (eg because the client disconnected)
839 Exim did stdio operations after fclose. This was exposed by a recent
840 change which nulled out the file handle after the fclose.
842 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
843 signed directly by the cert-signing cert, rather than an intermediate
844 OCSP-signing cert. This is the model used by LetsEncrypt.
846 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
848 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
849 an incoming connection.
851 HS/02 Bug 1802: Do not half-close the connection after sending a request
854 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
855 fallback to "prime256v1".
857 JH/34 SECURITY: Use proper copy of DATA command in error message.
858 Could leak key material. Remotely exploitable. CVE-2016-9963.
864 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
865 and 3.4.4 - once the server is enabled to respond to an OCSP request
866 it does even when not requested, resulting in a stapling non-aware
867 client dropping the TLS connection.
869 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
870 support variable-length bit vectors. No functional change.
872 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
873 The I= interface field on outgoing lines is now after the H= remote
874 host field, same as incoming lines. There is a separate
875 outgoing_interface log selector which allows you to disable the
878 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
879 If not running log_selector +smtp_connection the mainlog would be held
880 open indefinitely after a "too many connections" event, including to a
881 deleted file after a log rotate. Leave the per net connection logging
882 leaving it open for efficiency as that will be quickly detected by the
883 check on the next write.
885 HS/01 Bug 1671: Fix post transport crash.
886 Processing the wait-<transport> messages could crash the delivery
887 process if the message IDs didn't exist for some reason. When
888 using 'split_spool_directory=yes' the construction of the spool
889 file name failed already, exposing the same netto behaviour.
891 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
892 mime_regex ACL conditions.
894 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
895 to DSN fail messages (bounces): remote IP, remote greeting, remote response
896 to HELO, local diagnostic string.
898 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
899 log line to debug. Even when configured with a tls authenticator many
900 client connections are expected to not authenticate in this way, so
901 an authenticate fail is not an error.
903 HS/02 Add the Exim version string to the process info. This way exiwhat
904 gives some more detail about the running daemon.
906 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
907 matter for fast-change records such as DNSBLs.
909 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
910 retry record, even if constant. There may be multiple transports with
911 different interface settings and the retry behaviour needs to be kept
914 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
916 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
918 JH/10 Bug 840: fix log_defer_output option of pipe transport
920 JH/11 Bug 830: use same host for all RCPTS of a message, even under
921 hosts_randomize. This matters a lot when combined with mua_wrapper.
923 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
924 ${quote_pgsql:<string>} operator.
926 JH/13 Bug 1708: avoid misaligned access in cached lookup.
928 JH/14 Change header file name for freeradius-client. Relevant if compiling
929 with Radius support; from the Gentoo tree and checked under Fedora.
931 JH/15 Bug 1712: Introduce $prdr_requested flag variable
933 JH/16 Bug 1714: Permit an empty string as expansion result for transport
934 option transport_filter, meaning no filtering.
936 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
938 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
939 defaults to "*" (all hosts). The variable is now available when not built
940 with TLS, default unset, mainly to enable keeping the testsuite sane.
941 If a server certificate is not supplied (via tls_certificate) an error is
942 logged, and clients will find TLS connections fail on startup. Presumably
943 they will retry in-clear.
944 Packagers of Exim are strongly encouraged to create a server certificate
945 at installation time.
947 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
948 with the $config_file variable.
950 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
951 in transport context, after the attempt, and per-recipient. The latter type
952 is per host attempted. The event data is the error message, and the errno
953 information encodes the lookup type (A vs. MX) used for the (first) host,
954 and the trailing two digits of the smtp 4xx response.
956 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
957 to write to mainlog (or rejectlog, paniclog) in the window between file
958 creation and permissions/ownership being changed. Particularly affects
959 installations where exicyclog is run as root, rather than exim user;
960 result is that the running daemon panics and dies.
962 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
964 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
965 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
966 "pri" and "weight". Note that the previous implicit priority given by the
967 list order is no longer honoured.
969 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
972 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
973 by defining SUPPORT_SOCKS.
975 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
976 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
977 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
978 variables are renamed to proxy_{local,external}_{address,port}.
980 JH/27 Move Internationalisation support from Experimental to mainline, enabled
981 for a build by defining SUPPORT_I18N
983 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
984 of the query string, and make ${quote_redis:} do that quoting.
986 JH/29 Move Events support from Experimental to mainline, enabled by default
987 and removable for a build by defining DISABLE_EVENT.
989 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
991 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
992 cached by the daemon.
994 JH/32 Move Redis support from Experimental to mainline, enabled for a build
995 by defining LOOKUP_REDIS. The libhiredis library is required.
997 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
998 keys are given for lookup.
1000 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1001 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1002 only supported when built with TLS support. The PolarSSL SHA routines
1003 are still used when the TLS library is too old for convenient support.
1005 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1006 openssl_options), for security. OpenSSL forces this from version 1.1.0
1007 server-side so match that on older versions.
1009 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1010 allocation for $value could be released as the expansion processing
1011 concluded, but leaving the global pointer active for it.
1013 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1014 and to use the domains and local_parts ACL conditions.
1016 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1017 incorrectly not doubled on cutthrough transmission, hence seen as a
1018 body-termination at the receiving system - resulting in truncated mails.
1019 Commonly the sender saw a TCP-level error, and retransmitted the message
1020 via the normal store-and-forward channel. This could result in duplicates
1021 received - but deduplicating mailstores were liable to retain only the
1022 initial truncated version.
1024 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1026 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1028 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1029 we're in there, support oversigning also; bug 1309.
1031 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1033 HS/04 Add support for keep_environment and add_environment options.
1035 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1036 either intentional arithmetic overflow during PRNG, or testing config-
1039 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1040 delivery resulted in actual delivery. Cancel cutthrough before DATA
1043 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1044 rejects a recipient: pass the reject to the originator.
1046 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1047 Many were false-positives and ignorable, but it's worth fixing the
1050 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1051 for the new environment-manipulation done at startup. Move the routines
1052 from being local to tls.c to being global via the os.c file.
1054 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1055 an extract embedded as result-arg for a map, the first arg for extract
1056 is unavailable so we cannot tell if this is a numbered or keyed
1057 extraction. Accept either.
1063 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1066 JH/02 The smtp transport option "multi_domain" is now expanded.
1068 JH/03 The smtp transport now requests PRDR by default, if the server offers
1071 JH/04 Certificate name checking on server certificates, when exim is a client,
1072 is now done by default. The transport option tls_verify_cert_hostnames
1073 can be used to disable this per-host. The build option
1074 EXPERIMENTAL_CERTNAMES is withdrawn.
1076 JH/05 The value of the tls_verify_certificates smtp transport and main options
1077 default to the word "system" to access the system default CA bundle.
1078 For GnuTLS, only version 3.0.20 or later.
1080 JH/06 Verification of the server certificate for a TLS connection is now tried
1081 (but not required) by default. The verification status is now logged by
1082 default, for both outbound TLS and client-certificate supplying inbound
1085 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1088 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1089 Status Notification (bounce) messages are now MIME format per RFC 3464.
1090 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1091 under the control of the dsn_advertise_hosts option, and routers may
1092 have a dsn_lasthop option.
1094 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1095 default, modifiable by a malware= option. The list separator for
1096 the options can now be changed in the usual way. Bug 68.
1098 JH/10 The smtp_receive_timeout main option is now expanded before use.
1100 JH/11 The incoming_interface log option now also enables logging of the
1101 local interface on delivery outgoing connections.
1103 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1104 if the interface and destination host and port all match.
1106 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1109 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1110 Patch from Andrew Lewis.
1112 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1113 now supports optional time-restrictions, weighting, and priority
1114 modifiers per server. Patch originally by <rommer@active.by>.
1116 JH/16 The spamd_address main option now supports a mixed list of local
1117 and remote servers. Remote servers can be IPv6 addresses, and
1118 specify a port-range.
1120 JH/17 Bug 68: The spamd_address main option now supports an optional
1121 timeout value per server.
1123 JH/18 Bug 1581: Router and transport options headers_add/remove can
1124 now have the list separator specified.
1126 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1129 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1132 JH/21 Support for the A6 type of dns record is withdrawn.
1134 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1135 rather than the verbs used.
1137 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1138 from 255 to 1024 chars.
1140 JH/24 Verification callouts now attempt to use TLS by default.
1142 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1143 are generic router options now. The defaults didn't change.
1145 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1146 Original patch from Alexander Shikoff, worked over by JH.
1148 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1149 files if ZCAT_COMMAND is not executable.
1151 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1153 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1155 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1156 Normally benign, it bites when the pair was led to by a CNAME;
1157 modern usage is to not canonicalize the domain to a CNAME target
1158 (and we were inconsistent anyway for A-only vs AAAA+A).
1160 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1162 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1163 when evaluating $sender_host_dnssec.
1165 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1166 $sender_helo_dnssec variable.
1168 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1170 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1172 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1174 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1175 documented as working, but never had. Support all but $spam_report.
1177 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1178 added for tls authenticator.
1180 HS/03 Add perl_taintmode main config option
1186 TL/01 When running the test suite, the README says that variables such as
1187 no_msglog_check are global and can be placed anywhere in a specific
1188 test's script, however it was observed that placement needed to be near
1189 the beginning for it to behave that way. Changed the runtest perl
1190 script to read through the entire script once to detect and set these
1191 variables, reset to the beginning of the script, and then run through
1192 the script parsing/test process like normal.
1194 TL/02 The BSD's have an arc4random API. One of the functions to induce
1195 adding randomness was arc4random_stir(), but it has been removed in
1196 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1197 function when detected.
1199 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1200 cause callback expansion.
1202 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1203 syntax errors in an expansion can be treated as a string instead of
1204 logging or causing an error, due to the internal use of bool_lax
1205 instead of bool when processing it.
1207 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1208 server certificates when making smtp deliveries.
1210 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1212 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1214 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1216 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1217 Merged patch from Sebastian Wiedenroth.
1219 JH/05 Fix results-pipe from transport process. Several recipients, combined
1220 with certificate use, exposed issues where response data items split
1221 over buffer boundaries were not parsed properly. This eventually
1222 resulted in duplicates being sent. This issue only became common enough
1223 to notice due to the introduction of connection certificate information,
1224 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1226 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1227 size buffer was used, resulting in syntax errors when an expansion
1230 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1231 version 3.3.6 or later.
1233 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1234 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1235 both become "event_action", the variables become $event_name, $event_data
1236 and $event_defer_errno. There is a new variable $verify_mode, usable in
1237 routers, transports and related events. The tls:cert event is now also
1238 raised for inbound connections, if the main configuration event_action
1241 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1242 early OCSP support, but no stapling (appears to be less than 1.0.0).
1244 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1245 server certificate names available under the smtp transport option
1246 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1249 JH/10 Time-related extraction expansions from certificates now use the main
1250 option "timezone" setting for output formatting, and are consistent
1251 between OpenSSL and GnuTLS compilations. Bug 1541.
1253 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1254 encoded parameter in the incoming message. Bug 1558.
1256 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1257 include certificate info, eximon was claiming there were spoolfile
1260 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1262 JH/14 Log delivery-related information more consistently, using the sequence
1263 "H=<name> [<ip>]" wherever possible.
1265 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1266 are problematic for Debian distribution, omit them from the release
1269 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1271 JH/16 Fix string representation of time values on 64bit time_t architectures.
1274 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1275 output list separator was used.
1280 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1281 checkers that were complaining about end of non-void function with no
1284 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1285 This was a regression introduced in 4.83 by another bugfix.
1287 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1289 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1290 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1296 TF/01 Correctly close the server side of TLS when forking for delivery.
1298 When a message was received over SMTP with TLS, Exim failed to clear up
1299 the incoming connection properly after forking off the child process to
1300 deliver the message. In some situations the subsequent outgoing
1301 delivery connection happened to have the same fd number as the incoming
1302 connection previously had. Exim would try to use TLS and fail, logging
1303 a "Bad file descriptor" error.
1305 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1306 utilities have not been installed.
1308 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1309 temporary space as the ACL may create new global variables.
1311 TL/01 LDAP support uses per connection or global context settings, depending
1312 upon the detected version of the libraries at build time.
1314 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1315 to extract and use the src ip:port in logging and expansions as if it
1316 were a direct connection from the outside internet. PPv2 support was
1317 updated based on HAProxy spec change in May 2014.
1319 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1321 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1322 Properly escape header and check for NULL return.
1324 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1327 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1329 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1330 characters in header names, implemented as a verify condition.
1331 Contributed by Michael Fischer v. Mollard.
1333 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1334 results permerror and temperror. Previous values are deprecated but
1335 still accepted. In a future release, err_perm and err_temp will be
1336 completely removed, which will be a backward incompatibility if the
1337 ACL tests for either of these two old results. Patch contributed by
1338 user bes-internal on the mailing list.
1340 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1342 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1343 selectors, in both main and reject logs.
1345 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1348 JH/07 Add malware type "sock" for talking to simple daemon.
1350 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1352 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1353 routers/transports under cutthrough routing.
1355 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1356 numbers. Touch up "bool" conditional to keep the same definition.
1358 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1360 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1362 JH/12 Expand items in router/transport headers_add or headers_remove lists
1363 individually rather than the list as a whole. Bug 1452.
1365 Required for reasonable handling of multiple headers_ options when
1366 they may be empty; requires that headers_remove items with embedded
1367 colons must have them doubled (or the list-separator changed).
1369 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1370 view the policy declared in the DMARC record. Currently, $dmarc_status
1371 is a combined value of both the record presence and the result of the
1374 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1376 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1377 dnslookup router and the smtp transport (applying to the forward
1380 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1381 of ldap servers used for a specific lookup. Patch provided by Heiko
1384 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1385 New variable $lookup_dnssec_authenticated for observability.
1387 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1388 Patch submitted by Lars Timman.
1390 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1392 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1393 Requires trusted mode and valid format message id, aborts otherwise.
1394 Patch contributed by Heiko Schlichting.
1396 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1397 certextract with support for various fields. Bug 1358.
1399 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1400 is requested by default, modifiable by smtp transport option
1403 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1404 operate on certificate variables to give certificate fingerprints
1405 Also new ${sha256:cert_variable}.
1407 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1409 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1410 Christian Aistleitner.
1412 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1414 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1415 file. Patch from Wolfgang Breyha.
1417 JH/25 Expand the coverage of the delivery $host and $host_address to
1418 client authenticators run in verify callout. Bug 1476.
1420 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1421 align with daemon_smtp_ports. Bug 72.
1423 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1424 support and error reporting did not work properly.
1426 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1427 and is readable. Patch from Andrew Colin Kissa.
1429 TL/14 Enhance documentation of ${run expansion and how it parses the
1430 commandline after expansion, particularly in the case when an
1431 unquoted variable expansion results in an empty value.
1433 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1435 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1436 by renaming away from it. Addresses GH issue 15, reported by
1439 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1441 TL/15 SECURITY: prevent double expansion in math comparison functions
1442 (can expand unsanitized data). Not remotely exploitable.
1449 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1451 PP/02 Make -n do something, by making it not do something.
1452 When combined with -bP, the name of an option is not output.
1454 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1457 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1458 $sender_host_name and config options to manage this, and basic check
1461 PP/05 DSCP support for outbound connections and control modifier for inbound.
1463 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1464 (Only plugin which currently uses this is kerberos4, which nobody should
1465 be using, but we should make it available and other future plugins might
1466 conceivably use it, even though it would break NAT; stuff *should* be
1467 using channel bindings instead).
1469 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1470 name; added for Sendmail compatibility; requires admin caller.
1471 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1472 just ignore it); requires trusted caller.
1473 Also parse but ignore: -Ac -Am -X<logfile>
1476 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1478 TL/02 Add +smtp_confirmation as a default logging option.
1480 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1481 Patch by Magnus Holmgren from 2007-02-20.
1483 TL/04 Bugzilla 1281 - Spec typo.
1484 Bugzilla 1283 - Spec typo.
1485 Bugzilla 1290 - Spec grammar fixes.
1487 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1489 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1491 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1492 mailing list by by Dmitry Isaikin.
1494 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1496 JH/02 Support "G" suffix to numbers in ${if comparisons.
1498 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1500 NM/01 Bugzilla 1197 - Spec typo
1501 Bugzilla 1196 - Spec examples corrections
1503 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1505 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1506 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1509 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1510 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1512 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1513 "acl {{name}{arg}...}", and optional args on acl condition
1516 JH/05 Permit multiple router/transport headers_add/remove lines.
1518 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1520 JH/07 Avoid using a waiting database for a single-message-only transport.
1521 Performance patch from Paul Fisher. Bugzilla 1262.
1523 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1526 JH/09 Add $headers_added variable, with content from use of ACL modifier
1527 add_header (but not yet added to the message). Bugzilla 199.
1529 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1530 Pulled from Bugzilla 817 by Wolfgang Breyha.
1532 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1534 (nb: this is the same fix as in Exim 4.80.1)
1536 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1539 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1542 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1544 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1545 advertise SMTP AUTH mechanism to us, instead of a generic
1546 protocol violation error. Also, make Exim more robust to bad
1547 data from the Dovecot auth socket.
1549 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1551 When a queue runner is handling a message, Exim first routes the
1552 recipient addresses, during which it prunes them based on the retry
1553 hints database. After that it attempts to deliver the message to
1554 any remaining recipients. It then updates the hints database using
1557 So if a recipient address works intermittently, it can get repeatedly
1558 deferred at routing time. The retry hints record remains fresh so the
1559 address never reaches the final cutoff time.
1561 This is a fairly common occurrence when a user is bumping up against
1562 their storage quota. Exim had some logic in its local delivery code
1563 to deal with this. However it did not apply to per-recipient defers
1564 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1566 This change adds a proper retry rule check during routing so that the
1567 final cutoff time is checked against the message's age. We only do
1568 this check if there is an address retry record and there is not a
1569 domain retry record; this implies that previous attempts to handle
1570 the address had the retry_use_local_parts option turned on. We use
1571 this as an approximation for the destination being like a local
1572 delivery, as in LMTP.
1574 I suspect this new check makes the old local delivery cutoff check
1575 redundant, but I have not verified this so I left the code in place.
1577 TF/02 Correct gecos expansion when From: is a prefix of the username.
1579 Test 0254 submits a message to Exim with the header
1583 When I ran the test suite under the user fanf2, Exim expanded
1584 the header to contain my full name, whereas it should have added
1585 a Resent-Sender: header. It erroneously treats any prefix of the
1586 username as equal to the username.
1588 This change corrects that bug.
1590 GF/01 DCC debug and logging tidyup
1591 Error conditions log to paniclog rather than rejectlog.
1592 Debug lines prefixed by "DCC: " to remove any ambiguity.
1594 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1596 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1597 Bug spotted by Jeremy Harris; was flawed since initial commit.
1598 Would have resulted in OCSP responses post-SNI triggering an Exim
1599 NULL dereference and crash.
1601 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1603 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1604 Bug detection, analysis and fix by Samuel Thibault.
1605 Bugzilla 1331, Debian bug #698092.
1607 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1609 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1610 Server implementation by Todd Lyons, client by JH.
1611 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1612 config variable "prdr_enable" controls whether the server
1613 advertises the facility. If the client requests PRDR a new
1614 acl_data_smtp_prdr ACL is called once for each recipient, after
1615 the body content is received and before the acl_smtp_data ACL.
1616 The client is controlled by both of: a hosts_try_prdr option
1617 on the smtp transport, and the server advertisement.
1618 Default client logging of deliveries and rejections involving
1619 PRDR are flagged with the string "PRDR".
1621 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1622 fclose(). Diagnosis by Todd Lyons.
1624 PP/17 Update configure.default to handle IPv6 localhost better.
1625 Patch by Alain Williams (plus minor tweaks).
1628 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1629 This is now consistent with GnuTLS, and is now documented: the
1630 previous undocumented portable approach to treating the option as
1631 unset was to force an expansion failure. That still works, and
1632 an empty string is now equivalent.
1634 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1635 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1636 not performing validation itself.
1638 PP/20 Added force_command boolean option to pipe transport.
1639 Patch from Nick Koston, of cPanel Inc.
1641 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1644 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1646 PP/21 Fix eximon continuous updating with timestamped log-files.
1647 Broken in a format-string cleanup in 4.80, missed when I repaired the
1648 other false fix of the same issue.
1649 Report and fix from Heiko Schlichting.
1652 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1653 Report from Prashanth Katuri.
1655 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1656 It's SecureTransport, so affects any MacOS clients which use the
1657 system-integrated TLS libraries, including email clients.
1659 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1660 using a MIME ACL for non-SMTP local injection.
1661 Report and assistance in diagnosis by Warren Baker.
1663 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1665 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1667 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1668 last id that failed so it may be referenced in subsequent ACL's.
1670 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1673 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1674 ldap library initialization, allowing self-signed CA's to be
1675 used. Also properly sets require_cert option later in code by
1676 using NULL (global ldap config) instead of ldap handle (per
1677 session). Bug diagnosis and testing by alxgomz.
1679 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1680 the src/util/ subdirectory.
1682 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1683 renamed to Transport Post Delivery Action by Jeremy Harris, as
1686 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1687 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1688 redis_servers = needs to be configured which will be used by the redis
1689 lookup. Patch from Warren Baker, of The Packet Hub.
1691 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1693 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1694 hostname or reverse DNS when processing a host list. Used suggestions
1695 from multiple comments on this bug.
1697 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1699 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1700 Missed a few lines, added it to make the runtest require no keyboard
1703 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1704 contains upper case chars. Make router use caseful_local_part.
1706 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1707 support when GnuTLS has been built with p11-kit.
1713 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1715 This, or similar/improved, will also be change PP/11 of 4.82.
1721 PP/01 Handle short writes when writing local log-files.
1722 In practice, only affects FreeBSD (8 onwards).
1723 Bugzilla 1053, with thanks to Dmitry Isaikin.
1725 NM/01 Bugzilla 949 - Documentation tweak
1727 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1730 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1732 PP/02 Implemented gsasl authenticator.
1734 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1736 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1737 `pkg-config foo` for cflags/libs.
1739 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1740 with rest of GSASL and with heimdal_gssapi.
1742 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1743 `pkg-config foo` for cflags/libs for the TLS implementation.
1745 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1746 properties get this fed in as external SSF. A number of robustness
1747 and debugging improvements to the cyrus_sasl authenticator.
1749 PP/08 cyrus_sasl server now expands the server_realm option.
1751 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1752 Patch by Jeremy Harris.
1754 PP/10 Added dbmjz lookup type.
1756 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1758 PP/12 MAIL args handles TAB as well as SP, for better interop with
1759 non-compliant senders.
1760 Analysis and variant patch by Todd Lyons.
1762 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1763 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1764 Patch from Dirk Mueller <dmueller@suse.com>
1766 PP/13 tls_peerdn now print-escaped for spool files.
1767 Observed some $tls_peerdn in wild which contained \n, which resulted
1768 in spool file corruption.
1770 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1771 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1772 or write after TLS renegotiation, which otherwise led to messages
1775 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1776 as a tracking header (ie: a signed header comes before the signature).
1777 Patch from Wolfgang Breyha.
1779 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1780 comma-sep list; embedded commas doubled.
1782 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1784 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1786 Report and patch from Dmitry Banschikov.
1788 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1789 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1790 failures. We appear to now support TLS1.1+ with Exim.
1792 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1793 lets Exim select keys and certificates based upon TLS SNI from client.
1794 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1795 before an outbound SMTP session. New log_selector, +tls_sni.
1797 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1798 NULL dereference. Report and patch from Alun Jones.
1800 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1801 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1802 Not seeing resolver debug output on NetBSD, but suspect this is a
1803 resolver implementation change.
1805 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1806 Left warnings. Added "eximon gdb" invocation mode.
1808 PP/21 Defaulting "accept_8bitmime" to true, not false.
1810 PP/22 Added -bw for inetd wait mode support.
1812 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1813 locate the relevant includes and libraries. Made this the default.
1815 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1816 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1818 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1819 This may cause build issues on older platforms.
1821 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1822 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1823 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1824 Added SNI support via GnuTLS too.
1825 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1827 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1829 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1830 Applied second patch from Janne, implementing suggestion to default
1831 multiple-strings-in-record handling to match SPF spec.
1833 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1835 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1836 read-only, out of scope).
1837 Patch from Wolfgang Breyha, report from Stuart Northfield.
1839 PP/29 Fix three issues highlighted by clang analyser static analysis.
1840 Only crash-plausible issue would require the Cambridge-specific
1841 iplookup router and a misconfiguration.
1842 Report from Marcin Mirosław.
1844 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1846 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1847 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1848 As part of this, removing so much warning spew let me fix some minor
1849 real issues in debug logging.
1851 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1852 assignment on my part. Fixed.
1854 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1855 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1856 Janne Snabb (who went above and beyond: thank you).
1858 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1859 string otherwise requires a connection and a bunch more work and it's
1860 relatively easy to get wrong. Should also expose TLS library linkage
1863 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1864 64-bit ${eval} (JH/03).
1866 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1867 GNU libc to support some of the 64-bit stuff, should not lead to
1868 conflicts. Defined before os.h is pulled in, so if a given platform
1869 needs to override this, it can.
1871 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1872 protection layer was required, which is not implemented.
1873 Bugzilla 1254, patch from Wolfgang Breyha.
1875 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1876 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1877 tls_dhparam take prime identifiers. Also unbreak combination of
1878 OpenSSL+DH_params+TLSSNI.
1880 PP/39 Disable SSLv2 by default in OpenSSL support.
1886 PP/01 Solaris build fix for Oracle's LDAP libraries.
1887 Bugzilla 1109, patch from Stephen Usher.
1889 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1891 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1894 TF/02 Fix a couple more cases where we did not log the error message
1895 when unlink() failed. See also change 4.74-TF/03.
1897 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1898 lock up or crash if it happened to be inside a call to libc when it
1899 got a SIGUSR1 from exiwhat.
1901 The SIGUSR1 handler appends the current process status to the process
1902 log which is later printed by exiwhat. It used to use the general
1903 purpose logging code to do this, but several functions it calls are
1904 not safe for signals.
1906 The new output code in the SIGUSR1 handler is specific to the process
1907 log, and simple enough that it's easy to inspect for signal safety.
1908 Removing some special cases also simplifies the general logging code.
1909 Removing the spurious timestamps from the process log simplifies
1912 TF/04 Improved ratelimit ACL condition.
1914 The /noupdate option has been deprecated in favour of /readonly which
1915 has clearer semantics. The /leaky, /strict, and /readonly update modes
1916 are mutually exclusive. The update mode is no longer included in the
1917 database key; it just determines when the database is updated. (This
1918 means that when you upgrade Exim will forget old rate measurements.)
1920 Exim now checks that the per_* options are used with an update mode that
1921 makes sense for the current ACL. For example, when Exim is processing a
1922 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1923 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1924 must specify per_mail/readonly. If you omit the update mode it defaults to
1925 /leaky where that makes sense (as before) or /readonly where required.
1927 The /noupdate option is now undocumented but still supported for
1928 backwards compatibility. It is equivalent to /readonly except that in
1929 ACLs where /readonly is required you may specify /leaky/noupdate or
1930 /strict/noupdate which are treated the same as /readonly.
1932 A useful new feature is the /count= option. This is a generalization
1933 of the per_byte option, so that you can measure the throughput of other
1934 aggregate values. For example, the per_byte option is now equivalent
1935 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1937 The per_rcpt option has been generalized using the /count= mechanism
1938 (though it's more complicated than the per_byte equivalence). When it is
1939 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1940 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1941 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1942 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1943 Note that using per_rcpt with a non-readonly update mode in more than
1944 one ACL will cause the recipients to be double-counted. (The per_mail
1945 and per_byte options don't have this problem.)
1947 The handling of very low rates has changed slightly. If the computed rate
1948 is less than the event's count (usually one) then this event is the first
1949 after a long gap. In this case the rate is set to the same as this event's
1950 count, so that the first message of a spam run is counted properly.
1952 The major new feature is a mechanism for counting the rate of unique
1953 events. The new per_addr option counts the number of different
1954 recipients that someone has sent messages to in the last time period. It
1955 behaves like per_rcpt if all the recipient addresses are different, but
1956 duplicate recipient addresses do not increase the measured rate. Like
1957 the /count= option this is a general mechanism, so the per_addr option
1958 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1959 example, measure the rate that a client uses different sender addresses
1960 with the options per_mail/unique=$sender_address. There are further
1961 details in the main documentation.
1963 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1965 TF/06 Removed a few PCRE remnants.
1967 TF/07 Automatically extract Exim's version number from tags in the git
1968 repository when doing development or release builds.
1970 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1971 Bugzilla 879. Patch from Paul Fisher.
1973 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1974 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1977 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1979 PP/05 Variable $av_failed, true if the AV scanner deferred.
1980 Bugzilla 1078. Patch from John Horne.
1982 PP/06 Stop make process more reliably on build failure.
1983 Bugzilla 1087. Patch from Heiko Schlittermann.
1985 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1986 Bugzilla 1089. Patch from Heiko Schlittermann.
1988 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1989 Bugzilla 1131. Patch from Holger Weiß.
1991 PP/09 Handle IPv6 addresses with SPF.
1992 Bugzilla 860. Patch from Wolfgang Breyha.
1994 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1996 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1999 PP/11 match_* no longer expand right-hand-side by default.
2000 New compile-time build option, EXPAND_LISTMATCH_RHS.
2001 New expansion conditions, "inlist", "inlisti".
2003 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2005 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2007 PP/14 fix log_write() format string regression from TF/03.
2008 Bugzilla 1152. Patch from Dmitry Isaikin.
2014 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2016 PP/02 Harmonised TLS library version reporting; only show if debugging.
2017 Layout now matches that introduced for other libraries in 4.74 PP/03.
2019 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2021 PP/04 New "dns_use_edns0" global option.
2023 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2026 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2027 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2029 TK/01 Updated PolarSSL code to 0.14.2.
2030 Bugzilla 1097. Patch from Andreas Metzler.
2032 PP/07 Catch divide-by-zero in ${eval:...}.
2033 Fixes bugzilla 1102.
2035 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2038 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2039 format-string attack -- SECURITY: remote arbitrary code execution.
2041 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2042 time unintentionally subject to list matching rules, letting the header
2043 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2044 arbitrary string expansion). This allowed for information disclosure.
2046 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2047 INT_MIN/-1 -- value coerced to INT_MAX.
2053 NM/01 Workaround for PCRE version dependency in version reporting
2056 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2057 This fixes portability to compilers other than gcc, notably
2058 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2060 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2061 makefiles for portability to HP-UX and POSIX correctness.
2063 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2064 Also via indented variable definition in the Makefile.
2065 (Debugging by Oliver Heesakkers).
2067 PP/02 Restore caching of spamd results with expanded spamd_address.
2068 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2070 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2071 Improves build reliability. Fix from: Frank Elsner
2073 NM/02 Fix wide character breakage in the rfc2047 coding
2074 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2076 NM/03 Allow underscore in dnslist lookups
2077 Fixes bug 1026. Patch from Graeme Fowler
2079 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2080 Code patches from Adam Ciarcinski of NetBSD.
2082 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2085 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2086 is logged, to avoid truncation. Patch from John Horne.
2088 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2089 Patch from Jakob Hirsch.
2091 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2092 SQL string expansion failure details.
2093 Patch from Andrey Oktyabrski.
2095 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2096 Patch from Simon Arlott.
2098 PP/09 New lookups functionality failed to compile on old gcc which rejects
2099 extern declarations in function scope.
2100 Patch from Oliver Fleischmann
2102 PP/10 Use sig_atomic_t for flags set from signal handlers.
2103 Check getgroups() return and improve debugging.
2104 Fixed developed for diagnosis in bug 927 (which turned out to be
2107 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2108 Patch from Mark Zealey.
2110 PP/12 Bugzilla 1056: Improved spamd server selection.
2111 Patch from Mark Zealey.
2113 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2114 Based on patch from Heiko Schlittermann.
2116 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2117 Patch from Uwe Doering, sign-off by Michael Haardt.
2119 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2120 variable declaration deep within a block. Bug and patch from
2123 PP/15 lookups-Makefile IRIX compatibility coercion.
2125 PP/16 Make DISABLE_DKIM build knob functional.
2127 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2128 Patch by Simon Arlott
2130 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2131 variable argument macros. Our copy now differs from upstream.
2137 TF/01 Failure to get a lock on a hints database can have serious
2138 consequences so log it to the panic log.
2140 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2141 controlled using the smtp_confirmation log selector.
2143 TF/03 Include the error message when we fail to unlink a spool file.
2145 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2146 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2147 for maintaining out-of-tree patches for some time.
2149 PP/01 Bugzilla 139: Documentation and portability issues.
2150 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2151 Handle per-OS dynamic-module compilation flags.
2153 PP/02 Let /dev/null have normal permissions.
2154 The 4.73 fixes were a little too stringent and complained about the
2155 permissions on /dev/null. Exempt it from some checks.
2156 Reported by Andreas M. Kirchwitz.
2158 PP/03 Report version information for many libraries, including
2159 Exim version information for dynamically loaded libraries. Created
2160 version.h, now support a version extension string for distributors
2161 who patch heavily. Dynamic module ABI change.
2163 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2164 privilege escalation vulnerability whereby the Exim run-time user
2165 can cause root to append content of the attacker's choosing to
2168 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2171 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2172 If dropping privileges for untrusted macros, we disabled normal logging
2173 on the basis that it would fail; for the Exim run-time user, this is not
2174 the case, and it resulted in successful deliveries going unlogged.
2175 Fixed. Reported by Andreas Metzler.
2181 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2182 only prepend for the Resent-* case. Fixes regression introduced in
2183 Exim 4.70 by NM/22 for Bugzilla 607.
2185 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2186 increasing numbers of administrators be bitten by this.
2188 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2190 PP/03 Bugzilla 994: added openssl_options main configuration option.
2192 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2194 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2196 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2198 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2199 without a peer certificate, leading to a segfault because of an
2200 assumption that peers always have certificates. Be a little more
2201 paranoid. Problem reported by Martin Tscholak.
2203 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2204 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2205 NB: ClamAV planning to remove STREAM in "middle of 2010".
2206 CL also introduces -bmalware, various -d+acl logging additions and
2207 more caution in buffer sizes.
2209 PP/09 Implemented reverse_ip expansion operator.
2211 PP/10 Bugzilla 937: provide a "debug" ACL control.
2213 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2215 PP/12 Bugzilla 973: Implement --version.
2217 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2219 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2221 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2223 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2224 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2225 ignore trailing whitespace.
2227 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2229 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2232 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2233 Notification from Dr Andrew Aitchison.
2235 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2236 ExtendedDetectionInfo response format.
2237 Notification from John Horne.
2239 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2242 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2243 XSL and documented dependency on system catalogs, with examples of how
2246 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2249 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2250 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2251 configuration file which is writeable by the Exim user or group.
2253 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2254 of configuration files to cover files specified with the -C option if
2255 they are going to be used with root privileges, not just the default
2258 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2259 option (effectively making it always true).
2261 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2262 files to be used while preserving root privileges.
2264 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2265 that rogue child processes cannot use them.
2267 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2268 run-time user, instead of root.
2270 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2271 Exim run-time user without dropping privileges.
2273 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2274 result string, instead of calling string_vformat() twice with the same
2277 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2278 for other users. Others should always drop root privileges if they use
2279 -C on the command line, even for a whitelisted configure file.
2281 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2283 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2289 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2290 $data_path, and $header_path variables; fixed documentation bugs and
2293 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2294 exipick to access non-standard spools, including the "frozen" queue
2297 NM/01 Bugzilla 965: Support mysql stored procedures.
2298 Patch from Alain Williams
2300 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2302 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2303 Patch from Andreas Metzler
2305 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2306 Patch from Kirill Miazine
2308 NM/05 Bugzilla 671: Added umask to procmail example.
2310 JJ/03 installed exipick 20100323.0, fixing doc bug
2312 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2313 directory. Notification and patch from Dan Rosenberg.
2315 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2317 TK/02 Improve log output when DKIM signing operation fails.
2319 MH/01 Treat the transport option dkim_domain as a colon separated
2320 list, not as a single string, and sign the message with each element,
2321 omitting multiple occurences of the same signer.
2323 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2324 Bugzilla 985, 986. Patch by Simon Arlott
2326 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2327 Patch by Simon Arlott
2329 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2330 MBX locking. Notification from Dan Rosenberg.
2336 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2338 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2340 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2342 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2344 NM/04 Bugzilla 915: Flag broken perl installation during build.
2350 TK/01 Added patch by Johannes Berg that expands the main option
2351 "spamd_address" if it starts with a dollar sign.
2353 TK/02 Write list of recipients to X-Envelope-Sender header when building
2354 the mbox-format spool file for content scanning (suggested by Jakob
2357 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2358 (http://www.dcc-servers.net/) support via dccifd. Activated by
2359 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2361 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2362 by Mark Daniel Reidel <mr@df.eu>.
2364 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2365 When building exim an external PCRE library is now needed -
2366 PCRE is a system library on the majority of modern systems.
2367 See entry on PCRE_LIBS in EDITME file.
2369 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2370 conversation. Added nologin parameter to request.
2371 Patch contributed by Kirill Miazine.
2373 TF/01 Do not log submission mode rewrites if they do not change the address.
2375 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2377 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2378 log files in place. Contributed by Roberto Lima.
2380 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2382 TF/03 Bugzilla 615: When checking the local_parts router precondition
2383 after a local_part_suffix or local_part_prefix option, Exim now
2384 does not use the address's named list lookup cache, since this
2385 contains cached lookups for the whole local part.
2387 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2388 Robert Millan. Documentation is in experimental-spec.txt.
2390 TF/04 Bugzilla 668: Fix parallel build (make -j).
2392 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2394 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2395 Patch provided by Jan Srzednicki.
2397 TF/05 Leading white space used to be stripped from $spam_report which
2398 wrecked the formatting. Now it is preserved.
2400 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2401 that they are available at delivery time.
2403 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2405 TF/08 TLS error reporting now respects the incoming_interface and
2406 incoming_port log selectors.
2408 TF/09 Produce a more useful error message if an SMTP transport's hosts
2409 setting expands to an empty string.
2411 NM/06 Bugzilla 744: EXPN did not work under TLS.
2412 Patch provided by Phil Pennock.
2414 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2415 Patch provided by Richard Godbee.
2417 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2418 acl_smtp_notquit, added index entry.
2420 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2421 Patch provided by Eugene Bujak.
2423 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2424 accept(). Patch provided by Maxim Dounin.
2426 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2427 Patch provided by Phil Pennock.
2429 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2431 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2432 Patch provided by Brad "anomie" Jorsch.
2434 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2435 Patch provided by Dean Brooks.
2437 TK/05 Add native DKIM support (does not depend on external libraries).
2439 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2440 Patch provided by Graeme Fowler.
2442 NM/16 Bugzilla 851: Documentation example syntax fix.
2444 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2446 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2449 NM/19 Bugzilla 745: TLS version reporting.
2450 Patch provided by Phil Pennock.
2452 NM/20 Bugzilla 167: bool: condition support.
2453 Patch provided by Phil Pennock.
2455 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2456 clients. Patch provided by Phil Pennock.
2458 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2459 Patch provided by Brad "anomie" Jorsch.
2461 NM/23 Bugzilla 687: Fix misparses in eximstats.
2462 Patch provided by Heiko Schlittermann.
2464 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2465 Patch provided by Heiko Schlittermann.
2467 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2468 plus update to original patch.
2470 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2472 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2473 Patch provided by David Brownlee.
2475 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2477 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2479 NM/30 Bugzilla 888: TLS documentation bugfixes.
2481 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2483 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2484 Unlike the original bugzilla I have changed all shell scripts in src tree.
2486 NM/33 Bugzilla 898: Transport filter timeout fix.
2487 Patch by Todd Rinaldo.
2489 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2490 Patch by Serge Demonchaux.
2492 NM/35 Bugzilla 39: Base64 decode bug fixes.
2493 Patch by Jakob Hirsch.
2495 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2497 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2499 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2501 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2507 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2508 ALT-N's libdkim that I have put here:
2509 http://duncanthrax.net/exim-experimental/
2511 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2512 were called 'true' and it seems that C99 defines that as a reserved
2513 keyword to be used with 'bool' variable types. That means you could
2514 not include C99-style headers which use bools without triggering
2515 build errors in sieve.c.
2517 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2518 as mailq or other aliases. Changed the --help handling significantly
2519 to do whats expected. exim_usage() emits usage/help information.
2521 SC/01 Added the -bylocaldomain option to eximstats.
2523 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2525 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2527 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2533 PH/01 Another patch from the Sieve maintainer.
2535 PH/02 When an IPv6 address is converted to a string for single-key lookup
2536 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2537 dots are used instead of colons so that keys in lsearch files need not
2538 contain colons. This was done some time before quoting was made available
2539 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2540 (notated using the quote facility) so as to distinguish them from IPv4
2541 keys. This meant that lookups for IP addresses in host lists did not work
2542 for iplsearch lookups.
2544 This has been fixed by arranging for IPv6 addresses to be expressed with
2545 colons if the lookup type is iplsearch. This is not incompatible, because
2546 previously such lookups could never work.
2548 The situation is now rather anomalous, since one *can* have colons in
2549 ordinary lsearch keys. However, making the change in all cases is
2550 incompatible and would probably break a number of configurations.
2552 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2555 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2556 conversion specification without a maximum field width, thereby enabling
2557 a rogue spamd server to cause a buffer overflow. While nobody in their
2558 right mind would setup Exim to query an untrusted spamd server, an
2559 attacker that gains access to a server running spamd could potentially
2560 exploit this vulnerability to run arbitrary code as the Exim user.
2562 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2563 $primary_hostname instead of what libspf2 thinks the hosts name is.
2565 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2566 a directory entry by the name of the lookup key. Previously, if a
2567 symlink pointed to a non-existing file or a file in a directory that
2568 Exim lacked permissions to read, a lookup for a key matching that
2569 symlink would fail. Now it is enough that a matching directory entry
2570 exists, symlink or not. (Bugzilla 503.)
2572 PH/03 The body_linecount and body_zerocount variables are now exported in the
2575 PH/04 Added the $dnslist_matched variable.
2577 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2578 This means they are set thereafter only if the connection becomes
2581 PH/06 Added the client_condition to authenticators so that some can be skipped
2582 by clients under certain conditions.
2584 PH/07 The error message for a badly-placed control=no_multiline_responses left
2585 "_responses" off the end of the name.
2587 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2589 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2590 (without spaces) instead of just copying the configuration text.
2592 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2594 PH/11 Added $max_received_linelength.
2596 PH/12 Added +ignore_defer and +include_defer to host lists.
2598 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2599 way in which PCRE > 7.0 is built.
2601 PH/14 Implemented queue_only_load_latch.
2603 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2604 MAIL command. The effect was to mangle the value on 64-bit systems.
2606 PH/16 Another patch from the Sieve maintainer.
2608 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2610 PH/18 If a system quota error occurred while trying to create the file for
2611 a maildir delivery, the message "Mailbox is full" was not appended to the
2612 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2613 only to a quota excession during the actual writing of the file.
2615 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2616 characters?) which causes problems in log lines. The DN values are now
2617 passed through string_printing() before being added to log lines.
2619 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2620 and InterBase are left for another time.)
2622 PH/21 Added message_body_newlines option.
2624 PH/22 Guard against possible overflow in moan_check_errorcopy().
2626 PH/23 POSIX allows open() to be a macro; guard against that.
2628 PH/24 If the recipient of an error message contained an @ in the local part
2629 (suitably quoted, of course), incorrect values were put in $domain and
2630 $local_part during the evaluation of errors_copy.
2636 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2637 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2640 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2641 issue a MAIL command.
2643 PH/02 In an ACL statement such as
2645 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2647 if a client was not listed at all, or was listed with a value other than
2648 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2649 the condition was not true (as it should be), so access was not denied.
2650 The bug was that the ! inversion was incorrectly passed on to the second
2651 item. This has been fixed.
2653 PH/03 Added additional dnslists conditions == and =& which are different from
2654 = and & when the dns lookup returns more than one IP address.
2656 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2657 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2659 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2660 FSYNC, which compiles an option called disable_fsync that allows for
2661 bypassing fsync(). The documentation is heavily laced with warnings.
2663 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2665 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2666 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2667 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2668 including adding "make clean"; (3) Added -fPIC when compiling the test
2669 dynamically loaded module, to get rid of a warning.
2671 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2672 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2673 = 0s. The bug is otherwise harmless.
2675 PH/07 There was a bug in the dovecot authenticator such that the value of
2676 $auth1 could be overwritten, and so not correctly preserved, after a
2677 successful authentication. This usually meant that the value preserved by
2678 the server_setid option was incorrect.
2680 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2682 PH/09 Installed PCRE release 7.0.
2684 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2685 run for batched SMTP input. It is now run at the start of every message
2686 in the batch. While fixing this I discovered that the process information
2687 (output by running exiwhat) was not always getting set for -bs and -bS
2688 input. This is fixed, and it now also says "batched" for BSMTP.
2690 PH/11 Added control=no_pipelining.
2692 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2693 patch, slightly modified), and move the expansion of helo_data till after
2694 the connection is made in the smtp transport (so it can use these
2697 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2699 PH/14 Added log_selector = +pid.
2701 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2703 PH/16 Add ${if forany and ${if forall.
2705 PH/17 Added dsn_from option to vary the From: line in DSNs.
2707 PH/18 Flush SMTP output before performing a callout, unless control =
2708 no_callout_flush is set.
2710 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2711 was true (the default) a successful delivery failed to delete the retry
2712 item, thus causing premature timeout of the address. The bug is now
2715 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2717 PH/21 Long custom messages for fakedefer and fakereject are now split up
2718 into multiline responses in the same way that messages for "deny" and
2719 other ACL rejections are.
2721 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2722 with slight modification.
2724 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2725 draft, changing the syntax and factoring some duplicate code".
2727 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2728 for deliveries of the second and subsequent messages over the same SMTP
2731 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2732 ${reduce, with only minor "tidies".
2734 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2736 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2737 expansion side effects.
2739 PH/27 When a message times out after an over-quota error from an Exim-imposed
2740 quota, the bounce message says "mailbox is full". This message was not
2741 being given when it was a system quota that was exceeded. It now should
2744 MH/03 Made $recipients available in local_scan(). local_scan() already has
2745 better access to the recipient list through recipients_list[], but
2746 $recipients can be useful in postmaster-provided expansion strings.
2748 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2749 in the case of a MAIL command with additional options following the
2750 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2751 were accidentally chopped off.
2753 PH/29 SMTP synchronization checks are implemented when a command is read -
2754 there is a check that no more input is waiting when there shouldn't be
2755 any. However, for some commands, a delay in an ACL can mean that it is
2756 some time before the response is written. In this time, more input might
2757 arrive, invalidly. So now there are extra checks after an ACL has run for
2758 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2759 pipelining has not been advertised.
2761 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2763 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2764 returned in the SMTP response when smtp_return_error_details was false.
2765 This has been fixed.
2767 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2768 instead of the C I/O that was originally supplied, because problems were
2769 reported on Solaris.
2771 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2772 Exim which did not show up earlier: it was assuming that a call to
2773 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2774 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2775 was a macro that became an assignment, so it seemed to work. This has
2776 changed to a proper function call with a void return, hence the compile
2777 error. Exim's code has been fixed.
2779 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2782 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2784 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2786 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2787 because it turns out to be too big or there is a timeout) there is no
2788 "Completed" line in the log. When some messages of this type were
2789 selected by exigrep, they were listed as "not completed". Others were
2790 picked up by some special patterns. I have improved the selection
2791 criteria to be more general.
2793 PH/38 The host_find_failed option in the manualroute router can now be set
2794 to "ignore", to completely ignore a host whose IP address cannot be
2795 found. If all hosts are ignored, the behaviour is controlled by the new
2796 host_all_ignored option.
2798 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2799 homing or because of multiple MX records with /mx) generated more than
2800 one IP address, and the following item turned out to be the local host,
2801 all the secondary addresses of the first item were incorrectly removed
2802 from the list, along with the local host and any following hosts (which
2803 is what is supposed to happen).
2805 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2806 whoever called Exim into the -H file. In the case of the daemon it was
2807 behaving confusingly. When first started, it used values for whoever
2808 started the daemon, but after a SIGHUP it used the Exim user (because it
2809 calls itself on a restart). I have changed the code so that it now always
2812 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2813 message are rejected with the same error (e.g. no authentication or bad
2814 sender address), and a DATA command is nevertheless sent (as can happen
2815 with PIPELINING or a stupid MUA), the error message that was given to the
2816 RCPT commands is included in the rejection of the DATA command. This is
2817 intended to be helpful for MUAs that show only the final error to their
2820 PH/42 Another patch from the Sieve maintainer.
2822 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2823 Eximstats - Fixed some broken HTML links and added missing column headers
2825 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2826 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2828 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2830 PH/43 Yet another patch from the Sieve maintainer.
2832 PH/44 I found a way to check for a TCP/IP connection going away before sending
2833 the response to the final '.' that terminates a message, but only in the
2834 case where the client has not sent further data following the '.'
2835 (unfortunately, this is allowed). However, in many cases there won't be
2836 any further data because there won't be any more messages to send. A call
2837 to select() can be used: if it shows that the input is "ready", there is
2838 either input waiting, or the socket has been closed. An attempt to read
2839 the next input character can distinguish the two cases. Previously, Exim
2840 would have sent an OK response which the client would never have see.
2841 This could lead to message repetition. This fix should cure that, at
2842 least in a lot of common cases.
2844 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2845 advertised in response to EHLO.
2851 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2852 fixed by 4.65/MH/01 (is this a record?) are fixed:
2854 (i) An empty string was always treated as zero by the numeric comparison
2855 operators. This behaviour has been restored.
2857 (ii) It is documented that the numeric comparison operators always treat
2858 their arguments as decimal numbers. This was broken in that numbers
2859 starting with 0 were being interpreted as octal.
2861 While fixing these problems I realized that there was another issue that
2862 hadn't been noticed. Values of message_size_limit (both the global option
2863 and the transport option) were treated as octal if they started with 0.
2864 The documentation was vague. These values are now always treated as
2865 decimal, and I will make that clear in the documentation.
2871 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2872 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2875 MH/01 Don't check that the operands of numeric comparison operators are
2876 integers when their expansion is in "skipping" mode (fixes bug
2877 introduced by 4.64-PH/07).
2879 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2880 child addresses, Exim now panics and dies. Previously, because the count
2881 is held in a short int, deliveries were likely to be lost. As such a
2882 large number of recipients for a single message is ridiculous
2883 (performance will be very, very poor), I have chosen to impose a limit
2884 rather than extend the field.
2890 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2891 leftover -K file (the existence of which was triggered by #402).
2892 While we were at it, introduced process PID as part of the -K
2893 filename. This should rule out race conditions when creating
2896 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2897 processing considerably. Previous code took too long for large mails,
2898 triggering a timeout which in turn triggers #401.
2900 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2901 in the DK code in transports.c. sendfile() is not really portable,
2902 hence the _LINUX specificness.
2904 TF/01 In the add_headers option to the mail command in an Exim filter,
2905 there was a bug that Exim would claim a syntax error in any
2906 header after the first one which had an odd number of characters
2909 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2910 callout verification, Exim cached a "reject" for the entire domain. This
2911 is correct for most verifications, but it is not correct for a recipient
2912 verification with use_sender or use_postmaster set, because in that case
2913 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2914 case of MAIL FROM:<> rejection from other early rejections (e.g.
2915 rejection of HELO). When verifying a recipient using a non-null MAIL
2916 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2917 Whatever the result of the callout, the value of the domain cache is
2918 left unchanged (for any other kind of callout, getting as far as trying
2919 RCPT means that the domain itself is ok).
2921 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2924 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2925 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2926 session, the daemon ignores ECONNECT errors and logs others; it now
2927 ignores EPIPE as well.
2929 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2930 (quoted-printable decoding).
2932 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2933 later the small subsequent patch to fix an introduced bug.
2935 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2937 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2939 PH/08 An error is now given if message_size_limit is specified negative.
2941 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2942 to be given (somewhat) arbitrary names.
2944 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2947 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2948 miscellaneous code fixes
2950 PH/10 Added the log_reject_target ACL modifier to specify where to log
2953 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2954 hostname. This is wrong, because it relates to the incoming message (and
2955 probably the interface on which it is arriving) and not to the outgoing
2956 callout (which could be using a different interface). This has been
2957 changed to use the value of the helo_data option from the smtp transport
2958 instead - this is what is used when a message is actually being sent. If
2959 there is no remote transport (possible with a router that sets up host
2960 addresses), $smtp_active_hostname is used.
2962 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2963 tweaks were necessary in order to get it to work (see also 21 below):
2964 (a) The code assumed that strncpy() returns a negative number on buffer
2965 overflow, which isn't the case. Replaced with Exim's string_format()
2967 (b) There were several signed/unsigned issues. I just did the minimum
2968 hacking in of casts. There is scope for a larger refactoring.
2969 (c) The code used strcasecmp() which is not a standard C function.
2970 Replaced with Exim's strcmpic() function.
2971 (d) The code set only $1; it now sets $auth1 as well.
2972 (e) A simple test gave the error "authentication client didn't specify
2973 service in request". It would seem that Dovecot has changed its
2974 interface. Fortunately there's a specification; I followed it and
2975 changed what the client sends and it appears to be working now.
2977 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2980 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2981 address A is aliased to B and C, where B exists and C does not. Without
2982 -v the output is "A verified" because verification stops after a
2983 successful redirection if more than one address is generated. However,
2984 with -v the child addresses are also verified. Exim was outputting "A
2985 failed to verify" and then showing the successful verification for C,
2986 with its parentage. It now outputs "B failed to verify", showing B's
2987 parentage before showing the successful verification of C.
2989 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2990 look up a TXT record in a specific list after matching in a combined
2993 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2994 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2995 they consult the DNS. I had assumed they would set it the way they
2996 wanted; and indeed my experiments on Linux seem to show that in some
2997 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2998 To be on the safe side, however, I have now made the interface to
2999 host_find_byname() similar to host_find_bydns(), with an argument
3000 containing the DNS resolver options. The host_find_byname() function now
3001 sets these options at its start, just as host_find_bydns() does. The smtp
3002 transport options dns_qualify_single and dns_search_parents are passed to
3003 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3004 of host_find_byname() use the default settings of RES_DEFNAMES
3005 (qualify_single) but not RES_DNSRCH (search_parents).
3007 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3008 spool_read_header() do less string testing, by means of a preliminary
3009 switch on the second character of optional "-foo" lines. (This is
3010 overdue, caused by the large number of possibilities that now exist.
3011 Originally there were few.) While I was there, I also converted the
3012 str(n)cmp tests so they don't re-test the leading "-" and the first
3013 character, in the hope this might squeeze out yet more improvement.
3015 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3016 flag allowing group syntax was set by the header_syntax check but not
3017 turned off, possible causing trouble later; (2) The flag was not being
3018 set at all for the header_verify test, causing "group"-style headers to
3019 be rejected. I have now set it in this case, and also caused header_
3020 verify to ignore an empty address taken from a group. While doing this, I
3021 came across some other cases where the code for allowing group syntax
3022 while scanning a header line wasn't quite right (mostly, not resetting
3023 the flag correctly in the right place). These bugs could have caused
3024 trouble for malformed header lines. I hope it is now all correct.
3026 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3027 with the "reply" argument non-NULL. The code, however (which originally
3028 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3029 but it didn't always do it. This confused somebody who was copying the
3030 code for some other use. I have removed all the tests.
3032 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3033 feature that was used to support insecure browsers during the U.S. crypto
3034 embargo. It requires special client support, and Exim is probably the
3035 only MTA that supported it -- and would never use it because real RSA is
3036 always available. This code has been removed, because it had the bad
3037 effect of slowing Exim down by computing (never used) parameters for the
3038 RSA_EXPORT functionality.
3040 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3041 authenticator to fail if there's a tab character in the incoming data
3042 (there should never be unless someone is messing about, as it's supposed
3043 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3044 if the connection is using TLS or if the remote IP is the same as the
3045 local IP, and the "valid-client-cert option" if a client certificate has
3048 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3049 authenticators. This can be used for authorization after authentication
3050 succeeds. (In the case of plaintext, it servers for both authentication
3053 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3054 if any retry times were supplied.
3056 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3057 connection, where there is no client IP address to check. In this
3058 situation, the verify now always succeeds.
3060 PH/25 Applied John Jetmore's -Mset patch.
3062 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3064 PH/27 In a string expansion for a processed (not raw) header when multiple
3065 headers of the same name were present, leading whitespace was being
3066 removed from all of them, but trailing whitespace was being removed only
3067 from the last one. Now trailing whitespace is removed from each header
3068 before concatenation. Completely empty headers in a concatenation (as
3069 before) are ignored.
3071 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3072 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3074 PH/29 [Removed. This was a change that I later backed out, and forgot to
3075 correct the ChangeLog entry (that I had efficiently created) before
3076 committing the later change.]
3078 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3079 address errors (4xx response to RCPT) over the same connection as other
3080 messages routed to the same hosts. Such deliveries are always "forced",
3081 so retry times are not inspected. This resulted in far too many retries
3082 for the affected addresses. The effect occurred only when there were more
3083 hosts than the hosts_max_try setting in the smtp transport when it had
3084 the 4xx errors. Those hosts that it had tried were not added to the list
3085 of hosts for which the message was waiting, so if all were tried, there
3086 was no problem. Two fixes have been applied:
3088 (i) If there are any address or message errors in an SMTP delivery, none
3089 of the hosts (tried or untried) are now added to the list of hosts
3090 for which the message is waiting, so the message should not be a
3091 candidate for sending over the same connection that was used for a
3092 successful delivery of some other message. This seems entirely
3093 reasonable: after all the message is NOT "waiting for some host".
3094 This is so "obvious" that I'm not sure why it wasn't done
3095 previously. Hope I haven't missed anything, but it can't do any
3096 harm, as the worst effect is to miss an optimization.
3098 (ii) If, despite (i), such a delivery is accidentally attempted, the
3099 routing retry time is respected, so at least it doesn't keep
3100 hammering the server.
3102 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3103 in ${readsocket because some servers need this prod.
3105 PH/32 Added some extra debug output when updating a wait-xxx database.
3107 PH/33 The hint "could be header name not terminated by colon", which has been
3108 given for certain expansion errors for a long time, was not being given
3109 for the ${if def:h_colon_omitted{... case.
3111 PH/34 The spec says: "With one important exception, whenever a domain list is
3112 being scanned, $domain contains the subject domain." There was at least
3113 one case where this was not true.
3115 PH/35 The error "getsockname() failed: connection reset by peer" was being
3116 written to the panic log as well as the main log, but it isn't really
3117 panic-worthy as it just means the connection died rather early on. I have
3118 removed the panic log writing for the ECONNRESET error when getsockname()
3121 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3122 runs only) independently of the message's sender address. This meant
3123 that, if the 4xx error was in fact related to the sender, a different
3124 message to the same recipient with a different sender could confuse
3125 things. In particular, this can happen when sending to a greylisting
3126 server, but other circumstances could also provoke similar problems.
3127 I have changed the default so that the retry time for these errors is now
3128 based a combination of the sender and recipient addresses. This change
3129 can be overridden by setting address_retry_include_sender=false in the
3132 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3133 remote server are returned as part of bounce messages. This was not
3134 happening for LMTP over a pipe (the lmtp transport), but now it is the
3135 same for both kinds of LMTP.
3137 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3138 in header lines that were in fact CNAMEs. This is no longer the case.
3140 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3141 and queue runs started by the daemon processed all messages. This has
3142 been fixed so that -R and -S can now usefully be given with -q<time>.
3144 PH/40 Import PCRE release 6.7 (fixes some bugs).
3146 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3148 PH/42 Give an error if -q is specified more than once.
3150 PH/43 Renamed the variables $interface_address and $interface_port as
3151 $received_ip_address and $received_port, to make it clear that these
3152 values apply to message reception, and not to the outgoing interface when
3153 a message is delivered. (The old names remain recognized, of course.)
3155 PH/44 There was no timeout on the connect() call when using a Unix domain
3156 socket in the ${readsocket expansion. There now is.
3158 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3159 be meaningful with "accept".
3161 SC/01 Eximstats V1.43
3162 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3164 SC/02 Eximstats V1.44
3165 Use a glob alias rather than an array ref in the generated
3166 parser. This improves both readability and performance.
3168 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3169 Collect SpamAssassin and rejection statistics.
3170 Don't display local sender or destination tables unless
3171 there is data to show.
3172 Added average volumes into the top table text output.
3174 SC/04 Eximstats V1.46
3175 Collect data on the number of addresses (recipients)
3176 as well as the number of messages.
3178 SC/05 Eximstats V1.47
3179 Added 'Message too big' to the list of mail rejection
3180 reasons (thanks to Marco Gaiarin).
3182 SC/06 Eximstats V1.48
3183 Mainlog lines which have GMT offsets and are too short to
3184 have a flag are now skipped.
3186 SC/07 Eximstats V1.49 (Alain Williams)
3187 Added the -emptyok flag.
3189 SC/08 Eximstats V1.50
3190 Fixes for obtaining the IP address from reject messages.
3192 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3193 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3194 whitespace changes from 4.64-PH/27
3196 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3199 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3200 are found, allow negative numbers in numeric criteria)
3202 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3204 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3205 to match changes made in 4.64-PH/43
3207 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3209 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3210 "domain" is actually an IP address, and force a failure. This locks out
3211 those revolvers/nameservers that support "A-for-A" lookups, in
3212 contravention of the specifications.
3214 PH/48 When a host name was looked up from an IP address, and the subsequent
3215 forward lookup of the name timed out, the host name was left in
3216 $sender_host_name, contrary to the specification.
3218 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3219 restricted to single-key lookups, Exim was not diagnosing an error if
3220 * or *@ was used with a query-style lookup.
3222 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3224 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3225 long ago, but noone interested enough thought of it. Let's just say that
3226 the "1.1" means that there are some new functions that weren't there at
3227 some point in the past.
3229 PH/51 Error processing for expansion failure of helo_data from an smtp
3230 transport during callout processing was broken.
3232 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3233 tested/used via the -bh/-bhc/-bs options.
3235 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3236 bug, fixed in subsequent PCRE releases).
3238 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3239 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3241 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3247 SC/01 Use a glob alias rather than an array ref in eximstats generated
3248 parser. This improves both readability and performance.
3250 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3251 Don't display local sender or destination tables in eximstats unless
3252 there is data to show.
3253 Added average volumes into the eximstats top table text output.
3255 SC/03 Collect data on the number of addresses (recipients) as well
3256 as the number of messages in eximstats.
3258 TF/01 Correct an error in the documentation for the redirect router. Exim
3259 does (usually) call initgroups() when daemonizing.
3261 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3262 with consistent privilege compared to when running as a daemon.
3264 TF/03 Note in the spec that $authenticated_id is not set for local
3265 submissions from trusted users.
3267 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3268 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3270 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3271 by adding some example configuration directives to the default
3272 configuration file. A little bit of work is required to uncomment the
3273 directives and define how usernames and passwords are checked, but
3274 there is now a framework to start from.
3276 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3277 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3278 without this. I don't know how relevant this is to other LDAP libraries.
3280 PH/02 Add the verb name to the "unknown ACL verb" error.
3282 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3284 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3286 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3287 directory not expanded when it should be if an expanded home directory
3288 was set for the address (which is overridden by the transport).
3290 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3293 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3294 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3295 because it is too late at that time, and has no effect.
3297 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3298 security issue with \' (bugzilla #107). I could not use the
3299 PQescapeStringConn() function, because it needs a PGconn value as one of
3302 PH/08 When testing addresses using -bt, indicate those final addresses that
3303 are duplicates that would not cause an additional delivery. At least one
3304 person was confused, thinking that -bt output corresponded to deliveries.
3305 (Suppressing duplicates isn't a good idea as you lose the information
3306 about possibly different redirections that led to the duplicates.)
3308 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3309 systems where poll() doesn't work, in particular OS X.
3311 PH/10 Added more information to debugging output for retry time not reached.
3313 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3314 operations in malware.c.
3316 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3319 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3320 syslog_duplication set false, log lines that would normally be written
3321 both the the main log and to the reject log were not written to syslog at
3324 PH/14 In the default configuration, change the use of "message" in ACL warn
3325 statements to "add_header".
3327 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3328 not followed by a command (e.g. "seen endif").
3330 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3331 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3334 PH/17 Added extra conditions to the default value of delay_warning_condition
3338 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3339 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3340 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3343 The Auto-Submitted: and various List- headers are standardised, whereas I
3344 don't think Precedence: ever was.
3346 PH/18 Refactored debugging code in route_finduser() to show more information,
3347 in particular, the error code if getpwnam() issues one.
3349 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3350 This is apparently needed in addition to the PH/07 change above to avoid
3351 any possible encoding problems.
3353 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3354 but not after initializing Perl.
3356 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3357 output them only if debugging. By default they are written stderr,
3358 apparently, which is not desirable.
3360 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3363 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3366 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3368 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3369 authenticated or an ident call has been made. Suppress the default
3370 values for $authenticated_id and $authenticated_sender (but permit -oMai
3371 and -oMas) when testing with -bh.
3373 PH/24 Re-jigged the order of the tests in the default configuration so that the
3374 tests for valid domains and recipients precede the DNS black list and CSA
3375 tests, on the grounds that those ones are more expensive.
3377 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3378 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3379 This bug exists in every version of Exim that I still have, right back to
3382 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3383 However, an attempt to turn on case-sensitivity in a regex key by
3384 including (?-i) didn't work because the subject string was already
3385 lowercased, and the effects were non-intuitive. It turns out that a
3386 one-line patch can be used to allow (?-i) to work as expected.
3392 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3393 other effects) broke the use of negated acl sub-conditions.
3395 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3398 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3399 "Deny" causes Exim to reject the incoming connection with a 554 error.
3400 Unfortunately, if there is a major crisis, such as a disk failure,
3401 tcp-wrappers gives "deny", whereas what one would like would be some
3402 kind of temporary error. A kludge has been added to help with this.
3403 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3404 554 error is used if errno is still zero or contains ENOENT (which occurs
3405 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3408 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3410 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3411 errors. Otherwise a message that provokes a temporary error (when other
3412 messages do not) can cause a whole host to time out.
3414 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3415 addresses were routed directly to files or pipes from a redirect router.
3416 File deliveries just didn't batch; pipe deliveries might have suffered
3419 PH/06 A failure to get a lock for a hints database would erroneously always say
3420 "Failed to get write lock", even when it was really a read lock.
3422 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3423 of 0600. This has been changed to use the value of the lockfile_mode
3424 option (which defaults to 0600).
3426 PH/08 Applied small patch from the Sieve maintainer.
3428 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3429 folder from quota calculations, a direct delivery into this folder messed
3430 up the contents of the maildirsize file. This was because the regex was
3431 used only to exclude .Trash (or whatever) when the size of the mailbox
3432 was calculated. There was no check that a delivery was happening into an
3433 excluded directory. This bug has been fixed by ignoring all quota
3434 processing for deliveries into excluded directories.
3436 PH/10 Added the maildirfolder_create_regex option to appendfile.
3442 PH/01 The code for finding all the local interface addresses on a FreeBSD
3443 system running IPv6 was broken. This may well have applied to all BSD
3444 systems, as well as to others that have similar system calls. The broken
3445 code found IPv4 interfaces correctly, but gave incorrect values for the
3446 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3447 that it would not match correctly against @[] and not recognize the IPv6
3450 PH/02 The ipliteral router was not recognizing addresses of the form user@
3451 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3453 PH/03 Added disable_ipv6.
3455 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3456 decoded form, because it is most often used to construct To: headers
3457 lines in autoreplies, and the decoded form may well be syntactically
3458 invalid. However, $reply_address has leading white space removed, and all
3459 newlines turned into spaces so that the autoreply transport does not
3462 PH/05 If group was specified without a user on a router, and no group or user
3463 was specified on a transport, the group from the router was ignored.
3465 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3466 for visible compile-time settings that can be used to change these
3467 numbers, for those that want even more. Backwards compatibility with old
3468 spool files has been maintained. However, going back to a previous Exim
3469 release will lost any variables that are in spool files.
3471 PH/07 Two small changes when running in the test harness: increase delay when
3472 passing a TCP/IP connection to a new process, in case the original
3473 process has to generate a bounce, and remove special handling of
3474 127.0.0.2 (sic), which is no longer necessary.
3476 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3477 be the same on different OS.
3479 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3482 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3483 whether --show-vars was specified or not
3485 JJ/02 exipick: Added support for new ACL variable spool format introduced
3488 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3489 syntactically invalid From: or Reply-to: line, and a filter used this to
3490 generate an autoreply, and therefore failed to obtain an address for the
3491 autoreply, Exim could try to deliver to a non-existent relative file
3492 name, causing unrelated and misleading errors. What now happens is that
3493 it logs this as a hard delivery error, but does not attempt to create a
3496 PH/11 The exinext utility has a -C option for testing purposes, but although
3497 the given file was scanned by exinext itself; it wasn't being passed on
3498 when Exim was called.
3500 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3501 an end-of-file indication when reading a command response.
3503 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3504 compiled. In many other places in Exim, IPv6 addresses are always
3505 recognized, so I have changed this. It also means that IPv4 domain
3506 literals of the form [IPV4:n.n.n.n] are now always recognized.
3508 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3509 used if the router is not running as root, for example, when verifying at
3510 ACL time, or when using -bh. The debugging output from this situation was
3511 non-existent - all you got was a failure to exec. I have made two
3514 (a) Failures to set uid/gid, the current directory, or a process leader
3515 in a subprocess such as that created by queryprogram now generate
3516 suitable debugging output when -d is set.
3518 (b) The queryprogram router detects when it is not running as root,
3519 outputs suitable debugging information if -d is set, and then runs
3520 the subprocess without attempting to change uid/gid.
3522 PH/15 Minor change to Makefile for building test_host (undocumented testing
3525 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3526 additional section of a DNS packet that returns MX or SRV records.
3527 Instead, it always explicitly searches for A/AAAA records. This avoids
3528 major problems that occur when a DNS server includes only records of one
3529 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3530 fixed another bug: if SRV records were looked up and the corresponding
3531 address records were *not* found in the additional section, the port
3532 values from the SRV records were lost.
3534 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3535 using the correct key (the original address) when searching the retry
3536 rules in order to find which one to use for generating the retry hint.
3538 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3539 adding the default one. Similarly, if it contains a Reply-To: header, the
3540 errors_reply_to option, if set, is not used.
3542 PH/19 When calculating a retry time, Exim used to measure the "time since
3543 failure" by looking at the "first failed" field in the retry record. Now
3544 it does not use this if it is later than than the arrival time of the
3545 message. Instead it uses the arrival time. This makes for better
3546 behaviour in cases where some deliveries succeed, thus re-setting the
3547 "first failed" field. An example is a quota failure for a huge message
3548 when small messages continue to be delivered. Without this change, the
3549 "time since failure" will always be short, possible causing more frequent
3550 delivery attempts for the huge message than are intended.
3551 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3553 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3554 $1, $2, $3) because the numerical variables can be reset during some
3555 expansion items (e.g. "match"), thereby losing the authentication data.
3557 PH/21 Make -bV show the size of off_t variables so that the test suite can
3558 decide whether to run tests for quotas > 2G.
3560 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3561 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3562 filecount value is greater than 2G or if a quota value is greater than 2G
3563 on a system where the size of off_t is not greater than 4, a panic error
3566 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3567 never match. The debug and -bh output now contains an explicit error
3568 message indicating a malformed IPv4 address or mask.
3570 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3571 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3572 PH/23 above applies.
3574 PH/25 Do not write to syslog when running in the test harness. The only
3575 occasion when this arises is a failure to open the main or panic logs
3576 (for which there is an explicit test).
3578 PH/26 Added the /no_tell option to "control=freeze".
3580 PH/27 If a host name lookup failed very early in a connection, for example, if
3581 the IP address matched host_lookup and the reverse lookup yielded a name
3582 that did not have a forward lookup, an error message of the form "no IP
3583 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3584 could be logged. Now it outputs the IP address instead of "NULL".
3586 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3587 allows the sender and the authenticated sender to be set when
3588 submitting a message from within Exim. Since child_open_exim() is
3589 documented for local_scan(), the new function should be too.
3591 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3592 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3593 results in an empty string is now treated as unset.
3595 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3597 PH/31 Added sender_verify_fail logging option.
3599 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3600 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3601 tidied the source and removed it altogether.
3603 PH/33 When a queue run was abandoned because the load average was too high, a
3604 log line was always written; now it is written only if the queue_run log
3605 selector is set. In addition, the log line for abandonment now contains
3606 information about the queue run such as the pid. This is always present
3607 in "start" and "stop" lines but was omitted from the "abandon" line.
3609 PH/34 Omit spaces between a header name and the colon in the error message that
3610 is given when verify = headers_syntax fails (if there are lots of them,
3611 the message gets confusing).
3613 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3614 names, as there are now some PTR records that contain slashes. This check
3615 is only to protect against broken name servers that fall over on strange
3616 characters, so the fact that it applies to all lookups doesn't matter.
3618 PH/36 Now that the new test suite is complete, we can remove some of the
3619 special code in Exim that was needed for the old test suite. For example,
3620 sorting DNS records because real resolvers return them in an arbitrary
3621 order. The new test suite's fake resolver always returns records in the
3624 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3625 bounces) except when queue_only is set, to avoid logging races between
3626 the different processes.
3628 PH/38 Panic-die if .include specifies a non-absolute path.
3630 PH/39 A tweak to the "H" retry rule from its user.
3632 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3633 a label. They prevented compilation on older perls.
3635 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3636 a warning to be raised on newish perls.
3638 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3639 on queue. Changes to match documented behaviour of showing count of
3640 messages matching specified criteria.
3642 PH/40 Changed the default ident timeout from 30s to 5s.
3644 PH/41 Added support for the use of login_cap features, on those BSD systems
3645 that have them, for controlling the resources used by pipe deliveries.
3647 PH/42 The content-scanning code uses fopen() to create files in which to put
3648 message data. Previously it was not paying any attention to the mode of
3649 the files. Exim runs with umask(0) because the rest of the code creates
3650 files with open(), and sets the required mode explicitly. Thus, these
3651 files were ending up world-writeable. This was not a big issue, because,
3652 being within the spool directory, they were not world-accessible. I have
3653 created a function called modefopen, which takes an additional mode
3654 argument. It sets umask(777), creates the file, chmods it to the required
3655 mode, then resets the umask. All the relevant calls to fopen() in the
3656 content scanning code have been changed to use this function.
3658 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3659 to 24 hours. This avoids potential overflow problems when processing G
3660 and H retry rules. I suspect nobody ever tinkers with this value.
3662 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3664 PH/45 When the plaintext authenticator is running as a client, the server's
3665 challenges are checked to ensure they are valid base64 strings. By
3666 default, the authentication attempt is cancelled if an invalid string is
3667 received. Setting client_ignore_invalid_base64 true ignores these errors.
3668 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3669 they are received. Thus, the responses can be made to depend on the
3670 challenges. If an invalid string is ignored, an empty string is placed in
3673 PH/46 Messages that are created by the autoreply transport now contains a
3674 References: header, in accordance with RFCs 2822 and 3834.
3676 PH/47 Added authenticated_sender_force to the smtp transport.
3678 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3680 PH/49 Installed latest patch from the Sieve maintainer.
3682 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3683 was also set, the appendfile transport was unnecessarily scanning a
3684 directory of message files (e.g. for maildir delivery) to find the count
3685 of files (along with the size), even though it did not need this
3686 information. It now does the scan only if it needs to find either the
3687 size of the count of files.
3689 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3691 PH/52 Two bugs concerned with error handling when the smtp transport is
3694 (i) Exim was not creating retry information for temporary errors given
3695 for individual recipients after the DATA command when the smtp transport
3696 was used in LMTP mode. This meant that they could be retried too
3697 frequently, and not timed out correctly.
3699 (ii) Exim was setting the flag that allows error details to be returned
3700 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3701 recipients that were returned after the DATA command.
3703 PH/53 This is related to PH/52, but is more general: for any failing address,
3704 when detailed error information was permitted to be returned to the
3705 sender, but the error was temporary, then after the final timeout, only
3706 "retry timeout exceeded" was returned. Now it returns the full error as
3707 well as "retry timeout exceeded".
3709 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3710 do this, and (what is worse) MTAs that accept it.
3712 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3713 will now be deprecated.
3715 PH/56 New os.c-cygwin from the Cygwin maintainer.
3717 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3718 formats (previously only available in exim formats via -bpr, -bpru,
3719 and -bpra. Now also available in native and exiqgrep formats)
3721 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3722 with very large, slow to parse queues
3724 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3726 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3728 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3729 responses to authentication challenges, though it was showing the
3730 challenges; (ii) I've removed the CR characters from the debug output for
3733 PH/58 Allow for the insertion of a newline as well as a space when a string
3734 is turned into more than one encoded-word during RFC 2047 encoding. The
3735 Sieve code now uses this.
3737 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3738 data_4xx, lost_connection, tls_required.
3740 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3741 message was being sent as an SMTP response.
3743 PH/61 Add -l and -k options to exicyclog.
3745 PH/62 When verifying, if an address was redirected to one new address, so that
3746 verification continued, and the new address failed or deferred after
3747 having set something in $address_data, the value of $address_data was not
3748 passed back to the ACL. This was different to the case when no
3749 redirection occurred. The value is now passed back in both cases.
3751 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3752 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3753 use login_cap.h, so on its own it isn't the distinguishing feature. The
3754 new name refers directly to the setclassresources() function.
3756 PH/65 Added configuration files for NetBSD3.
3758 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3760 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3761 is preferred over IPv4.
3763 PH/68 The bounce_return_message and bounce_return_body options were not being
3764 honoured for bounces generated during the reception of non-SMTP messages.
3765 In particular, this applied to messages rejected by the ACL. This bug has
3766 been fixed. However, if bounce_return_message is true and bounce_return_
3767 body is false, the headers that are returned for a non-SMTP message
3768 include only those that have been read before the error was detected.
3769 (In the case of an ACL rejection, they have all been read.)
3771 PH/69 The HTML version of the specification is now built in a directory called
3772 spec_html instead of spec.html, because the latter looks like a path with
3773 a MIME-type, and this confuses some software.
3775 PH/70 Catch two compiler warnings in sieve.c.
3777 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3778 function verify_get_ident() calls ip_connect() to connect a socket, but
3779 if the "connect()" function timed out, ip_connect() used to close the
3780 socket. However, verify_get_ident() also closes the socket later, and in
3781 between Exim writes to the log, which may get opened at this point. When
3782 the socket was closed in ip_connect(), the log could get the same file
3783 descriptor number as the socket. This naturally causes chaos. The fix is
3784 not to close the socket in ip_connect(); the socket should be closed by
3785 the function that creates it. There was only one place in the code where
3786 this was missing, in the iplookup router, which I don't think anybody now
3787 uses, but I've fixed it anyway.
3789 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3790 well as to direct DNS lookups. Otherwise the handling of names in host
3791 lists is inconsistent and therefore confusing.
3797 PH/01 Two changes to the default runtime configuration:
3799 (1) Move the checks for relay_from_hosts and authenticated clients from
3800 after to before the (commented out) DNS black list checks.
3802 (2) Add control=submission to the relay_from_hosts and authenticated
3803 clients checks, on the grounds that messages accepted by these
3804 statements are most likely to be submissions.
3806 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3808 (1) Generate an error if the third argument for the ${prvs expansion is
3811 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3814 (3) Reset the variables that are obtained from the first argument of
3815 ${prvscheck and used in the second argument before leaving the code,
3816 because their memory is reclaimed, so using them afterwards may do
3819 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3820 one (it's much easier than Tom thought :-).
3822 (5) Because of (4), we can now allow for the use of $prvscheck_result
3823 inside the third argument.
3825 PH/03 For some reason, the default setting of PATH when running a command from
3826 a pipe transport was just "/usr/bin". I have changed it to
3829 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3830 anything to be listed in the output from -bV.
3832 PH/05 When a filter generated an autoreply, the entire To: header line was
3833 quoted in the delivery log line, like this:
3835 => >A.N.Other <ano@some.domain> <original@ddress> ...
3837 This has been changed so that it extracts the operative address. There
3838 may be more than one such address. If so, they are comma-separated, like
3841 => >ano@some.domain,ona@other.domain <original@ddress> ...
3843 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3844 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3845 not being looked up in the rDNS to get a host name, Exim was showing the
3846 IP address twice in Received: lines, even though the IP addresses were
3847 identical. For example:
3849 Received: from [1.2.3.4] (helo=[1.2.3.4])
3851 However, if the real host name was known, it was omitting the HELO data
3852 if it matched the actual IP address. This has been tidied up so that it
3853 doesn't show the same IP address twice.
3855 PH/07 When both +timestamp and +memory debugging was on, the value given by
3856 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3857 called by the debug printing, thereby overwriting the timestamp buffer.
3858 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3860 PH/08 When the original message was included in an autoreply transport, it
3861 always said "this is a copy of the message, including all the headers",
3862 even if body_only or headers_only was set. It now gives an appropriate
3865 PH/09 Applied a patch from the Sieve maintainer which:
3867 o fixes some comments
3868 o adds the (disabled) notify extension core
3869 o adds some debug output for the result of if/elsif tests
3870 o points to the current vacation draft in the documentation
3871 and documents the missing references header update
3875 o fixes a bug in processing the envelope test (when testing
3876 multiple envelope elements, the last element determined the
3879 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3880 Electronic Mail") by including:
3882 Auto-submitted: auto-generated
3884 in the messages that it generates (bounce messages and others, such as
3885 warnings). In the case of bounce messages for non-SMTP messages, there was
3886 also a typo: it was using "Auto_submitted" (underscore instead of
3887 hyphen). Since every message generated by Exim is necessarily in response
3888 to another message, thes have all been changed to:
3890 Auto-Submitted: auto-replied
3892 in accordance with these statements in the RFC:
3894 The auto-replied keyword:
3896 - SHOULD be used on messages sent in direct response to another
3897 message by an automatic process,
3899 - MUST NOT be used on manually-generated messages,
3901 - MAY be used on Delivery Status Notifications (DSNs) and Message
3902 Disposition Notifications (MDNs),
3904 - MUST NOT be used on messages generated by automatic or periodic
3905 processes, except for messages which are automatic responses to
3908 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3909 to the default Received: header definition.
3911 PH/12 Added log selector acl_warn_skipped (default on).
3913 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3914 variables because (a) they are in the wrong storage pool and (b) even if
3915 they were copied, it wouldn't work properly because of the caching.
3917 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3918 checking when decoding. Apparently there are clients that generate
3919 overlong encoded strings. Why am I not surprised?
3921 PH/15 If the first argument of "${if match_address" was not empty, but did not
3922 contain an "@" character, Exim crashed. Now it writes a panic log message
3923 and treats the condition as false.
3925 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3927 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3928 extension "envelope-auth". The code is finished and in agreement with
3929 other implementations, but there is no documentation so far and in fact,
3930 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3931 not changing the active code.
3933 Print executed "if" and "elsif" statements when debugging is used. This
3934 helps a great deal to understand what a filter does.
3936 Document more things not specified clearly in RFC3028. I had all this
3937 sorted out, when out of a sudden new issues came to my mind. Oops."
3939 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3942 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3943 canonical form (as documented). However, after a host name lookup from
3944 the IP address, check_host() was doing a simple string comparison with
3945 addresses acquired from the DNS when checking that the found name did
3946 have the original IP as one of its addresses. Since any found IPv6
3947 addresses are likely to be in abbreviated form, the comparison could
3948 fail. Luckily, there already exists a function for doing the comparison
3949 by converting both addresses to binary, so now that is used instead of
3950 the text comparison.
3952 PH/20 There was another similar case to PH/19, when a complete host name was
3953 given in a host list; looking up its IP address could give an abbreviated
3954 form, whereas the current host's name might or might not be abbreviated.
3955 The same fix has been applied.
3961 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3962 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3965 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3966 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3968 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3970 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3971 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3972 Also, for "Auto-Submitted", treat anything other than "no" as
3973 non-personal, in accordance with RFC 3834. (Previously it treated
3974 anything starting "auto-" as non-personal.)
3976 TF/01 The control=submission/name=... option had a problem with syntax
3977 errors if the name included a slash character. The /name= option
3978 now slurps the rest of the string, so it can include any characters
3979 but it must come last in the list of options (after /sender_retain
3982 PH/05 Some modifications to the interface to the fake nameserver for the new
3990 TK/01 Added the "success_on_redirect" address verification option. See
3991 NewStuff for rationale and an example.
3993 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3995 PH/02 Patch to exigrep to allow it to work on syslog lines.
3997 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3998 fread() to skip over the body file's header line, because in Cygwin the
3999 header line is locked and is inaccessible.
4001 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4002 co-exist for some time) to make it clear that it is the Exim ID that is
4003 referenced, not the Message-ID: header line.
4005 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4006 string_format() function, because snprintf() does not exist on all
4009 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4010 ${stat: expansion item.
4012 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4013 protocol synchronization error", to keep the pedants happy.
4015 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4016 well as for IRIX systems, when gcc is being used. See the host.c source
4019 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4021 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4024 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4025 transport and to the smtp transport in LMTP mode.
4027 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4029 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4030 run a filter in a subprocess. This could lead to confusion in subsequent
4031 lookups in the parent process. There should also be a search_tidyup() at
4032 the end of the subprocess.
4034 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4035 only if the host matched helo_try_verify_hosts, which caused the
4036 verification to occur when the EHLO/HELO command was issued. The ACL just
4037 tested the remembered result. Now, if a previous verification attempt has
4038 not happened, "verify = helo" does it there and then.
4040 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4042 TK/03 Fix log output including CR from clamd.
4044 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4045 exist provoked a memory error which could cause a segfault.
4047 PH/15 Installed PCRE 6.2
4049 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4051 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4052 of the problem. Specifically, suggested +O2 rather than +O1 for the
4055 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4057 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4058 record, the port was not correctly logged when the outgoing_port log
4059 selector was set (it logged the transort's default port).
4061 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4062 fallback_hosts, and "hosts" in the smtp transport.
4064 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4065 host errors such as "Connection refused".
4067 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4068 authentication with radiusclient 0.4.9:
4070 - Error returned from rc_read_config was caught wrongly
4071 - Username/password not passed on to radius server due to wrong length.
4073 The presumption is that some radiusclient API changes for 4.51/PH/17
4074 were not taken care of correctly. The code is still untested by me (my
4075 Linux distribution still has 0.3.2 of radiusclient), but it was
4076 contributed by a Radius user.
4078 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4079 expanding the "port" option of the smtp transport.
4081 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4082 while reading a MIME header. Thanks to Tom Hughes for a patch.
4084 PH/24 Include config.h inside local_scan.h so that configuration settings are
4087 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4088 that in an ACL for RCPT (for example), you can examine exactly what was
4091 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4092 commands, but it was not correctly comparing the address with the actual
4093 client host address. Thus, it would show the EHLO address in Received:
4094 header lines when this was not necessary.
4096 PH/27 Added the % operator to ${eval:}.
4098 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4099 it should be ignoring failures (because with -C, for example, it has lost
4100 privilege). It wasn't ignoring creation failures other than "already
4103 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4106 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4107 because an input line was too long, either on its own, or by virtue of
4108 too many continuations, the temporary file was not being removed, and the
4109 return code was incorrect.
4111 PH/31 Missing "BOOL" in function definition in filtertest.c.
4113 PH/32 Applied Sieve patches from the maintainer.
4115 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4117 PH/33 Added "verify = not_blind".
4119 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4120 Local/Makefile (with some defaults set). These are used in built scripts
4121 such as exicyclog, but they have never been used in the exim_install
4122 script (though there are many overriding facilities there). I have
4123 arranged that the exim_install script now takes note of these two
4126 PH/35 Installed configuration files for Dragonfly.
4128 PH/36 When a locally submitted message by a trusted user did not contain a
4129 From: header, and the sender address was obtained from -f or from an SMTP
4130 MAIL command, and the trusted user did not use -F to supply a sender
4131 name, $originator_name was incorrectly used when constructing a From:
4132 header. Furthermore, $originator_name was used for submission mode
4133 messages from external hosts without From: headers in a similar way,
4134 which is clearly wrong.
4136 PH/37 Added control=suppress_local_fixups.
4138 PH/38 When log_selector = +received_sender was set, and the addition of the
4139 sender made the log line's construction buffer exactly full, or one byte
4140 less than full, an overflow happened when the terminating "\n" was
4143 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4144 when the result of a list match is failure because a DNS lookup failed.
4146 PH/40 RM_COMMAND is now used in the building process.
4148 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4149 the "build-* directories that it finds.
4151 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4152 address in a domain literal was a prefix of an interface address.
4154 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4155 when verifying a sender address, unless rewrite_headers is false.
4157 PH/44 Wrote a long comment about why errors_to addresses are verified as
4158 recipients, not senders.
4160 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4161 the ratelimit ACL was added.
4163 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4165 PH/46 Added extra information about PostgreSQL errors to the error string.
4167 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4168 avoiding the need to install special zones in a real server. This is
4169 backwards compatible; if it can't find the fake resolver, it drops back.
4170 Thus, both old and new test suites can be run.
4172 TF/02 Added util/ratelimit.pl
4174 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4175 clock is set back in time.
4177 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4178 Candler <B.Candler@pobox.com>.
4180 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4181 OK for addresses that are the result of redirections.
4183 PH/48 A number of further additions for the benefit of the new test suite,
4184 including a fake gethostbyname() that interfaces to the fake DNS resolver
4187 TF/06 The fix for widen_domains has also been applied to qualify_single and
4188 search_parents which are the other dnslookup options that can cause
4191 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4194 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4196 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4197 service" if it encountered an MX record with an empty target hostname.
4198 The message is now "an MX or SRV record indicated no SMTP service".
4200 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4201 if the DNS of the sending site is misconfigured. This is quite a
4202 common situation. This change restores the behaviour of treating a
4203 helo verification defer as a failure.
4205 PH/51 If self=fail was set on a router, the bounce message did not include the
4206 actual error message.
4212 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4214 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4215 command itself ended in error, the underlying message about the transport
4216 filter timeout was being overwritten with the pipe command error. Now the
4217 underlying error message should be appended to the second error message.
4219 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4221 PH/02 Reduce the amount of output that "make" produces by default. Full output
4222 can still be requested.
4224 PH/03 The warning log line about a condition test deferring for a "warn" verb
4225 was being output only once per connection, rather than after each
4226 occurrence (because it was using the same function as for successful
4227 "warn" verbs). This seems wrong, so I have changed it.
4229 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4230 it should not have, which might have caused a crash in the right
4231 circumstances, but probably never did.
4233 PH/04 Installed a modified version of Tony Finch's patch to make submission
4234 mode fix the return path as well as the Sender: header line, and to
4235 add a /name= option so that you can make the user's friendly name appear
4238 TF/03 Added the control = fakedefer ACL modifier.
4240 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4241 Mark Lowes for thorough testing.
4243 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4245 TK/03 Merged latest SRS patch from Miles Wilton.
4247 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4248 with the definition in sysexits.h (which is #included earlier).
4249 Fortunately, Exim does not actually use EX_OK. The code used to try to
4250 preserve the sysexits.h value, by assuming that macro definitions were
4251 scanned for macro replacements. I have been disabused of this notion,
4252 so now the code just undefines EX_OK before #including unistd.h.
4254 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4255 in the smtp transport. When a block could not be written in a single
4256 write() function, the timeout was being re-applied to each part-write.
4257 This seems wrong - if the receiver was accepting one byte at a time it
4258 would take for ever. The timeout is now adjusted when this happens. It
4259 doesn't have to be particularly precise.
4261 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4262 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4264 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4265 without a domain if the check to <postmaster@domain> fails.
4267 SC/01 Eximstats: added -xls and the ability to specify output files
4268 (patch written by Frank Heydlauf).
4270 SC/02 Eximstats: use FileHandles for outputting results.
4272 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4274 SC/04 Eximstats: fixed display of large numbers with -nvr option
4276 SC/05 Eximstats: fixed merging of reports with empty tables.
4278 SC/06 Eximstats: added the -include_original_destination flag
4280 SC/07 Eximstats: removed tabs and trailing whitespace.
4282 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4284 TK/06 MBOX spool code: Add real "From " MBOX separator line
4285 so the .eml file is really in mbox format (even though
4286 most programs do not really care). Patch from Alex Miller.
4288 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4289 The latter is generated from $received_to and is only set if the
4290 message has one envelope recipient. SA can use these headers,
4291 obviously out-of-the-box. Patch from Alex Miller.
4293 PH/08 The ${def test on a variable was returning false if the variable's
4294 value was "0", contrary to what the specification has always said!
4295 The result should be true unless the variable is empty.
4297 PH/09 The syntax error of a character other than { following "${if
4298 def:variable_name" (after optional whitespace) was not being diagnosed.
4299 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4300 accidental colon was present, for example, could give incorrect results.
4302 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4303 result is used (not including appendfile, where other changes are about
4306 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4307 This involved changing a lot of size variables from int to off_t. It
4308 should work with maildirs and everything.
4310 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4311 spamd dying while we are connected to it.
4313 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4316 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4317 function for BDB 4.3.
4319 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4321 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4322 the sources for PCRE have been re-arranged and more files are now
4325 PH/15 The code I had for printing potentially long long variables in PH/11
4326 above was not the best (it lost precision). The length of off_t variables
4327 is now inspected at build time, and an appropriate printing format (%ld
4328 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4329 to be "long long int" or "long int". This is needed for the internal
4330 formatting function string_vformat().
4332 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4333 the configuration file to be ":syslog", then the script "guesses" where
4334 the logs files are, rather than using the compiled in default. In our
4335 case the guess is not the same as the compiled default, so the script
4336 suddenly stopped working when I started to use syslog. The patch checks
4337 to see if log_file_path is "". If so, it attempts to read it from exim
4338 with no configuration file to get the compiled in version, before it
4339 falls back to the previous guessing code."
4341 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4342 implementing BATV in an Exim configuration. See NewStuff for the gory
4345 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4346 Makefile that are specific to HP-UX.
4348 PH/18 If the "use_postmaster" option was set for a recipient callout together
4349 with the "random" option, the postmaster address was used as the MAIL
4350 FROM address for the random test, but not for the subsequent recipient
4351 test. It is now used for both.
4353 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4354 patch removes a few documentation additions to RFC 3028, because the
4355 latest draft now contains them. It adds the new en;ascii-case comparator
4356 and a new error check for 8bit text in MIME parts. Comparator and
4357 require names are now matched exactly. I enabled the subaddress
4358 extension, but it is not well tested yet (read: it works for me)."
4360 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4361 rework some of the code of TK/09 above to avoid the hardwired use of
4362 "%lld" and "long long". Replaced the call to snprintf() with a call to
4365 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4366 records point to non-existent hosts", "retry timeout exceeded", and
4367 "retry time not reached for any host after a long failure period".
4369 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4370 experimental DomainKeys support:
4372 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4373 (2) On an error such as an illegally used "control", the wrong name for
4374 the control was given.
4376 These problems did NOT occur unless DomainKeys support was compiled.
4378 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4380 PH/24 Added ${if match_ip condition.
4382 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4383 databases so that it will be absolutely obvious if a crash occurs in the
4384 DB library. This is a regular occurrence (often caused by mis-matched
4387 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4388 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4389 picked up on a user's system that detects such things. There doesn't seem
4390 to be a gcc warning option for this - only an attribute that has to be
4391 put on the function's prototype. It seems that in Fedora Core 4 they have
4392 set this on a number of new functions. No doubt there will be more in due
4395 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4396 specify a transport. However, if an address that was verified by such a
4397 router was the subject of a callout, Exim crashed because it tried to
4398 read the rcpt_include_affixes from the non-existent transport. Now it
4399 just assumes that the setting of that option is false. This bug was
4400 introduced by 4.51/PH/31.
4402 PH/28 Changed -d+all to exclude +memory, because that information is very
4403 rarely of interest, but it makes the output a lot bigger. People tend to
4404 do -d+all out of habit.
4406 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4407 code in os-type was giving problems when libc.so lives in lib64, like on
4410 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4411 aren't the modern standard, and it seems that some systems' include files
4412 don't always have them. Exim was already checking for some of the newer
4413 ones like T_AAAA, and defining it itself. I've added checks for all the
4414 record types that Exim uses.
4416 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4417 not automatically generating a new one, as it is supposed to. This
4418 prevented TLS from working. If the file did exist, but contained invalid
4419 data, a new version was generated, as expected. It was only the case of a
4420 non-existent file that was broken.
4422 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4423 with a change in libdomainkeys > 0.64.
4425 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4426 from DNS. If the selector record carries the flag, it now has
4427 precedence over the domain-wide flag.
4429 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4431 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4432 the use of a port name that isn't defined in /etc/services) occurred, the
4433 message was deferred as in a normal delivery, and thus remained on the
4434 spool, instead of being failed because of the mua_wrapper setting. This
4435 is now fixed, and I tidied up some of the mua_wrapper messages at the
4438 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4439 the messages in a hash of arrays rather than using individual hashes.
4440 This is a bit cleaner and results in dramatic memory savings, albeit
4441 at a slight CPU cost.
4443 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4444 as requested by Marc Sherman.
4446 SC/10 Eximstats: added histograms for user specified patterns as requested
4449 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4451 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4452 fopen() in the content-scanning modules that did not already have it.
4458 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4459 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4461 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4463 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4465 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4466 to test Sieve filters that use "vacation".
4468 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4469 that changes the way the GnuTLS parameters are stored in the cache file.
4470 The new format can be generated externally. For backward compatibility,
4471 if the data in the cache doesn't make sense, Exim assumes it has read an
4472 old-format file, and it generates new data and writes a new file. This
4473 means that you can't go back to an older release without removing the
4476 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4477 work if there are any delivery delays because "one_time" forces the
4478 parent to be marked "delivered", so its unseen clone is never tried
4479 again. For this reason, Exim now forbids the simultaneous setting of
4482 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4483 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4484 entry if you want to know the details. The fix, however, neglected to
4485 consider the case when local delivery batching is involved. The test for
4486 "previously delivered" was not happening when checking to see if an
4487 address could be batched with a previous (undelivered) one; under
4488 certain circumstances this could lead to multiple deliveries to the same
4491 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4492 in its include files, and this causes problems building Exim.
4494 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4495 header_syntax) but Exim was just ignoring anything given after a slash.
4496 In particular, this caused confusion with an attempt to use "verify =
4497 reverse_host_lookup/defer_ok". An error is now given when options are
4498 supplied for verify items that do not have them. (Maybe reverse_host_
4499 lookup should have a defer_ok option, but that's a different point.)
4501 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4502 defined by RFC 821) to 2048, because there were problems with some AUTH
4503 commands, and RFC 1869 says the size should be increased for extended
4504 SMTP commands that take arguments.
4506 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4509 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4510 "unknown" error; now it says that the functionality isn't in the binary.
4512 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4513 an address' error message when a string expansion fails (syntax or
4514 whatever). Otherwise the password may appear in the log. Following change
4515 PH/42 below, there is no longer a chance of it appearing in a bounce
4518 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4520 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4521 of its addresses was ever tried. (Bugzilla bug #2.)
4523 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4524 the result incorrectly in the debug output. (It correctly added a newline
4525 to what was transported.)
4527 TF/01 Added $received_time.
4529 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4530 commented out examples of how to interface to a virus scanner and to
4531 SpamAssassin. Also added commented examples of av_scanner and
4532 spamd_address settings.
4534 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4535 and controls are allowed in which ACLs. There were a couple of minor
4536 errors. Some of the entries in the conditions table (which is a table of
4537 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4538 negation of where the condition IS allowed.
4540 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4542 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4543 header file does not have a version number, so I've had to invent a new
4544 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4545 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4546 radiusclient), but it was contributed by a Radius user.
4548 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4549 files or directories, for OpenSSL.
4551 PH/19 When an Exim process that is running as root has to create an Exim log
4552 file, it does so in a subprocess that runs as exim:exim so as to get the
4553 ownership right at creation (otherwise, other Exim processes might see
4554 the file with the wrong ownership). There was no test for failure of this
4555 fork() call, which would lead to the process getting stuck as it waited
4556 for a non-existent subprocess. Forks do occasionally fail when resources
4557 run out. I reviewed all the other calls to fork(); they all seem to check
4560 PH/20 When checking for unexpected SMTP input at connect time (before writing
4561 the banner), Exim was not dealing correctly with a non-positive return
4562 from the read() function. If the client had disconnected by this time,
4563 the result was a log entry for a synchronization error with an empty
4564 string after "input=" when read() returned zero. If read() returned -1
4565 (an event I could not check), uninitialized data bytes were printed.
4566 There were reports of junk text (parts of files, etc) appearing after
4569 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4571 PH/22 Added support for macro redefinition, and (re)definition in between
4572 driver and ACL definitions.
4574 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4575 forgetting to use the resulting value; it was using the unexpanded value.
4577 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4578 hadn't been configured. The fix is from Juergen Kreileder, who
4579 understands it better than I do:
4581 "Here's what I see happening with three configured cyrus_sasl
4582 authenticators configured (plain, login, cram-md5):
4584 On startup auth_cyrus_sasl_init() gets called for each of these.
4585 This means three calls to sasl_listmech() without a specified mech_list.
4586 => SASL tests which mechs of all available mechs actually work
4587 => three warnings about OTP not working
4588 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4590 With the patch, sasl_listmech() also gets called three times. But now
4591 SASL's mech_list option is set to the server_mech specified in the the
4592 authenticator. Or in other words, the answer from sasl_listmech()
4593 gets limited to just the mech you're testing for (which is different
4595 => the return list contains just 'plain' or 'login', 'cram-md5' or
4596 nothing depending on the value of ob->server_mech.
4598 I've just tested the patch: Authentication still works fine,
4599 unavailable mechs specified in the exim configuration are still
4600 caught, and the auth.log warnings about OTP are gone."
4602 PH/25 When debugging is enabled, the contents of the command line are added
4603 to the debugging output, even when log_selector=+arguments is not
4606 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4607 answer is "GNU", and only if the return is "GNU/something" is the answer
4610 PH/27 $acl_verify_message is now set immediately after the failure of a
4611 verification in an ACL, and so is available in subsequent modifiers. In
4612 particular, the message can be preserved by coding like this:
4614 warn !verify = sender
4615 set acl_m0 = $acl_verify_message
4617 Previously, $acl_verify_message was set only while expanding "message"
4618 and "log_message" when a very denied access.
4620 PH/28 Modified OS/os.c-Linux with
4622 -#ifndef OS_LOAD_AVERAGE
4623 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4625 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4626 nomenclature these days.)
4628 PH/29 Installed patch from the Sieve maintainer that adds the options
4629 sieve_useraddress and sieve_subaddress to the redirect router.
4631 PH/30 In these circumstances:
4632 . Two addresses routed to the same list of hosts;
4633 . First host does not offer TLS;
4634 . First host accepts first address;
4635 . First host gives temporary error to second address;
4636 . Second host offers TLS and a TLS session is established;
4637 . Second host accepts second address.
4638 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4639 and peerdn, if requested) that were in fact used only for the second
4642 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4643 attention to any local part prefix or suffix that was matched by the
4644 router that accepted the address. It now behaves in the same way as it
4645 does for delivery: the affixes are removed from the local part unless
4646 rcpt_include_affixes is set on the transport.
4648 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4649 timeout during the DATA phase of an incoming message.
4651 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4652 applied a patch sanctioned by the Sieve maintainer.
4654 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4655 the uid or gid is negative. A case of a negative gid caused this to be
4656 noticed. The fix allows for either to be negative.
4658 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4659 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4662 PH/36 The change PH/12 above was broken. Fixed it.
4664 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4665 the grounds that routing the same address twice would always produce the
4666 same answer. This might have been true once, but it is certainly no
4667 longer true now. Routing a child address may depend on the previous
4668 routing that produced that child. Some complicated redirection strategies
4669 went wrong when messages had multiple recipients, and made Exim's
4670 behaviour dependent on the order in which the addresses were given.
4672 I have moved the duplicate checking until after the routing is complete.
4673 Exim scans the addresses that are assigned to local and remote
4674 transports, and removes any duplicates. This means that more work will be
4675 done, as duplicates will always all be routed, but duplicates are
4676 presumably rare, so I don't expect this is of any significance.
4678 For deliveries to pipes, files, and autoreplies, the duplicate checking
4679 still happens during the routing process, since they are not going to be
4682 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4683 It corrects a timeout issue with spamd. This is Ian's comment: "The
4684 background is that sometimes spamd either never reads data from a
4685 connection it has accepted, or it never writes response data. The exiscan
4686 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4687 blindly assumes that writes won't block so it may never time out."
4689 PH/39 Allow G after quota size as well as K and M.
4691 PH/40 The value set for $authenticated_id in an authenticator may not contain
4692 binary zeroes or newlines because the value is written to log lines and
4693 to spool files. There was no check on this. Now the value is run through
4694 the string_printing() function so that such characters are converted to
4695 printable escape sequences.
4697 PH/41 $message_linecount is a new variable that contains the total number of
4698 lines in the message. Compare $body_linecount, which is the count for the
4701 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4702 bounce and delay warning messages, except in certain special cases, which
4705 (a) An SMTP error message from a remote host;
4706 (b) A message specified in a :fail: redirection;
4707 (c) A message specified in a "fail" command in a system filter;
4708 (d) A message specified in a FAIL return from the queryprogram router;
4709 (e) A message specified by the cannot_route_message router option.
4711 In these cases only, Exim does include the error details in bounce and
4712 warning messages. There are also a few cases where bland messages such
4713 as "unrouteable address" or "local delivery error" are given.
4715 PH/43 $value is now also set for the "else" part of a ${run expansion.
4717 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4718 being worked on, but at least Exim now implements the latest version to
4721 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4722 process to complete was treated as a delivery failure, a timeout while
4723 writing the message to the pipe was logged, but erroneously treated as a
4724 successful delivery. Such timeouts include transport filter timeouts. For
4725 consistency with the overall process timeout, these timeouts are now
4726 treated as errors, giving rise to delivery failures by default. However,
4727 there is now a new Boolean option for the pipe transport called
4728 timeout_defer, which, if set TRUE, converts the failures into defers for
4729 both kinds of timeout. A transport filter timeout is now identified in
4732 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4733 systems where "make" and "gmake" are different, calling "gmake" at top
4734 level broke things. I've arranged for the value of $(MAKE) to be passed
4735 from the Makefile to this script so that it can call the same version of
4739 A note about Exim versions 4.44 and 4.50
4740 ----------------------------------------
4742 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4743 changes of various kinds. As a consequence, a big documentation update was
4744 needed. This delayed the release for rather longer than seemed good, especially
4745 in the light of a couple of (minor) security issues. Therefore, the changes
4746 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4747 release. So 4.44 and 4.50 are in effect two different branches that both start
4750 I have left the 4.50 change log unchanged; it contains all the changes since
4751 4.43. The change log for 4.44 is below; many of its items are identical to
4752 those for 4.50. This seems to be the most sensible way to preserve the
4753 historical information.
4759 1. Minor wording change to the doc/README.SIEVE file.
4761 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4762 computation of the current number of files was incorrect.
4764 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4765 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4768 4. Give more explanation in the error message when the command for a transport
4769 filter fails to execute.
4771 5. There are several places where Exim runs a non-Exim command in a
4772 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4773 was being done only for the command run by the queryprogram router. It is
4774 now done for all such subprocesses. The other cases are: ${run, transport
4775 filters, and the commands run by the lmtp and pipe transports.
4777 6. Added CONFIGURE_GROUP build-time option.
4779 7. Some older OS have a limit of 256 on the maximum number of file
4780 descriptors. Exim was using setrlimit() to set 1000 as a large value
4781 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4782 systems. I've change it so that if it can't get 1000, it tries for 256.
4784 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4785 was an oversight, and furthermore, ever since the addition of extra
4786 controls (e.g. 4.43/32), the checks on when to allow different forms of
4787 "control" were broken. There should now be diagnostics for all cases when a
4788 control that does not make sense is encountered.
4790 9. Added the /retain_sender option to "control=submission".
4792 10. $recipients is now available in the predata ACL (oversight).
4794 11. Tidy the search cache before the fork to do a delivery from a message
4795 received from the command line. Otherwise the child will trigger a lookup
4796 failure and thereby defer the delivery if it tries to use (for example) a
4797 cached ldap connection that the parent has called unbind on.
4799 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4800 of $address_data from the recipient verification was clobbered by the
4801 sender verification.
4803 13. The value of address_data from a sender verification is now available in
4804 $sender_address_data in subsequent conditions in the ACL statement.
4806 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4808 15. Added a new option "connect=<time>" to callout options, to set a different
4811 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4812 was its contents. (It was OK if the option was not defined at all.)
4814 17. A "Completed" log line is now written for messages that are removed from
4815 the spool by the -Mrm option.
4817 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4818 information about exactly what failed.
4820 19. Added -dd to debug only the daemon process.
4822 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4823 handles timeouts, both on the server side and network timeouts. Renamed the
4824 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4826 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4827 It is now set to "smtps".
4829 22. $host_address is now set to the target address during the checking of
4830 ignore_target_hosts.
4832 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4833 being passed; this would have caused $sender_host_name to have been used if
4834 matching the list had actually called for a host name (not very likely,
4835 since this list is usually IP addresses). A host name is now passed as
4838 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4839 code that specifies a non-restarting handler (typically sigaction() in
4840 modern systems) in an attempt to fix a rare and obscure crash bug.
4842 25. Narrowed the window for a race in the daemon that could cause it to ignore
4843 SIGCHLD signals. This is not a major problem, because they are used only to
4844 wake it up if nothing else does.
4846 26. A malformed maildirsize file could cause Exim to calculate negative values
4847 for the mailbox size or file count. Odd effects could occur as a result.
4848 The maildirsize information is now recalculated if the size or filecount
4851 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4852 support for a long time. Removed HAVE_SYS_VFS_H.
4854 28. Installed the latest version of exipick from John Jetmore.
4856 29. In an address list, if the pattern was not a regular expression, an empty
4857 subject address (from a bounce message) matched only if the pattern was an
4858 empty string. Non-empty patterns were not even tested. This was the wrong
4859 because it is perfectly reasonable to use an empty address as part of a
4860 database query. An empty address is now tested by patterns that are
4861 lookups. However, all the other forms of pattern expect the subject to
4862 contain a local part and a domain, and therefore, for them, an empty
4863 address still always fails if the pattern is not itself empty.
4865 30. Exim went into a mad DNS loop when attempting to do a callout where the
4866 host was specified on an smtp transport, and looking it up yielded more
4867 than one IP address.
4869 31. Re-factored the code for checking spool and log partition space into a
4870 function that finds that data and another that does the check. The former
4871 is then used to implement four new variables: $spool_space, $log_space,
4872 $spool_inodes, and $log_inodes.
4874 32. The RFC2047 encoding function was originally intended for short strings
4875 such as real names; it was not keeping to the 75-character limit for
4876 encoded words that the RFC imposes. It now respects the limit, and
4877 generates multiple encoded words if necessary. To be on the safe side, I
4878 have increased the buffer size for the ${rfc2047: expansion operator from
4881 33. It is now permitted to omit both strings after an "if" condition; if the
4882 condition is true, the result is "true". As before, when the second string
4883 is omitted, a false condition yields an empty string. This makes it less
4884 cumbersome to write custom ACL and router conditions.
4886 34. Failure to deliver a bounce message always caused it to be frozen, even if
4887 there was an errors_to setting on the router. The errors_to setting is now
4890 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4891 canonical form (fully expanded) before being placed in
4892 $sender_host_address.
4894 36. The table in the code that translates DNS record types into text (T_A to
4895 "A" for instance) was missing entries for NS and CNAME. It is just possible
4896 that this could have caused confusion if both these types were looked up
4897 for the same domain, because the text type is used as part of Exim's
4898 per-process caching. But the chance of anyone hitting this buglet seems
4901 37. The dnsdb lookup has been extended in a number of ways.
4903 (1) There is a new type, "zns", which walks up the domain tree until it
4904 finds some nameserver records. It should be used with care.
4906 (2) There is a new type, "mxh", which is like "mx" except that it returns
4907 just the host names, not the priorities.
4909 (3) It is now possible to give a list of domains (or IP addresses) to be
4910 looked up. The behaviour when one of the lookups defers can be
4911 controlled by a keyword.
4913 (4) It is now possible to specify the separator character for use when
4914 multiple records are returned.
4916 38. The dnslists ACL condition has been extended: it is now possible to supply
4917 a list of IP addresses and/or domains to be looked up in a particular DNS
4920 39. Added log_selector=+queue_time_overall.
4922 40. When running the queue in the test harness, wait just a tad after forking a
4923 delivery process, to get repeatability of debugging output.
4925 41. Include certificate and key file names in error message when GnuTLS fails
4926 to set them up, because the GnuTLS error message doesn't include the name
4927 of the failing file when there is a problem reading it.
4929 42. Allow both -bf and -bF in the same test run.
4931 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4933 44. The "Exiscan patch" is now merged into the mainline Exim source.
4935 45. Sometimes the final signoff response after QUIT could fail to get
4936 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4937 < 0 before doing a fflush(). This bug looks as though it goes back to the
4938 introduction of TLS in release 3.20, but "sometimes" must have been rare
4939 because the tests only now provoked it.
4941 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4942 (this can affect the format of dates).
4944 47. exim_tidydb, when checking for the continued existence of a message for
4945 which it has found a message-specific retry record, was not finding
4946 messages that were in split spool directories. Consequently, it was
4947 deleting retry records that should have stayed in existence.
4949 48. Steve fixed some bugs in eximstats.
4951 49. The SPA authentication driver was not abandoning authentication and moving
4952 on to the next authenticator when an expansion was forced to fail,
4953 contradicting the general specification for all authenticators. Instead it
4954 was generating a temporary error. It now behaves as specified.
4956 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4957 (the order specifies the preference for clients). The order is now AES256,
4958 AES128, 3DES, ARCFOUR128.
4960 51. Small patch to Sieve code - explicitly set From: when generating an
4963 52. Exim crashed if a remote delivery caused a very long error message to be
4964 recorded - for instance if somebody sent an entire SpamAssassin report back
4965 as a large number of 550 error lines. This bug was coincidentally fixed by
4966 increasing the size of one of Exim's internal buffers (big_buffer) that
4967 happened as part of the Exiscan merge. However, to be on the safe side, I
4968 have made the code more robust (and fixed the comments that describe what
4971 53. Now that there can be additional text after "Completed" in log lines (if
4972 the queue_time_overall log selector is set), a one-byte patch to exigrep
4973 was needed to allow it to recognize "Completed" as not the last thing in
4976 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4977 patch that reportedly fixes this has been added. I am not expert enough to
4978 create a test for it. This is what the patch creator wrote:
4980 "I found a little strange behaviour of ldap code when working with
4981 Windows 2003 AD Domain, where users was placed in more than one
4982 Organization Units. When I tried to give exim partial DN, the exit code
4983 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4984 But simultaneously result of request was absolutely normal ldap result,
4985 so I produce this patch..."
4987 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4988 REFERENCE, so I have modified the code to exclude the patch when that macro
4991 55. Some experimental protocols are using DNS PTR records for new purposes. The
4992 keys for these records are domain names, not reversed IP addresses. The
4993 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4994 leaves it alone. Component reversal etc. now happens only for IP addresses.
4997 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4999 57. Double the size of the debug message buffer (to 2048) so that more of very
5000 long debug lines gets shown.
5002 58. The exicyclog utility now does better if the number of log files to keep
5003 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5005 59. Two changes related to the smtp_active_hostname option:
5007 (1) $smtp_active_hostname is now available as a variable.
5008 (2) The default for smtp_banner uses $smtp_active_hostname instead
5009 of $primary_hostname.
5011 60. The host_aton() function is supposed to be passed a string that is known
5012 to be a valid IP address. However, in the case of IPv6 addresses, it was
5013 not checking this. This is a hostage to fortune. Exim now panics and dies
5014 if the condition is not met. A case was found where this could be provoked
5015 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5016 components; fortuitously, this particular loophole had already been fixed
5017 by change 4.50/55 above.
5019 If there are any other similar loopholes, the new check in host_aton()
5020 itself should stop them being exploited. The report I received stated that
5021 data on the command line could provoke the exploit when Exim was running as
5022 exim, but did not say which command line option was involved. All I could
5023 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5024 running as the user.
5027 61. There was a buffer overflow vulnerability in the SPA authentication code
5028 (which came originally from the Samba project). I have added a test to the
5029 spa_base64_to_bits() function which I hope fixes it.
5032 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5033 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5035 63. The daemon start-up calls getloadavg() while still root for those OS that
5036 need the first call to be done as root, but it missed one case: when
5037 deliver_queue_load_max is set with deliver_drop_privilege. This is
5038 necessary for the benefit of the queue runner, because there is no re-exec
5039 when deliver_drop_privilege is set.
5041 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5042 This has been fixed.
5044 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5045 was in use, was not putting the data itself into the right store pool;
5046 consequently, it could be overwritten for a subsequent message in the same
5047 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5050 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5052 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5053 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5054 the function were treating the return as a boolean value, which happened to
5055 work because 0=false and not-0=true, but is not correct code.
5057 68. The host_aton() function was not handling scoped IPv6 addresses (those
5058 with, for example, "%eth0" on the end) correctly.
5060 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5061 negated items (that is, ~something) in unsigned ints. Some compilers
5062 apparently mutter when there is no cast.
5064 70. If an address verification called from an ACL failed, and did not produce a
5065 user-specific message (i.e. there was only a "system" message), nothing was
5066 put in $acl_verify_message. In this situation, it now puts the system
5069 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5070 session; change 4.31/43 added the unwanted input to the log line - except
5071 that it did not do this in the start of session case. It now does.
5073 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5074 This is wrong and can cause the other end to generate a synchronization
5075 error if it is another Exim or anything else that does the synchronization
5076 check. A QUIT command is no longer sent after a timeout.
5078 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5079 during host lookups.
5081 74. The defer_ok option of callout verification was not working if it was used
5082 when verifying addresses in header lines, that is, for this case:
5084 verify = header_sender/callout=defer_ok
5086 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5087 those file descriptors could be used for SMTP connections. If anything
5088 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5089 could be sent to the SMTP client, causing chaos. The daemon now opens
5090 stdin, stdout, and stderr to /dev/null when it puts itself into the
5093 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5094 log by default. The user can override this with suitable Perl magic.
5096 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5097 the log message when discard triggers, was not working for the DATA ACL or
5098 for the non-SMTP ACL.
5100 78. Error message wording change in sieve.c.
5102 79. If smtp_accept_max_per_host was set, the number of connections could be
5103 restricted to fewer than expected, because the daemon was trying to set up
5104 a new connection before checking whether the processes handling previous
5105 connections had finished. The check for completed processes is now done
5106 earlier. On busy systems, this bug wouldn't be noticed because something
5107 else would have woken the daemon, and it would have reaped the completed
5110 80. If a message was submitted locally by a user whose login name contained one
5111 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5112 It caused a spool format error. I have fixed the spool reading code. A
5113 related problem was that the "from" clause in the Received: line became
5114 illegal because of the space(s). It is now covered by ${quote_local_part.
5116 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5119 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5121 83. Patch from Sieve maintainer to fix unterminated string problem in
5122 "vacation" handling.
5124 84. Some minor changes to the Linux configuration files to help with other
5125 OS variants using glibc.
5127 85. One more patch for Sieve to update vacation handling to latest spec.
5130 ----------------------------------------------------
5131 See the note above about the 4.44 and 4.50 releases.
5132 ----------------------------------------------------
5138 1. Change 4.43/35 introduced a bug that caused file counts to be
5139 incorrectly computed when quota_filecount was set in an appendfile
5142 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5143 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5146 3. Give more explanation in the error message when the command for a transport
5147 filter fails to execute.
5149 4. There are several places where Exim runs a non-Exim command in a
5150 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5151 was being done only for the command run by the queryprogram router. It is
5152 now done for all such subprocesses. The other cases are: ${run, transport
5153 filters, and the commands run by the lmtp and pipe transports.
5155 5. Some older OS have a limit of 256 on the maximum number of file
5156 descriptors. Exim was using setrlimit() to set 1000 as a large value
5157 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5158 systems. I've change it so that if it can't get 1000, it tries for 256.
5160 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5161 was an oversight, and furthermore, ever since the addition of extra
5162 controls (e.g. 4.43/32), the checks on when to allow different forms of
5163 "control" were broken. There should now be diagnostics for all cases when a
5164 control that does not make sense is encountered.
5166 7. $recipients is now available in the predata ACL (oversight).
5168 8. Tidy the search cache before the fork to do a delivery from a message
5169 received from the command line. Otherwise the child will trigger a lookup
5170 failure and thereby defer the delivery if it tries to use (for example) a
5171 cached ldap connection that the parent has called unbind on.
5173 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5174 of $address_data from the recipient verification was clobbered by the
5175 sender verification.
5177 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5178 was its contents. (It was OK if the option was not defined at all.)
5180 11. A "Completed" log line is now written for messages that are removed from
5181 the spool by the -Mrm option.
5183 12. $host_address is now set to the target address during the checking of
5184 ignore_target_hosts.
5186 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5187 being passed; this would have caused $sender_host_name to have been used if
5188 matching the list had actually called for a host name (not very likely,
5189 since this list is usually IP addresses). A host name is now passed as
5192 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5193 code that specifies a non-restarting handler (typically sigaction() in
5194 modern systems) in an attempt to fix a rare and obscure crash bug.
5196 15. Narrowed the window for a race in the daemon that could cause it to ignore
5197 SIGCHLD signals. This is not a major problem, because they are used only to
5198 wake it up if nothing else does.
5200 16. A malformed maildirsize file could cause Exim to calculate negative values
5201 for the mailbox size or file count. Odd effects could occur as a result.
5202 The maildirsize information is now recalculated if the size or filecount
5205 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5206 support for a long time. Removed HAVE_SYS_VFS_H.
5208 18. Updated exipick to current release from John Jetmore.
5210 19. Allow an empty sender to be matched against a lookup in an address list.
5211 Previously the only cases considered were a regular expression, or an
5214 20. Exim went into a mad DNS lookup loop when doing a callout where the
5215 host was specified on the transport, if the DNS lookup yielded more than
5218 21. The RFC2047 encoding function was originally intended for short strings
5219 such as real names; it was not keeping to the 75-character limit for
5220 encoded words that the RFC imposes. It now respects the limit, and
5221 generates multiple encoded words if necessary. To be on the safe side, I
5222 have increased the buffer size for the ${rfc2047: expansion operator from
5225 22. Failure to deliver a bounce message always caused it to be frozen, even if
5226 there was an errors_to setting on the router. The errors_to setting is now
5229 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5230 canonical form (fully expanded) before being placed in
5231 $sender_host_address.
5233 24. Updated eximstats to version 1.33
5235 25. Include certificate and key file names in error message when GnuTLS fails
5236 to set them up, because the GnuTLS error message doesn't include the name
5237 of the failing file when there is a problem reading it.
5239 26. Expand error message when OpenSSL has problems setting up cert/key files.
5242 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5243 (this can affect the format of dates).
5245 28. exim_tidydb, when checking for the continued existence of a message for
5246 which it has found a message-specific retry record, was not finding
5247 messages that were in split spool directories. Consequently, it was
5248 deleting retry records that should have stayed in existence.
5250 29. eximstats updated to version 1.35
5251 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5252 1.35 - bugfix such that pie charts by volume are generated correctly
5254 30. The SPA authentication driver was not abandoning authentication and moving
5255 on to the next authenticator when an expansion was forced to fail,
5256 contradicting the general specification for all authenticators. Instead it
5257 was generating a temporary error. It now behaves as specified.
5259 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5260 (the order specifies the preference for clients). The order is now AES256,
5261 AES128, 3DES, ARCFOUR128.
5263 31. Small patch to Sieve code - explicitly set From: when generating an
5266 32. Exim crashed if a remote delivery caused a very long error message to be
5267 recorded - for instance if somebody sent an entire SpamAssassin report back
5268 as a large number of 550 error lines. This bug was coincidentally fixed by
5269 increasing the size of one of Exim's internal buffers (big_buffer) that
5270 happened as part of the Exiscan merge. However, to be on the safe side, I
5271 have made the code more robust (and fixed the comments that describe what
5274 33. Some experimental protocols are using DNS PTR records for new purposes. The
5275 keys for these records are domain names, not reversed IP addresses. The
5276 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5277 leaves it alone. Component reversal etc. now happens only for IP addresses.
5280 34. The host_aton() function is supposed to be passed a string that is known
5281 to be a valid IP address. However, in the case of IPv6 addresses, it was
5282 not checking this. This is a hostage to fortune. Exim now panics and dies
5283 if the condition is not met. A case was found where this could be provoked
5284 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5285 components; fortuitously, this particular loophole had already been fixed
5286 by change 4.50/55 or 4.44/33 above.
5288 If there are any other similar loopholes, the new check in host_aton()
5289 itself should stop them being exploited. The report I received stated that
5290 data on the command line could provoke the exploit when Exim was running as
5291 exim, but did not say which command line option was involved. All I could
5292 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5293 running as the user.
5296 35. There was a buffer overflow vulnerability in the SPA authentication code
5297 (which came originally from the Samba project). I have added a test to the
5298 spa_base64_to_bits() function which I hope fixes it.
5301 36. The daemon start-up calls getloadavg() while still root for those OS that
5302 need the first call to be done as root, but it missed one case: when
5303 deliver_queue_load_max is set with deliver_drop_privilege. This is
5304 necessary for the benefit of the queue runner, because there is no re-exec
5305 when deliver_drop_privilege is set.
5307 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5308 was in use, was not putting the data itself into the right store pool;
5309 consequently, it could be overwritten for a subsequent message in the same
5310 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5313 38. Sometimes the final signoff response after QUIT could fail to get
5314 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5315 < 0 before doing a fflush(). This bug looks as though it goes back to the
5316 introduction of TLS in release 3.20, but "sometimes" must have been rare
5317 because the tests only now provoked it.
5323 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5324 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5325 Now it returns an integer. A number of calls were still expecting a T/F
5326 return. Fortuitously, in all cases, the tests worked in OK situations,
5327 which is the norm. However, things would have gone wrong on any write
5328 failures on the smtp file descriptor. This function is used when sending
5329 messages over SMTP and also when doing verify callouts.
5331 2. When Exim is called to do synchronous delivery of a locally submitted
5332 message (the -odf or -odi options), it no longer closes stderr before doing
5335 3. Implemented the mua_wrapper option.
5337 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5339 5. Implemented the functions header_remove(), header_testname(),
5340 header_add_at_position(), and receive_remove_recipient(), and exported them
5343 6. If an ACL "warn" statement specified the addition of headers, Exim already
5344 inserted X-ACL-Warn: at the start if there was no header name. However, it
5345 was not making this test for the second and subsequent header lines if
5346 there were newlines in the string. This meant that an invalid header could
5347 be inserted if Exim was badly configured.
5349 7. Allow an ACL "warn" statement to add header lines at the start or after all
5350 the Received: headers, as well as at the end.
5352 8. Added the rcpt_4xx retry error code.
5354 9. Added postmaster_mailfrom=xxx to callout verification option.
5356 10. Added mailfrom=xxxx to the callout verification option, for verify=
5359 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5360 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5362 12. Inserted some casts to stop certain compilers warning when using pointer
5363 differences as field lengths or precisions in printf-type calls (mostly
5364 affecting debugging statements).
5366 13. Added optional readline() support for -be (dynamically loaded).
5368 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5369 same clock tick as a message's arrival, so that its received time was the
5370 same as the "first fail" time on the retry record, and that message
5371 remained on the queue past the ultimate address timeout, every queue runner
5372 would try a delivery (because it was past the ultimate address timeout) but
5373 after another failure, the ultimate address timeout, which should have then
5374 bounced the address, did not kick in. This was a "< instead of <=" error;
5375 in most cases the first failure would have been in the next clock tick
5376 after the received time, and all would be well.
5378 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5379 being recognized when the domain list was tested by the match_domain
5380 condition in an expansion string.
5382 16. Added the ${str2b64: operator.
5384 17. Exim was always calling setrlimit() to set a large limit for the number of
5385 processes, without checking whether the existing limit was already
5386 adequate. (It did check for the limit on file descriptors.) Furthermore,
5387 errors from getrlimit() and setrlimit() were being ignored. Now they are
5388 logged to the main and panic logs, but Exim does carry on, to try to do its
5389 job under whatever limits there are.
5391 18. Imported PCRE 5.0.
5393 19. Trivial typo in log message " temporarily refused connection" (the leading
5396 20. If the log selector return_path_on_delivery was set and an address was
5397 redirected to /dev/null, the delivery process crashed because it assumed
5398 that a return path would always be set for a "successful" delivery. In this
5399 case, the whole delivery is bypassed as an optimization, and therefore no
5402 21. Internal re-arrangement: the function for sending a challenge and reading
5403 a response while authentication was assuming a zero-terminated challenge
5404 string. It's now changed to take a pointer and a length, to allow for
5405 binary data in such strings.
5407 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5409 23. Exim was not respecting finduser_retries when seeking the login of the
5410 uid under which it was called; it was always trying 10 times. (The default
5411 setting of finduser_retries is zero.) Also, it was sleeping after the final
5412 failure, which is pointless.
5414 24. Implemented tls_on_connect_ports.
5416 25. Implemented acl_smtp_predata.
5418 26. If the domain in control=submission is set empty, Exim assumes that the
5419 authenticated id is a complete email address when it generates From: or
5420 Sender: header lines.
5422 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5423 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5424 chown and chgrp in /bin and hostname in /usr/bin.
5426 28. Exim was keeping the "process log" file open after each use, just as it
5427 does for the main log. This opens the possibility of it remaining open for
5428 long periods when the USR1 signal hits a daemon. Occasional processlog
5429 errors were reported, that could have been caused by this. Anyway, it seems
5430 much more sensible not to leave this file open at all, so that is what now
5433 29. The long-running daemon process does not normally write to the log once it
5434 has entered its main loop, and it closes the log before doing so. This is
5435 so that log files can straightforwardly be renamed and moved. However,
5436 there are a couple of unusual error situations where the daemon does write
5437 log entries, and I had neglected to close the log afterwards.
5439 30. The text of an SMTP error response that was received during a remote
5440 delivery was being truncated at 512 bytes. This is too short for some of
5441 the long messages that one sometimes sees. I've increased the limit to
5444 31. It is now possible to make retry rules that apply only when a message has a
5445 specific sender, in particular, an empty sender.
5447 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5448 it possible to be selective about when SMTP synchronization is enforced.
5450 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5452 32. Implemented hosts_connection_nolog.
5454 33. Added an ACL for QUIT.
5456 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5459 35. Added mailbox_size and mailbox_filecount to appendfile.
5461 36. Added control = no_multiline_responses to ACLs.
5463 37. There was a bug in the logic of the code that waits for the clock to tick
5464 in the case where the clock went backwards by a substantial amount such
5465 that the microsecond fraction of "now" was more than the microsecond
5466 fraction of "then" (but the whole seconds number was less).
5468 38. Added support for the libradius Radius client library this is found on
5469 FreeBSD (previously only the radiusclient library was supported).
5475 1. When certain lookups returned multiple values in the form name=value, the
5476 quoting of the values was not always being done properly. Specifically:
5477 (a) If the value started with a double quote, but contained no whitespace,
5479 (b) If the value contained whitespace other than a space character (i.e.
5480 tabs or newlines or carriage returns) it was not quoted.
5481 This fix has been applied to the mysql and pgsql lookups by writing a
5482 separate quoting function and calling it from the lookup code. The fix
5483 should probably also be applied to nisplus, ibase and oracle lookups, but
5484 since I cannot test any of those, I have not disturbed their existing code.
5486 2. A hit in the callout cache for a specific address caused a log line with no
5487 reason for rejecting RCPT. Now it says "Previous (cached) callout
5488 verification failure".
5490 3. There was an off-by-one bug in the queryprogram router. An over-long
5491 return line was truncated at 256 instead of 255 characters, thereby
5492 overflowing its buffer with the terminating zero. As well as fixing this, I
5493 have increased the buffer size to 1024 (and made a note to document this).
5495 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5496 when Exim is waiting for an SMTP response from a remote server, Exim
5497 restarts its select() call on the socket, thereby resetting its timeout.
5498 This is not a problem when such interrupts are rare. Somebody set up a cron
5499 job to run exiwhat every 2 minutes, which is less than the normal select()
5500 timeout (5 or 10 minutes). This meant that the select() timeout never
5501 kicked in because it was always reset. I have fixed this by comparing the
5502 time when an interrupt arrives with the time at the start of the first call
5503 to select(). If more time than the timeout has elapsed, the interrupt is
5504 treated as a timeout.
5506 5. Some internal re-factoring in preparation for the addition of Sieve
5507 extensions (by MH). In particular, the "personal" test is moved to a
5508 separate function, and given an option for scanning Cc: and Bcc: (which is
5509 not set for Exim filters).
5511 6. When Exim created an email address using the login of the caller as the
5512 local part (e.g. when creating a From: or Sender: header line), it was not
5513 quoting the local part when it contained special characters such as @.
5515 7. Installed new OpenBSD configuration files.
5517 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5518 try to make them clearer.
5520 9. Callout options, other than the timeout value, were being ignored when
5521 verifying sender addresses in header lines. For example, when using
5523 verify = header_sender/callout=no_cache
5525 the cache was (incorrectly) being used.
5527 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5528 only the Cygwin environment.
5530 11. When return_path_on_delivery was set as a log selector, if different remote
5531 addresses in the same message used different return paths and parallel
5532 remote delivery occurred, the wrong values would sometimes be logged.
5533 (Whenever a remote delivery process finished, the return path value from
5534 the most recently started remote delivery process was logged.)
5536 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5537 lines when AUTH and/or TLS are in use. This is the "received protocol"
5538 field. Exim used to use "asmtp" for authenticated SMTP, without any
5539 indication (in the protocol name) for TLS use. Now it follows the RFC and
5540 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5541 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5542 names appear in log lines as well as in Received: header lines.
5544 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5545 extensions, and comparison tests, and to fix some bugs.
5547 14. Changes to the "personal" filter test:
5549 (1) The test was buggy in that it was just doing the equivalent of
5550 "contains" tests on header lines. For example, if a user's address was
5551 anne@some.where, the "personal" test would incorrectly be true for
5553 To: susanne@some.where
5555 This test is now done by extracting each address from the header in turn,
5556 and checking the entire address. Other tests that are part of "personal"
5557 are now done using regular expressions (for example, to check local parts
5558 of addresses in From: header lines).
5560 (2) The list of non-personal local parts in From: addresses has been
5561 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5562 taken from the Sieve specification recommendations.
5564 (3) If the message contains any header line starting with "List-" it is
5565 treated as non-personal.
5567 (4) The test for "circular" in the Subject: header line has been removed
5568 because it now seems ill-conceived.
5570 15. Minor typos in src/EDITME comments corrected.
5572 16. Installed latest exipick from John Jetmore.
5574 17. If headers_add on a router specified a text string that was too long for
5575 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5576 of string_sprintf() is now avoided.
5578 18. $message_body_size was not set (it was always zero) when running the DATA
5579 ACL and the local_scan() function.
5581 19. For the "mail" command in an Exim filter, no default was being set for
5582 the once_repeat time, causing a random time value to be used if "once" was
5583 specified. (If the value happened to be <= 0, no repeat happened.) The
5584 default is now 0s, meaning "never repeat". The "vacation" command was OK
5585 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5586 (I found it when inspecting the code).
5588 20. There is now an overall timeout for performing a callout verification. It
5589 defaults to 4 times the callout timeout, which applies to individual SMTP
5590 commands during the callout. The overall timeout applies when there is more
5591 than one host that can be tried. The timeout is checked before trying the
5592 next host. This prevents very long delays if there are a large number of
5593 hosts and all are timing out (e.g. when the network connections are timing
5594 out). The value of the overall timeout can be changed by specifying an
5595 additional sub-option for "callout", called "maxwait". For example:
5597 verify = sender/callout=5s,maxwait=20s
5599 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5600 to the end before writing, but this should make it even safer).
5602 22. Exim was forgetting that it had advertised PIPELINING for the second and
5603 subsequent messages on an SMTP connection. It was also not resetting its
5604 memory on STARTTLS and an internal HELO.
5606 23. When Exim logs an SMTP synchronization error within a session, it now
5607 records whether PIPELINING has been advertised or not.
5609 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5610 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5611 rather than long int.
5613 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5615 26. Added the never_mail option to autoreply.
5621 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5622 crash if the getsockname() call failed; this can happen if a connection is
5623 closed very soon after it is established. The problem was simply in the
5624 order in which certain operations were done, causing Exim to try to write
5625 to the SMTP stream before it had set up the file descriptor. The bug has
5626 been fixed by making things happen in the correct order.
5632 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5633 before the connection was closed, thus losing the rejection response.
5635 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5636 some early Solaris releases, but causes trouble in current releases where
5637 socklen_t is defined.
5639 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5642 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5645 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5646 that releasing the top of it at the end releases what was used for sub-
5647 expansions (unless the block got too big). However, discard this block if
5648 the first thing is a variable or header, so that we can use its block when
5649 it is dynamic (useful for very large $message_headers, for example).
5651 6. Lookups now cache *every* query, not just the most recent. A new, separate
5652 store pool is used for this. It can be recovered when all lookup caches are
5653 flushed. Lookups now release memory at the end of their result strings.
5654 This has involved some general refactoring of the lookup sources.
5656 7. Some code has been added to the store_xxx() functions to reduce the amount
5657 of flapping under certain conditions.
5659 8. log_incoming_interface used to affect only the <= reception log lines. Now
5660 it causes the local interface and port to be added to several more SMTP log
5661 lines, for example "SMTP connection from", and rejection lines.
5663 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5665 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5667 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5668 could be overwritten at the end of the current message (or the start of a
5669 new message if it was set in a HELO ACL). The value is now preserved for
5670 the duration of the SMTP connection.
5672 12. If a transport had a headers_rewrite setting, and a matching header line
5673 contained an unqualified address, that address was qualified, even if it
5674 did not match any rewriting rules. The underlying bug was that the values
5675 of the flags that permit the existence of unqualified sender and recipient
5676 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5677 non-local messages, and by -bnq for local messages) were not being
5678 preserved with the message after it was received.
5680 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5681 "next input=" as part of the text comprising the host identity instead of
5682 the correct text. The code was using the same buffer for two different
5683 strings. However, depending on which order the printing function evaluated
5684 its arguments, the bug did not always show up. Under Linux, for example, my
5685 test suite worked just fine.
5687 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5688 doesn't work with some older versions of Perl. It has been changed to "my",
5689 which in any case is probably the better facility to use.
5691 15. A really picky compiler found some instances of statements for creating
5692 error messages that either had too many or two few arguments for the format
5695 16. The size of the buffer for calls to the DNS resolver has been increased
5696 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5697 for addresses that have a lot of PTR records. This alleviates a problem; it
5698 does not fully solve it.
5700 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5701 buffer now truncates the list and logs the incident, which is the same
5702 action as happens when Exim is looking up a host name and its aliases.
5703 Previously in this situation something unpredictable would happen;
5704 sometimes it was "internal error: store_reset failed".
5706 18. If a server dropped the connection unexpectedly when an Exim client was
5707 using GnuTLS and trying to read a response, the client delivery process
5708 crashed while trying to generate an error log message.
5710 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5711 string, for example:
5713 warn message = H1: something\nH2: something
5715 the text was added as a single header line from Exim's point of view
5716 though it ended up OK in the delivered message. However, searching for the
5717 second and subsequent header lines using $h_h2: did not work. This has been
5718 fixed. Similarly, if a system filter added multiple headers in this way,
5719 the routers could not see them.
5721 20. Expanded the error message when iplsearch is called with an invalid key to
5722 suggest using net-iplsearch in a host list.
5724 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5725 ACLs are no longer actually imposed (and a message to that effect is
5728 22. If a "gecos" field in a passwd entry contained escaped characters, in
5729 particular, if it contained a \" sequence, Exim got it wrong when building
5730 a From: or a Sender: header from that name. A second bug also caused
5731 incorrect handling when an unquoted " was present following a character
5732 that needed quoting.
5734 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5735 was not being matched caselessly.
5737 24. Arranged for all hyphens in the exim.8 source to be escaped with
5740 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5741 properly. Recipient callout cache records were still being keyed to include
5742 the sender, even when use_sender was set false. This led to far more
5743 callouts that were necessary. The sender is no longer included in the key
5744 when use_sender is false.
5746 26. Added "control = submission" modifier to ACLs.
5748 27. Added the ${base62d: operator to decode base 62 numbers.
5750 28. dnsdb lookups can now access SRV records.
5752 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5753 the configuration file.
5755 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5756 (-v) mode. This makes the output for a verbose queue run more intelligible.
5758 31. Added a use_postmaster feature to recipient callouts.
5760 32. Added the $body_zerocount variable, containing the number of binary zero
5761 bytes in the message body.
5763 33. The time of last modification of the "new" subdirectory is now used as the
5764 "mailbox time last read" when there is a quota error for a maildir
5767 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5769 35. Added +ignore_unknown as a special item in host lists.
5771 36. Code for decoding IPv6 addresses in host lists is now included, even if
5772 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5773 address was recognized as an IP address, but was then not correctly decoded
5774 into binary, causing unexpected and incorrect effects when compared with
5781 1. Very minor rewording of debugging text in manualroute to say "list of
5782 hosts" instead of "hostlist".
5784 2. If verify=header_syntax was set, and a header line with an unqualified
5785 address (no domain) and a large number of spaces between the end of the
5786 name and the colon was received, the reception process suffered a buffer
5787 overflow, and (when I tested it) crashed. This was caused by some obsolete
5788 code that should have been removed. The fix is to remove it!
5790 3. When running in the test harness, delay a bit after writing a bounce
5791 message to get a bit more predictability in the log output.
5793 4. Added a call to search_tidyup() just before forking a reception process. In
5794 theory, someone could use a lookup in the expansion of smtp_accept_max_
5795 per_host which, without the tidyup, could leave open a database connection.
5797 5. Added the variables $recipient_data and $sender_data which get set from a
5798 lookup success in an ACL "recipients" or "senders" condition, or a router
5799 "senders" option, similar to $domain_data and $local_part_data.
5801 6. Moved the writing of debug_print from before to after the "senders" test
5804 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5805 problems for message scanning, either using a data ACL, or using
5806 local_scan() because the Received: header was not generated till after they
5807 were called (in order to set the time as the time of reception completion).
5808 I have revised the way this works. The header is now generated after the
5809 body is received, but before the ACL or local_scan() are called. After they
5810 are run, the timestamp in the header is updated.
5816 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5817 before starting a queue runner without re-exec. This happened only when
5818 deliver_drop_privilege was set or when the Exim user was set to root. The
5819 effect of the bug was that timeouts during subsequent deliveries caused
5820 crashes instead of being properly handled. The handler is now left at its
5821 default (and expected) setting.
5823 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5824 message, again when deliver_drop_privilege is set or Exim is run as root.
5825 The bug described in (1) was not present in this case, but the tidying up
5826 of the other signals was missing. I have made the two cases consistent.
5828 3. The ignore_target_hosts setting on a manualroute router was being ignored
5829 for hosts that were looked up using the /MX notation.
5831 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5834 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5835 operated on the sender address. After changing the $sender_address to <>
5836 for the sender address verify, Exim was re-instated it as the original
5837 (before rewriting) address, but remembering that it had rewritten it, so it
5838 wasn't rewriting it again. This bug also had the effect of breaking the
5839 sender address verification caching when the sender address was rewritten.
5841 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5842 This has been changed so that if the ip literal address matches
5843 ignore_target_hosts, the router declines.
5845 7. Added expansion conditions match_domain, match_address, and match_local_
5846 part (NOT match_host).
5848 8. The placeholder for the Received: header didn't have a length field set.
5850 9. Added code to Exim itself and to exim_lock to test for a specific race
5851 condition that could lead to file corruption when using MBX delivery. The
5852 issue is with the lockfile that is created in /tmp. If this file is removed
5853 after a process has opened it but before that process has acquired a lock,
5854 there is the potential for a second process to recreate the file and also
5855 acquire a lock. This could lead to two Exim processes writing to the file
5856 at the same time. The added code performs the same test as UW imapd; it
5857 checks after acquiring the lock that its file descriptor still refers to
5858 the same named file.
5860 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5861 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5864 11. Added the smtp_active_hostname option. If used, this will typically be made
5865 to depend on the incoming interface address. Because $interface_address is
5866 not set up until the daemon has forked a reception process, error responses
5867 that can happen earlier (such as "too many connections") no longer contain
5870 12. If an expansion in a condition on a "warn" statement fails because a lookup
5871 defers, the "warn" statement is abandoned, and the next ACL statement is
5872 processed. Previously this caused the whole ACL to be aborted.
5874 13. Added the iplsearch lookup type.
5876 14. Added ident_timeout as a log selector.
5878 15. Added tls_certificate_verified as a log selector.
5880 16. Added a global option tls_require_ciphers (compare the smtp transport
5881 option of the same name). This controls incoming TLS connections.
5883 17. I finally figured out how to make tls_require_ciphers do a similar thing
5884 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5885 before starting the TLS session.
5887 18. Tabs are now shown as \t in -bP output.
5889 19. If the log selector return_path_on_delivery was set, Exim crashed when
5890 bouncing a message because it had too many Received: header lines.
5892 20. If two routers both had headers_remove settings, and the first one included
5893 a superfluous trailing colon, the final name in the first list and the
5894 first name in the second list were incorrectly joined into one item (with a
5895 colon in the middle).
5901 1. Added -C and -D options to the exinext utility, mainly to make it easier
5902 to include in the automated testing, but these could be helpful when
5903 multiple configurations are in use.
5905 2. The exinext utility was not formatting the output nicely when there was
5906 an alternate port involved in the retry record key, nor when there was a
5907 message id as well (for retries that were specific to a specific message
5908 and a specific host). It was also confused by IPv6 addresses, because of
5909 the additional colons they contain. I have fixed the IPv4 problem, and
5910 patched it up to do a reasonable job for IPv6.
5912 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5913 delivery, the log line now contains "pipelined" if PIPELINING was used.
5915 4. An SMTP transport process used to panic and die if the bind() call to set
5916 an explicit outgoing interface failed. This has been changed; it is now
5917 treated in the same way as a connect() failure.
5919 5. A reference to $sender_host_name in the part of a conditional expansion
5920 that was being skipped was still causing a DNS lookup. This no longer
5923 6. The def: expansion condition was not recognizing references to header lines
5924 that used bh_ and bheader_.
5926 7. Added the _cache feature to named lists.
5928 8. The code for checking quota_filecount in the appendfile transport was
5929 allowing one more file than it should have been.
5931 9. For compatibility with Sendmail, the command line option
5939 and sets the incoming protocol and host name (for trusted callers). The
5940 host name and its colon can be omitted when only the protocol is to be set.
5941 Note the Exim already has two private options, -pd and -ps, that refer to
5942 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5943 "s", but I don't think that's a major issue.
5945 10. A number of refactoring changes to the code, none of which should affect
5948 (a) The number of logging options was getting close to filling up the
5949 32-bit word that was used as a bit map. I have split them into two classes:
5950 those that are passed in the argument to log_write(), and those that are
5951 only ever tested independently outside of that function. These are now in
5952 separate 32-bit words, so there is plenty of room for expansion again.
5953 There is no change in the user interface or the logging behaviour.
5955 (b) When building, for example, log lines, the code previously used a
5956 macro that called string_cat() twice, in order to add two strings. This is
5957 not really sufficiently general. Furthermore, there was one instance where
5958 it was actually wrong because one of the argument was used twice, and in
5959 one call a function was used. (As it happened, calling the function twice
5960 did not affect the overall behaviour.) The macro has been replaced by a
5961 function that can join an arbitrary number of extra strings onto a growing
5964 (c) The code for expansion conditions now uses a table and a binary chop
5965 instead of a serial search (which was left over from when there were very
5966 few conditions). Also, it now recognizes conditions like "pam" even when
5967 the relevant support is not compiled in: a suitably worded error message is
5968 given if an attempt is made to use such a condition.
5970 11. Added ${time_interval:xxxxx}.
5972 12. A bug was causing one of the ddress fields not to be passed back correctly
5973 from remote delivery subprocesses. The field in question was not being
5974 subsequently used, so this caused to problems in practice.
5976 13. Added new log selectors queue_time and deliver_time.
5978 14. Might have fixed a bug in maildirsizefile handling that threw up
5979 "unexpected character" debug warnings, and recalculated the data
5980 unnecessarily. In any case, I expanded the warning message to give more
5983 15. Added the message "Restricted characters in address" to the statements in
5984 the default ACL that block characters like @ and % in local parts.
5986 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5987 Three changes have been made:
5989 (a) There was a serious bug; a negative response to MAIL caused the whole
5990 recipient domain to be cached as invalid, thereby blocking all messages
5991 to all local parts at the same domain, from all senders. This bug has
5992 been fixed. The domain is no longer cached after a negative response to
5993 MAIL if the sender used is not empty.
5995 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5998 (c) A new callout option, "use_sender" has been added for people who want
5999 the modified behaviour.
6005 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6008 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6009 indeed breaks things for older releases.
6011 3. Added additional logging to the case where there is a problem reading data
6012 from a filter that is running in a subprocess using a pipe, in order to
6013 try to track down a specific problem.
6015 4. Testing facility fudge: when running in the test harness and attempting
6016 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6017 getting "No route to host". Convert this to a timeout.
6019 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6022 6. Some OS don't have socklen_t but use size_t instead. This affects the
6023 fifth argument of getsockopt() amongst other things. This is now
6024 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6025 can be set for individual OS. I have set it for SunOS5, OSF1, and
6026 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6027 some earlier ones do not.
6029 7. Change 4.30/15 was not doing the test caselessly.
6031 8. The standard form for an IPv6 address literal was being rejected by address
6032 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6033 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6034 this, as well as the form without the "IPv6" on the front (but only when
6035 address literals are enabled, of course).
6037 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6039 10. Exim crashed if a message with an empty sender address specified by -f
6040 encountered a router with an errors_to setting. This could be provoked only
6041 by a command such as
6045 where an empty string was supplied; "<>" did not hit this bug.
6047 11. Installed PCRE release 4.5.
6049 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6050 remained set. It is now erased.
6052 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6053 times from message ids (which are base 36 rather than the normal 62).
6055 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6056 were being counted as actual protocol errors, and logged if the log
6057 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6058 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6059 and DATA following a set of rejected RCPTs do not count as protocol errors.
6060 In other words, Exim assumes they were pipelined, though this may not
6061 actually be the case. Of course, in all cases the client gets an
6062 appropriate error code.
6064 15. If a lookup fails in an ACL condition, a message about the failure may
6065 be available; it is used if testing the ACL cannot continue, because most
6066 such messages specify what the cause of the deferral is. However, some
6067 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6068 that caused an old message to be retained and used if a later statement
6069 caused a defer, replacing the real cause of the deferral.
6071 16. If an IP address had so many PTR records that the DNS lookup buffer
6072 was not large enough to hold them, Exim could crash while trying to process
6073 the truncated data. It now detects and logs this case.
6075 17. Further to 4.21/58, another change has been made: if (and only if) the
6076 first line of a message (the first header line) ends with CRLF, a bare LF
6077 in a subsequent header line has a space inserted after it, so as not to
6078 terminate the header.
6080 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6081 unnecessarily, used atoi() instead of strtol(), and didn't check the
6082 termination when getting file sizes from file names by regex.
6084 19. Completely re-implemented the support for maildirsize files, in the light
6085 of a number of problems with the previous contributed implementation
6086 (4.30/29). In particular:
6088 . If the quota is zero, the maildirsize file is maintained, but no quota is
6091 . If the maildir directory does not exist, it is created before any attempt
6092 to write a maildirsize file.
6094 . The quota value in the file is just a cache; if the quota is changed in
6095 the transport, the new value overrides.
6097 . A regular expression is available for excluding directories from the
6100 20. The autoreply transport checks the characters in options that define the
6101 message's headers; it allows continued headers, but it was checking with
6102 isspace() after an embedded newline instead of explicitly looking for a
6105 21. If all the "regular" hosts to which an address was routed had passed their
6106 expiry times, and had not reached their retry times, the address was
6107 bounced, even if fallback hosts were defined. Now Exim should go on to try
6110 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6111 equivalent code in the SMTP transport. Some hosts send humungous responses
6112 to HELO/EHLO, more than 1024 it seems.
6114 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6115 gives compiler warnings in some environments. I've now done it "properly",
6118 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6119 (because of problems with the built-in one) was declared to return uschar *
6120 instead of char *, causing compiler failure.
6122 25. Fixed a file descriptor leak when processing alias/forward files.
6124 26. Fixed a minor format string issue in dbfn.c.
6126 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6128 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6129 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6132 29. When a sender address is verified, it is cached, to save repeating the test
6133 when there is more than one recipient in a message. However, when the
6134 verification involves a callout, it is possible for different callout
6135 options to be set for different recipients. It is too complicated to keep
6136 track of this in the cache, so now Exim always runs a verification when a
6137 callout is required, relying on the callout cache for the optimization.
6138 The overhead is duplication of the address routing, but this should not be
6141 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6142 to be verified with callout=postmaster, and the main callout worked but the
6143 postmaster check failed, the verification correctly failed. However, if a
6144 subsequent RCPT command asked for sender verification *without* the
6145 postmaster check, incorrect caching caused this verification also to fail,
6148 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6149 it was not caching the DNS options (qualify_single, search_parents) that
6150 were used when the lookup failed. A subsequent lookup with different
6151 options therefore always gave the same answer, though there were cases
6152 where it should not have. (Example: a "domains = !$mx_any" option on a
6153 dnslookup router: the "domains" option is always processed without any
6154 widening, but the router might have qualify_single set.) Now Exim uses the
6155 cached value only when the same options are set.
6157 32. Added John Jetmore's "exipick" utility to the distribution.
6159 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6160 than a timeout (e.g. a certificate is required, and is not provided), an
6161 Exim server now closes the connection immediately. Previously it waited for
6162 the client to close - but if the client is SSL, it seems that they each
6163 wait for each other, leading to a delay before one of them times out.
6165 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6166 maintained 0.8.x compatibility because I don't think many are using it, and
6167 it is clearly obsolete.
6169 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6172 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6173 client certificate was expired. A simple patch fixes this, though I don't
6174 understand the full logic of why the verify callback is called multiple
6177 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6178 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6179 which causes problems with some clients (such as the Certicom SSL Plus
6180 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6181 disables the coutermeasure allowing Eudora to connect."
6183 38. Exim was not checking that a write() to a log file succeeded. This could
6184 lead to Bad Things if a log got too big, in particular if it hit a file
6185 size limit. Exim now panics and dies if it cannot write to a log file, just
6186 as it does if it cannot open a log file.
6188 39. Modified OS/Makefile-Linux so that it now contains
6190 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6192 The two -D definitions ensure that Exim is compiled with large file
6193 support, which makes it possible to handle log files that are bigger than
6196 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6197 instance) a domain was checked against a named list that involved a lookup,
6198 causing $domain_data to be set, then another domain was checked against the
6199 same list, then the first domain was re-checked, the value of $domain_data
6200 after the final check could be wrong. In particular, if the second check
6201 failed, it could be set empty. This bug probably also applied to
6204 41. The strip_trailing_dot option was not being applied to the address given
6205 with the -f command-line option.
6207 42. The code for reading a message's header from the spool was incrementing
6208 $received_count, but never initializing it. This meant that the value was
6209 incorrect (doubled) while delivering a message in the same process in which
6210 it was received. In the most common configuration of Exim, this never
6211 happens - a fresh exec is done - but it can happen when
6212 deliver_drop_privilege is set.
6214 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6215 it now includes up to 150 characters of the unexpected data in the log
6218 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6219 and building data strings. The size of both of these buffers was 10 000
6220 bytes - far larger than anybody would *ever* want, thought I. Needless to
6221 say, somebody hit the limit. I have increased the maximum line length to
6222 20 000 and the maximum data length of concatenated lines to 100 000. I have
6223 also fixed two bugs, because there was no checking on these buffers. Tsk,
6224 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6225 buffer is too small.
6227 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6228 lsearch lookups. Now it does.
6230 46. When parsing a route_list item in a manualroute router, a fixed-length
6231 buffer was used for the list of hosts. I made this 1024 bytes long,
6232 thinking that nobody would ever have a list of hosts that long. Wrong.
6233 Somebody had a whole pile of complicated expansion conditions, and the
6234 string was silently truncated, leading to an expansion error. It turns out
6235 that it is easier to change to an unlimited length (owing to other changes
6236 that have happened since this code was originally written) than to build
6237 structure for giving a limitation error. The length of the item that
6238 expands into the list of hosts is now unlimited.
6240 47. The lsearch lookup could not handle data where the length of text line was
6241 more than 4095 characters. Such lines were truncated, leading to shortened
6242 data being returned. It should now handle lines of any length.
6244 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6245 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6248 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6249 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6250 no longer happens in comment lines. A list of replacements is now placed
6251 at the head of all of the source files, except those whose only change is
6252 to replace PERL_COMMAND in the very first #! line.
6254 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6255 messages on the queue, with a bottom-up merge sort, using code contributed
6256 by Michael Haardt. This should make operations like -bp somewhat faster on
6257 large queues. It won't affect queue runners, except when queue_run_in_order
6260 51. Installed eximstats 1.31 in the distribution.
6262 52. Added support for SRV lookups to the dnslookup router.
6264 53. If an ACL referred to $message_body or $message_body_end, the value was not
6265 reset for any messages that followed in the same SMTP session.
6267 54. The store-handling optimization for building very long strings was not
6268 differentiating between the different store pools. I don't think this
6269 actually made any difference in practice, but I've tidied it.
6271 55. While running the routers to verify a sender address, $sender_address
6272 was still set to the sender address. This is wrong, because when routing to
6273 send a bounce to the sender, it would be empty. Therefore, I have changed
6274 it so that, while verifying a sender address, $sender_address is set to <>.
6275 (There is no change to what happens when verifying a recipient address.)
6277 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6278 target A or AAAA records (if not already returned) without resetting the
6279 qualify_single or search_parents options of the DNS resolver. These are
6280 inappropriate in this case because the targets of MX and SRV records must
6281 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6282 target that, when qualified, matched something in the local domain. These
6283 two options are now turned off when doing these lookups.
6285 57. It seems that at least some releases of Reiserfs (which does not have the
6286 concept of a fixed number of inodes) returns zero and not -1 for the
6287 number of available inodes. This interacted badly with check_spool_inodes,
6288 which assumed that -1 was the "no such thing" setting. What I have done is
6289 to check that the total number of inodes is greater than zero before doing
6290 the test of how many are available.
6292 58. When a "warn" ACL statement has a log_message modifier, the message is
6293 remembered, and not repeated. This is to avoid a lot of repetition when a
6294 message has many recipients that cause the same warning to be written.
6295 However, Exim was preserving the list of already written lines for an
6296 entire SMTP session, which doesn't seem right. The memory is now reset if a
6297 new message is started.
6299 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6300 debugging output unless log_rewrite was also set.
6302 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6303 of (int)(handle) when we know that handle contains (void *)(-1).
6305 61. The Exim daemon panic-logs an error return when it closes the incoming
6306 connection. However "connection reset by peer" seems to be common, and
6307 isn't really an error worthy of noting specially, so that particular error
6310 62. When Exim is trying to find all the local interfaces, it used to panic and
6311 die if the ioctl to get the interface flags failed. However, it seems that
6312 on at least one OS (Solaris 9) it is possible to have an interface that is
6313 included in the list of interfaces, but for which you get a failure error
6314 for this call. This happens when the interface is not "plumbed" into a
6315 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6316 failure of the "get flags" call assumes that the interface is down.
6318 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6319 makes life easier for people who are doing arithmetic on fields extracted
6320 from dates, where you often get leading zeros that should not be
6321 interpreted as octal.
6323 64. Added qualify_domain to the redirect router, to override the global
6326 65. If a pathologically long header line contained very many addresses (the
6327 report of this problem mentioned 10 000) and each of them was rewritten,
6328 Exim could use up a very large amount of memory. (It kept on making new
6329 copies of the header line as it rewrote, and never released the old ones.)
6330 At the expense of a bit more processing, the header rewriting function has
6331 been changed so that it no longer eats memory in this way.
6333 66. The generation of the Received: header has been moved from the time that a
6334 message starts to be received, to the time that it finishes. The timestamp
6335 in the Received: header should now be very close to that of the <= log
6336 line. There are two side-effects of this change:
6338 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6339 logged header lines no longer include the local Received: line, because
6340 it has not yet been created. The same applies to a copy of the message
6341 that is returned to a non-SMTP sender when a message is rejected.
6343 (b) When a filter file is tested using -bf, no additional Received: header
6344 is added to the test message. After some thought, I decided that this
6347 This change does not affect the value of $received_for. It is still set
6348 after address rewriting, but before local_scan() is called.
6350 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6352 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6353 gave an unhelpful panic error message, and a defer error. I have managed to
6354 change this behaviour so that it now rejects any supplied certificate,
6355 which seems right, as the list of acceptable certificates is empty.
6357 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6358 gave an unhelpful defer error. I have not managed to make this reject any
6359 supplied certificates, but the error message it gives is "no certificate
6360 supplied", which is not helpful.
6362 70. exigrep's output now also includes lines that are not associated with any
6363 message, but which match the given pattern. Implemented by a patch from
6364 Martin Sluka, which also tidied up the Perl a bit.
6366 71. Recipient callout verification, like sender verification, was using <> in
6367 the MAIL FROM command. This isn't really the right thing, since the actual
6368 sender may affect whether the remote host accepts the recipient or not. I
6369 have changed it to use the actual sender in the callout; this means that
6370 the cache record is now keyed on a recipient/sender pair, not just the
6371 recipient address. There doesn't seem to be a real danger of callout loops,
6372 since a callout by the remote host to check the sender would use <>.
6373 [SEE ABOVE: changed after hitting problems.]
6375 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6376 temporary errors. However, in the case of such a code being given after
6377 the end of a data transmission (i.e. after ".") Exim was failing to write
6378 a retry record for the message. (Yes, there was some broken host that was
6379 actually sending 8xx at this point.)
6381 73. An unknown lookup type in a host list could cause Exim to panic-die when
6382 the list was checked. (An example that provoked this was putting <; in the
6383 middle of a list instead of at the start.) If this happened during a DATA
6384 ACL check, a -D file could be left lying around. This kind of configuration
6385 error no longer causes Exim to die; instead it causes a defer error. The
6386 incident is still logged to the main and panic logs.
6388 74. Buglet left over from Exim 3 conversion. The message "too many messages
6389 in one connection" was written to the rejectlog but not the mainlog, except
6390 when address rewriting (yes!) was being logged.
6392 75. Added write_rejectlog option.
6394 76. When a system filter was run not as root (that is, when system_filter_user
6395 was set), the values of the $n variables were not being returned to the
6396 main process; thus, they were not subsequently available in the $sn
6399 77. Added +return_path_on_delivery log selector.
6401 78. A connection timeout was being treated differently from recipients deferred
6402 when testing hosts_max_try with a message that was older than the host's
6403 retry timeout. (The host should not be counted, thus allowing all hosts to
6404 be tried at least once before bouncing.) This may have been the cause of an
6405 occasionally reported bug whereby a message would remain on the queue
6406 longer than the retry timeout, but would be bounced if a delivery was
6407 forced. I say "may" because I never totally pinned down the problem;
6408 setting up timeout/retry tests is difficult. See also the next item.
6410 79. The ultimate address timeout was not being applied to errors that involved
6411 a combination of host plus message (for example, a timeout on a MAIL
6412 command). When an address resolved to a number of possible hosts, and they
6413 were not all tried for each delivery (e.g. because of hosts_max_try), a
6414 message could remain on the queue longer than the retry timeout.
6416 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6419 81. Fixed an obscure SMTP outgoing bug which required at least the following
6420 conditions: (a) there was another message waiting for the same server;
6421 (b) the server returned 5xx to all RCPT commands in the first message so
6422 that the message was not completed; (c) the server dropped the connection
6423 or gave a negative response to the RSET that Exim sends to abort the
6424 transaction. The observed case was a dropped connection after DATA that had
6425 been sent in pipelining mode. That is, the server had advertised PIPELINING
6426 but was not implementing it correctly. The effect of the bug was incorrect
6427 behaviour, such as trying another host, and this could lead to a crash.
6433 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6434 and daemon.c were passed as pointers to ints; they should have been
6435 pointers to socklen_t variables (which are typically unsigned ints).
6437 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6440 3. Fixed a really odd bug that affected only the testing scheme; patching a
6441 certain fixed string in the binary changed the value of another string that
6442 happened to be identical to the end of the original first string.
6444 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6445 name", it returns that address as the IP address. On some operating
6446 systems (e.g. Solaris), it also passes back the IP address string as the
6447 "host name". However, on others (e.g. Linux), it passes back an empty
6448 string. Exim wasn't checking for this, and was changing the host name to an
6449 empty string, assuming it had been canonicalized.
6451 5. Although rare, it is permitted to have more than one PTR record for a given
6452 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6453 all the names associated with an address, because they do in Solaris.
6454 However, it seems that they do not in Linux for data that comes from the
6455 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6456 I found this out when I moved to a new Linux workstation and tried to run
6457 the Exim test suite.
6459 To get round this problem I have changed the code so that it now does its
6460 own call to the DNS to look up PTR records when searching for a host name.
6461 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6462 addresses that are only in /etc/hosts are still found.
6464 This behaviour is, however, controlled by an option called host_lookup_
6465 order, which defaults to "bydns:byaddr". If people want to use the other
6466 order, or indeed, just use one or the other means of lookup, they can
6467 specify it in this variable.
6469 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6470 some operating systems, this comes back from gethostbyaddr() as an empty
6471 string, and this is what Exim used to test for. However, it seems that in
6472 other systems, "." is yielded. Exim now tests for this case too.
6474 7. The values of check_spool_space and check_log_space are now held internally
6475 as a number of kilobytes instead of an absolute number of bytes. If a
6476 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6477 kilobyte. This means that much larger values can be stored.
6479 8. Exim monitor: an attempt to get the action menu when not actually pointing
6480 at a message produces an empty menu entitled "No message selected". This
6481 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6482 no entries in it ("Shell widget menu has zero width and/or height"). So I
6483 have added a single, blank menu entry in this case.
6485 9. Added ${quote_local_part.
6487 10. MIME decoding is now applied to the contents of Subject: header lines when
6490 11. Now that a reference to $sender_host_address automatically causes a reverse
6491 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6492 host lookup before query-style lookups in lists that might use this
6493 variable. This has therefore been abolished, and the "net-" prefix is no
6494 longer necessary for query-style lookups.
6496 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6497 have been a typo for LFLAGS, so it has been changed.
6499 13. The install script calls Exim with "-C /dev/null" in order to find the
6500 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6501 to be output. However, since Exim outputs its version number before the
6502 error, it didn't break the script. It just looked ugly. I fixed this by
6503 always allowing "-C /dev/null" if the caller is root.
6505 14. Ignore overlarge ACL variable number when reading spool file - insurance
6506 against a later release with more variables having written the file.
6508 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6509 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6510 this, as well as the form without the "IPv6" on the front.
6512 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6513 OS/Makefile-Darwin file.
6515 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6516 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6518 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6519 variable. However, some OS use ints for inodes. I've added cast to long int
6520 to get rid of the compiler warning.
6522 19. I had forgotten to lock out "/../" in configuration file names when
6523 ALT_CONFIG_PREFIX was set.
6525 20. Routers used for verification do not need to specify transports. However,
6526 if such a router generated a host list, and callout was configured, Exim
6527 crashed, because it could not find a port number from the (non-existent)
6528 transport. It now assumes port 25 in this circumstance.
6530 21. Added the -t option to exigrep.
6532 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6533 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6534 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6535 them is set, LOOKUP_LSEARCH is forced.
6537 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6539 24. Added sender and host information to the "rejected by local_scan()" log
6540 line; previously there was no indication of these.
6542 25. Added .include_if_exists.
6544 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6545 Linux. It turns out that not all file systems support this. Apparently some
6546 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6547 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6548 not supported on the file descriptor, is now ignored when Exim is trying to
6549 sync a directory. This applies only to Linux.
6551 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6553 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6554 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6557 29. Added support for maildirsize files from supplied patch (modified a bit).
6559 30. The use of :fail: followed by an empty string could lead Exim to respond to
6560 sender verification failures with (e.g.):
6562 550 Verification failed for <xxx>
6563 550 Sender verify failed
6565 where the first response line was missing the '-' that indicates it is not
6566 the final line of the response.
6568 31. The loop for finding the name of the user that called Exim had a hardwired
6569 limit of 10; it now uses the value of finduser_retries, which is used for
6570 all other user lookups.
6572 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6575 33. Exim was neglecting to zero errno before one call of strtol() when
6576 expanding a string and expecting an integer value. On some systems this
6577 resulted in spurious "integer overflow" errors. Also, it was casting the
6578 result into an int without checking.
6580 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6581 did not work. The code looks as if it has *never* worked, though it appears
6582 to have been documented since at least release 1.62. I have made it work.
6584 35. The "timeout_DNS" error in retry rules, also documented since at least
6585 1.62, also never worked. As it isn't clear exactly what this means, and
6586 clearly it isn't a major issue, I have abolished the feature by treating it
6587 as "timeout", and writing a warning to the main and panic logs.
6589 36. The display of retry rules for -brt wasn't always showing the error code
6592 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6593 timeout_connect_A, timeout_connect_MX.
6595 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6596 to the empty sender.
6598 39. The daemon was not analysing the content of -oX till after it had closed
6599 stderr and disconnected from the controlling terminal. This meant that any
6600 syntax errors were only noted on the panic log, and the return code from
6601 the command was 0. By re-arranging the code a little, I've made the
6602 decoding happen first, so such errors now appear on stderr, and the return
6603 code is 1. However, the actual setting up of the sockets still happens in
6604 the disconnected process, so errors there are still only recorded on the
6607 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6608 connections (as happens on some IP stacks) was logged at start up time as
6609 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6610 it from "IPv6 and IPv4", which means that two separate sockets are being
6613 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6614 says whether AF_INET or AF_INET6 was passed as an argument.
6616 42. Exiwhat output was messed up when time zones were included in log
6619 43. Exiwhat now gives more information about the daemon's listening ports,
6620 and whether -tls-on-connect was used.
6622 44. The "port" option of the smtp transport is now expanded.
6624 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6625 silently ignored. Now an error message is written to the main and panic
6628 46. There's a new ACL modifier called "logwrite" which writes to a log file
6629 as soon as it is encountered.
6631 47. Added $local_user_uid and $local_user_gid at routing time.
6633 48. Exim crashed when trying to verify a sender address that was being
6636 49. Exim was recognizing only a space character after ".include". It now also
6637 recognizes a tab character.
6639 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6640 extracting the relevant information from the specification. The man page no
6641 longer contains scrambled data for the -d option, and I've added a section
6642 at the front about calling Exim under different names.
6644 51. Added "extra_headers" argument to the "mail" command in filter files.
6646 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6649 53. Installed eximstats 1.29.
6651 54. Added transport_filter_timeout as a generic transport option.
6653 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6654 Cc: header lines. This was required by RFC 822, but it not required by RFC
6657 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6658 incoming messages that did not have them. Now it does so only if the
6659 message originates locally, that is, if there is no associated remote host
6660 address. When Resent- header lines are present, this applies to the Resent-
6661 lines rather than the non-Resent- lines.
6663 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6664 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6666 58. Messages for configuration errors now include the name of the main
6667 configuration file - useful now that there may be more than one file in a
6668 list (.included file names were always shown).
6670 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6671 for those rare installations that do not start the daemon as root or run it
6672 setuid root. I've cut out the call to initgroups() if the daemon is not
6675 60. The Exim user and group can now be bound into the binary as text strings
6676 that are looked up at the start of Exim's processing.
6678 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6680 62. Added $mailstore_basename variable.
6682 63. Installed patch to sieve.c from Michael Haardt.
6684 64. When Exim failed to open the panic log after failing to open the main log,
6685 the original message it was trying to log was written to stderr and debug
6686 output, but if they were not available (the usual case in production), it
6687 was lost. Now it is written to syslog before the two lines that record the
6688 failures to open the logs.
6690 65. Users' Exim filters run in subprocesses under the user's uid. It is
6691 possible for a "deliver" command or an alias in a "personal" command to
6692 provoke an address rewrite. If logging of address rewriting is configured,
6693 this fails because the process is not running as root or exim. There may be
6694 a better way of dealing with this, but for the moment (because 4.30 needs
6695 to be released), I have disabled address rewrite logging when running a
6696 filter in a non-root, non-exim process.
6702 1. The buildconfig auxiliary program wasn't quoting the value set for
6703 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6704 not defined. This bug was masked in 4.22 by the effect that was fixed in
6707 2. Some messages that were rejected after a message id was allocated were
6708 shown as "incomplete" by exigrep. It no longer does this for messages that
6709 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6711 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6712 have allow_domain_literals set, the ID did not get logged in the <= line.
6713 Domain literals are now always recognized in Message-ID: header lines.
6715 4. The first argument for a ${extract expansion item is the key name or field
6716 number. Leading and trailing spaces in this item were not being ignored,
6717 causing some misleading effects.
6719 5. When deliver_drop_privilege was set, single queue runner processes started
6720 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6721 same command in the process it spins off) were not dropping privilege.
6723 6. When the daemon running as "exim" started a queue runner, it always
6724 re-executed Exim in the spun-off process. This is a waste of effort when
6725 deliver_drop_privilege is set. The new process now just calls the
6726 queue-runner function directly.
6732 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6735 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6736 ignored. Though the use of -oP was forcing the writing of a pid file, it
6737 was always written to the default place.
6739 3. If the message "no IP address found for host xxxx" is generated during
6740 incoming verification, it is now followed by identification of the incoming
6741 connection (so you can more easily find what provoked it).
6743 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6745 5. Added some features to "harden" Exim a bit more against certain attacks:
6747 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6748 be put in Local/Makefile. This is like the never_users runtime option,
6749 but it cannot be overridden. The default setting is "root".
6751 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6752 prefix string with which any file named in a -C command line option
6755 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6756 is retained for -C and -D only if the caller of Exim is root. Without
6757 it, the exim user may also use -C and -D and retain privilege.
6759 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6760 command line option is disabled.
6762 6. Macro names set by the -D option must start with an upper case letter, just
6763 like macro names defined in the configuration file.
6765 7. Added "dereference=" facility to LDAP.
6767 8. Two instances of the typo "uknown" in the source files are fixed.
6769 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6770 the Configure-Makefile script screwed up while processing it.
6772 10. Incorporated PCRE 4.4.
6774 11. The SMTP synchronization check was not operating right at the start of an
6775 SMTP session. For example, it could not catch a HELO sent before the client
6776 waited for the greeting. There is now a check for outstanding input at the
6777 point when the greeting is written. Because of the duplex, asynchronous
6778 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6779 way, but not yet received, when the check is performed.
6781 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6782 on TCP/IP sockets, because this apparently causes some broken clients to
6785 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6786 unchanged) from the Cygwin maintainer.
6788 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6789 was supported instead of testing for mbx. Effectively a typo.
6791 15. The spa authenticator server code was not checking that the input it
6792 received was valid base64.
6794 16. The debug output line for the "set" modifier in ACLs was not showing the
6795 name of the variable that was being set.
6797 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6799 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6800 lookup on its own. Something else was needed to trigger the lookup. For
6801 example, a match in host_lookup or the need for a host name in a host list.
6802 Now, if $sender_host_name is referenced and the host name has not yet been
6803 looked up, a lookup is performed. If the lookup fails, the variable remains
6804 empty, and $host_lookup_failed is set to "1".
6806 19. Added "eqi" as a case-independent comparison operator.
6808 20. The saslauthd authentication condition could segfault if neither service
6809 nor realm was specified.
6811 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6812 error message that was logged was misleading, and incoming SMTP
6813 connections were dropped. The message is now more accurate, and temporary
6814 errors are given to SMTP connections.
6816 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6817 (except RSET) with a 421 error, until QUIT is received. However, it was
6818 failing to send a response to QUIT.
6820 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6821 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6822 ACL is now run after the helo_try_verify_hosts code.
6824 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6825 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6826 case-independent, so other case variants are also recognized). Apparently
6827 some systems use these upper case variants.
6829 25. If more than two messages were waiting for the same host, and a transport
6830 filter was specified for the transport, Exim sent two messages over the
6831 same TCP/IP connection, and then failed with "socket operation on non-
6832 socket" when it tried to send the third.
6834 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6836 27. The extern definition of crypt16() in expand.c was not being excluded when
6837 the OS had its own crypt16() function.
6839 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6840 as a preferred synonym for return_size_limit, both as an option and as an
6843 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6845 30. Changed the default configuration ACL to relax the local part checking rule
6846 for addresses that are not in any local domains. For these addresses,
6847 slashes and pipe symbols are allowed within local parts, but the sequence
6848 /../ is explicitly forbidden.
6850 31. SPA server authentication was not clearing the challenge buffer before
6853 32. log_message in a "warn" ACL statement was writing to the reject log as
6854 well as to the main log, which contradicts the documentation and doesn't
6855 seem right (because no rejection is happening). So I have stopped it.
6857 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6858 However, I am unable to do any testing of this.
6860 34. Fixed an infelicity in the appendfile transport. When checking directories
6861 for a mailbox, to see if any needed to be created, it was accidentally
6862 using path names with one or more superfluous leading slashes; tracing
6863 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6865 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6866 contents are added to the log line that is written for every discarded
6867 recipient. (Previously a log_message setting was ignored.)
6869 36. The ${quote: operator now quotes the string if it is empty.
6871 37. The install script runs exim in order to find its version number. If for
6872 some reason other than non-existence or emptiness, which it checks, it
6873 could not run './exim', it was installing it with an empty version number,
6874 i.e. as "exim-". This error state is now caught, and the installation is
6877 38. An argument was missing from the function that creates an error message
6878 when Exim fails to connect to the socket for saslauthd authentication.
6879 This could cause Exim to crash, or give a corrupted message.
6881 39. Added isip, isip4, and isip6 to ${if conditions.
6883 40. The ACL variables $acl_xx are now saved with the message, and can be
6884 accessed later in routers, transports, and filters.
6886 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6887 strings in the file are not string-expanded.
6889 42. If a MAIL command specified a SIZE value that was too large to fit into an
6890 int variable, the check against message_size_limit failed. Such values are
6891 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6892 day this will have to be increased, but I don't think I want to be around
6893 when emails are that large.
6900 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6901 iconv() is not standard in FreeBSD.
6903 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6904 IPv6 enabled. The observed symptom was a segmentation fault on return from
6905 the function os_common_find_running_interfaces() in src/os.c.
6907 3. In the check_special_case() function in daemon.c I had used "errno" as an
6908 argument name, which causes warnings on some systems. This was basically a
6909 typo, since it was named "eno" in the comments!
6911 4. The code that waits for the clock to tick (at a resolution of some fraction
6912 of a second) so as to ensure message-id uniqueness was always waiting for
6913 at least one whole tick, when it could have waited for less. [This is
6914 almost certainly not relevant at current processor speeds, where it is
6915 unlikely to ever wait at all. But we try to future-proof.]
6917 5. The function that sleeps for a time interval that includes fractions of a
6918 second contained a race. It did not block SIGALRM between setting the
6919 timer, and suspending (a couple of lines later). If the interval was short
6920 and the sigsuspend() was delayed until after it had expired, the suspension
6921 never ended. On busy systems this could lead to processes getting stuck for
6924 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6925 process, before it forks any delivery processes. The open lookup caching
6926 mechanism meant that the open file or database connection was passed into
6927 the delivery process. The problem was that delivery processes always tidy
6928 up cached lookup data. This could cause a problem for the next delivery
6929 process started by the queue runner, because the external queue runner
6930 process does not know about the closure. So the next delivery process
6931 still has data in the lookup cache. In the case of a file lookup, there was
6932 no problem because closing a file descriptor in a subprocess doesn't affect
6933 the parent. However, if the lookup was caching a connection to a database,
6934 the connection was closed, and the second delivery process was likely to
6935 see errors such as "PGSQL: query failed: server closed the connection
6936 unexpectedly". The problem has been fixed by closing all cached lookups
6937 in a queue runner before running a delivery process.
6939 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6940 seem to have the "const" qualifier which it has on other OS. I've
6943 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6944 *of the same type* (client or server) with the same public name that an
6945 error should be diagnosed.
6947 9. When Exim looked up a host name for an IP address, but failed to find the
6948 original IP address when looking up the host name (a safety check), it
6949 output the message "<ip address> does not match any IP for NULL", which was
6950 confusing, to say the least. The bug was that the host name should have
6951 appeared instead of "NULL".
6953 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6954 user that is built into the binary, and the -C or -D options is used, root
6955 privilege is dropped before the configuration file is read. In addition,
6956 logging is switched to stderr instead of the normal log files. If the
6957 configuration then re-defines the Exim user, the unprivileged environment
6958 is probably not what is expected, so Exim logs a panic warning message (but
6961 However, if deliver_drop_privilege is set, the unprivileged state may well
6962 be exactly what is intended, so the warning has been cut out in that case,
6963 and Exim is allowed to try to write to its normal log files.
6969 1. smtp_return_error_details was not giving details for temporary sender
6970 or receiver verification errors.
6972 2. Diagnose a configuration error if two authenticators have the same public
6975 3. Exim used not to create the message log file for a message until the first
6976 delivery attempt. This could be confusing when incoming messages were held
6977 for policy or load reasons. The message log file is now created at the time
6978 the message is received, and an initial "Received" line is written to it.
6980 4. The automatically generated man page for command line options had a minor
6981 bug that caused no ill effects; however, a more serious problem was that
6982 the procedure for building the man page automatically didn't always
6983 operate. Consequently, release 4.20 contains an out-of-date version. This
6984 shouldn't happen again.
6986 5. When building Exim with embedded Perl support, the script that builds the
6987 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6988 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6990 6. The freeze_tell option was not being used for messages that were frozen on
6991 arrival, either by an ACL or by local_scan().
6993 7. Added the smtp_incomplete_transaction log selector.
6995 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6996 was accepting AUTH without a new EHLO.
6998 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6999 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7002 10. Logging of TCP/IP connections (when configured) now happens in the main
7003 daemon process instead of the child process, so that the TCP/IP connection
7004 count is more accurate (but it can never be perfect).
7006 11. The use of "drop" in a nested ACL was not being handled correctly in the
7007 outer ACL. Now, if condition failure induced by the nested "drop" causes
7008 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7009 or "require"), the connection is dropped.
7011 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7012 that yield "discard" can now be used with an "accept" or a "discard" verb,
7013 but an error is generated for any others (because I can't see a useful way
7014 to define what should happen).
7016 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7017 are now processed in the same way as lines in the Exim configuration file.
7018 In particular, continuation lines are supported.
7020 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7022 15. Added -ti meaning -t -i.
7024 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7025 domains, and warn by logging if others are found.
7027 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7028 returned from GIFCONF when Exim is trying to find the list of interfaces on
7029 a host. The code in os.c has been modified to copy each ifreq to an aligned
7030 structure in all cases.
7032 Also, in some cases, the returned ifreq's were being copied to a 'struct
7033 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7034 means the last couple of bytes of an IPv6 address could be chopped if the
7035 ifreq contained only a normal sockaddr (14 bytes storage).
7037 18. Named domain lists were not supported in the hosts_treat_as_local option.
7038 An entry such as +xxxx was not recognized, and was treated as a literal
7041 19. Ensure that header lines added by a DATA ACL are included in the reject log
7042 if the ACL subsequently rejects the message.
7044 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7045 MD5 (which is deprecated).
7047 21. When testing a filter file using -bf, Exim was writing a message when it
7048 took the sender from a "From " line in the message, but it was not doing so
7049 when it took $return_path from a Return-Path: header line. It now does.
7051 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7052 with a valid header line field name (a series of printing characters
7053 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7055 23. Changed "disc" in the source to "disk" to conform to the documentation and
7056 the book and for uniformity.
7058 24. Ignore Sendmail's -Ooption=value command line item.
7060 25. When execve() failed while trying to run a command in a pipe transport,
7061 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7062 could be confused with a return value of 69 from the command itself. This
7063 has been changed to 127, the value the shell returns if it is asked to run
7064 a non-existent command. The wording for the related log line suggests a
7065 non-existent command as the problem.
7067 26. If received_header_text expands to an empty string, do not add a Received:
7068 header line to the message. (Well, it adds a token one on the spool, but
7069 marks it "old" so that it doesn't get used or transmitted.)
7071 27. Installed eximstats 1.28 (addition of -nt option).
7073 28. There was no check for failure on the call to getsockname() in the daemon
7074 code. This can fail if there is a shortage of resources on the system, with
7075 ENOMEM, for example. A temporary error is now given on failure.
7077 29. Contrary to the C standard, it seems that in some environments, the
7078 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7079 program. Exim now does this explicitly; it affects the formatting of
7080 timestamps using strftime().
7082 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7083 complaints. I've now initialized all the variables, to avoid this.
7085 32. Header lines added by a system filter were not being "seen" during
7086 transport-time rewrites.
7088 33. The info_callback() function passed to OpenSSL is set up with type void
7089 (*)(SSL *, int, int), as described somewhere. However, when calling the
7090 function (actually a macro) that sets it up, the type void(*)() is
7091 expected. I've put in a cast to prevent warnings from picky compilers.
7093 34. If a DNS black list lookup found a CNAME record, but there were no A
7094 records associated with the domain it pointed at, Exim crashed.
7096 35. If a DNS black list lookup returned more than one A record, Exim ignored
7097 all but the first. It now scans all returned addresses if a particular IP
7098 value is being sought. In this situation, the contents of the
7099 $dnslist_value variable are a list of all the addresses, separated by a
7102 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7103 to accept a host name and all its aliases if the forward lookup for any of
7104 them yielded the IP address of the incoming connection. Now it accepts only
7105 those names whose forward lookup yields the correct IP address. Any other
7106 names are discarded. This closes a loophole whereby a rogue DNS
7107 administrator could create reverse DNS records to break through a
7108 wildcarded host restriction in an ACL.
7110 37. If a user filter or a system filter that ran in a subprocess used any of
7111 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7112 the wrong values were passed to the pipe command ($thisaddress had the
7113 value of $0, $0 had the value of $1, etc). This bug was introduced by
7114 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7116 38. Improved the line breaking for long SMTP error messages from ACLs.
7117 Previously, if there was no break point between 40 and 75 characters, Exim
7118 left the rest of the message alone. Two changes have been made: (a) I've
7119 reduced the minimum length to 35 characters; (b) if it can't find a break
7120 point between 35 and 75 characters, it looks ahead and uses the first one
7121 that it finds. This may give the occasional overlong line, but at least the
7122 remaining text gets split now.
7124 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7125 file descriptors might be low, and that setting 1000 would always raise it.
7126 It turns out that in some environments, the limit is already over 1000 and
7127 that lowering it causes trouble. So now Exim takes care not to decrease it.
7129 40. When delivering a message, the value of $return_path is set to $sender_
7130 address at the start of routing (routers may change the value). By an
7131 oversight, this default was not being set up when an address was tested by
7132 -bt or -bv, which affected the outcome if any router or filter referred to
7135 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7136 log only when "message" or "log_message" are set. However, if one of the
7137 conditions was an address verification, or a call to a nested ACL, the
7138 messages generated by the underlying test were being passed through. This
7139 no longer happens. The underlying message is available in $acl_verify_
7140 message for both "message" and "log_message" expansions, so it can be
7141 passed through if needed.
7143 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7144 new expansion $bh_ to give the encoded byte string without charset
7145 translation. Translation happens only if iconv() is available; HAVE_ICONV
7146 indicates this at build time. HEADERS_CHARSET gives the charset to
7147 translate to; headers_charset can change it in the configuration, and
7148 "headers charset" can change it in an individual filter file.
7150 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7151 that creates RFC 2047 encoded "words" labels them as that charset instead
7152 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7153 expansion operator; (ii) when Exim creates a From: line for a local
7154 message; (iii) when a header line is rewritten to include a "phrase" part.
7156 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7157 buggy, causing it to skip the first lines of messages whose message ID
7158 ended in 'D'. This would not have bitten before Exim release 4.14, because
7159 message IDs were unlikely to end in 'D' before then. The effect was to have
7160 incorrect size information for certain domains.
7162 45. #include "config.h" was missing at the start of the crypt16.c module. This
7163 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7166 46. If there was a timeout during a "random" callout check, Exim treated it as
7167 a failure of the random address, and carried on sending RSET and the real
7168 address. If the delay was just some slowness somewhere, the response to the
7169 original RCPT would be taken as a response to RSET and so on, causing
7170 mayhem of various kinds.
7172 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7173 when I implemented it. It didn't allow for the fact that some option values
7174 may legitimately be negative (e.g. size_addition), and it didn't even do
7175 the right test for positive values.
7177 48. Domain names in DNS records are case-independent. Exim always looks them up
7178 in lower case. Some resolvers return domain names in exactly the case they
7179 appear in the zone file, that is, they may contain uppercase letters. Not
7180 all resolvers do this - some return always lower case. Exim was treating a
7181 change of case by a resolver as a change of domain, similar to a widening
7182 of a domain abbreviation. This triggered its re-routing code and so it was
7183 trying to route what was effectively the same domain again. This normally
7184 caused routing to fail (because the router wouldn't handle the domain
7185 twice). Now Exim checks for this case specially, and just changes the
7186 casing of the domain that it ultimately uses when it transmits the message
7189 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7192 50. If a filter generated a file delivery with a non-absolute name (possible if
7193 no home directory exists for the router), the forbid_file option was not
7196 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7197 the existing equality matching.
7199 52. Exim was using ints instead of ino_t variables in some places where it was
7200 dealing with inode numbers.
7202 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7203 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7204 called TMPDIR, and if it finds it is different, it changes its value.
7206 54. The smtp_printf() function is now made available to local_scan() so
7207 additional output lines can be written before returning. There is also an
7208 smtp_fflush() function to enable the detection of a dropped connection.
7209 The variables smtp_input and smtp_batched_input are exported to
7212 55. Changed the default runtime configuration: the message "Unknown user"
7213 has been removed from the ACL, and instead placed on the localuser router,
7214 using the cannot_route_message feature. This means that any verification
7215 failures that generate their own messages won't get overridden. Similarly,
7216 the "Unrouteable address" message that was in the ACL for unverifiable
7217 relay addresses has also been removed.
7219 56. Added hosts_avoid_esmtp to the smtp transport.
7221 57. The exicyclog script was not checking for the esoteric option
7222 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7223 will work only if exicyclog is run under the appropriate euid.
7225 58. Following a discussion on the list, the rules by which Exim recognises line
7226 endings on incoming messages have been changed. The -dropcr and drop_cr
7227 options are now no-ops, retained only for backwards compatibility. The
7228 following line terminators are recognized: LF CRLF CR. However, special
7229 processing applies to CR:
7231 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7232 nor a local message in the state where . is a terminator.
7234 (ii) If a bare CR is encountered in a header line, an extra space is added
7235 after the line terminator so as not to end the header. The reasoning
7236 behind this is that bare CRs in header lines are most likely either
7237 to be mistakes, or people trying to play silly games.
7239 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7240 was being incorrectly given as 18 bytes larger than it should have been.
7241 This is a VOB (very old bug).
7243 60. This may never have affected anything current, but just in case it has:
7244 When the local host is found other than at the start of a list of hosts,
7245 the local host, those with the same MX, and any that follow, are discarded.
7246 When the list in question was part of a longer list of hosts, the following
7247 hosts (not currently being processed) were also being discarded. This no
7248 longer happens. I'm not sure if this situation could ever has previously
7251 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7254 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7255 header. This is something that is recommended in a new Internet Draft, and
7256 is something that is documented as being done by Sendmail. There are two
7257 possible values. For messages generated by the autoreply transport, Exim
7260 Auto-Submitted: auto-replied
7262 whereas for all other generated messages (e.g. bounces) it adds
7264 Auto-Submitted: auto-generated
7266 63. The "personal" condition in filters now includes a test for the
7267 Auto-Submitted: header. If it contains the string "auto-" the message it
7268 not considered personal.
7270 64. Added rcpt_include_affixes as a generic transport option.
7272 65. Added queue_only_override (default true).
7274 66. Added the syslog_duplication option.
7276 67. If what should have been the first header line of a message consisted of
7277 a space followed by a colon, Exim was mis-interpreting it as a header line.
7278 It isn't of course - it is syntactically invalid and should therefore be
7279 treated as the start of the message body. The misbehaviour could have
7280 caused a number of strange effects, including loss of data in subsequent
7281 header lines, and spool format errors.
7283 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7284 client host had authenticated. This control can now be exercised by an ACL
7285 for more flexibility.
7287 69. By default, callouts do not happen when testing with -bh. There is now a
7288 variant, -bhc, which does actually run the callout code, including
7289 consulting and updating the callout cache.
7291 70. Added support for saslauthd authentication, courtesy of Alexander
7294 71. If statvfs() failed on the spool or log directories while checking their
7295 size for availability, Exim confusingly gave the error "space shortage".
7296 Furthermore, in debugging mode it crashed with a floating point exception.
7297 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7298 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7299 serious problem, Exim now writes to the main and panic logs when this
7300 happens, with details of the failure. It then refuses to accept the
7301 incoming message, giving the message "spool directory problem" or "log
7302 directory problem" with a 421 code for SMTP messages.
7304 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7305 0, 1, and 2 exist, because some OS complain for execs without them (see
7306 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7307 descriptors. However, the code omitted to check that the open succeeded,
7308 causing mysterious errors if for some reason the permissions on /dev/null
7309 got screwed. Now Exim writes a message to the main and panic logs, and
7310 bombs out if it can't open /dev/null.
7312 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7313 interact so that it is all more flexible. It is supposed to remain
7314 backwards compatible. Also added extra_local_interfaces.
7316 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7317 to bomb out with an assertion failure - to the client this appears as a
7318 connection drop. This problem occurs in the part of the code that was taken
7319 from the Samba project. Fortunately, the assertion is in a very simple
7320 function, so I have fixed this by reproducing the function inline in the
7321 one place where it is called, and arranging for authentication to fail
7322 instead of killing the process with assert().
7324 75. The SPA client code was not working when the server requested OEM rather
7325 than Unicode encoding.
7327 76. Added code to make require_files with a specific uid setting more usable in
7328 the case where statting the file as root fails - usually a non-root-mounted
7329 NFS file system. When this happens and the failure is EACCES, Exim now
7330 forks a subprocess and does the per-uid checking as the relevant uid.
7332 77. Added process_log_path.
7334 78. If log_file_path was not explicitly set, a setting of check_log_space or
7335 check_log_inodes was ignored.
7337 79. If a space check for the spool or log partitions fails, the incident is now
7338 logged. Of course, in the latter case the data may get lost...
7340 80. Added the %p formatting code to string_format() so that it can be used to
7341 print addresses in debug_print(). Adjusted all the address printing in the
7342 debugging in store.c to use %p rather than %d.
7344 81. There was a concern that a line of code in smtp_in.c could overflow a
7345 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7346 initially expressed, the concern was not well-founded, because trailing
7347 spaces are removed early. However, if the trailing spaces were followed by
7348 a NULL, they did not get removed, so the overflow was possible. Two fixes
7351 (a) I re-wrote the offending code in a cleaner fashion.
7352 (b) If an incoming SMTP command contains a NULL character, it is rejected
7355 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7356 runs initgroups(), so that if the Exim user is in any additional groups,
7357 they will be used during message reception.
7363 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob