1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
12 side, to decode a TLS 1.3 packet capture.
14 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
15 Previously the default library behaviour applied, sending two, each in
18 JH/03 Debug output for ACL now gives the config file name and line number for
21 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
23 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
25 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
26 buffer overrun for (non-chunking) other transports.
28 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
29 TLS1.3, means that a server rejecting a client certificate is not visible
30 to the client until the first read of encrypted data (typically the
31 response to EHLO). Add detection for that case and treat it as a failed
32 TLS connection attempt, so that the normal retry-in-clear can work (if
35 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
36 and/or domain. Found and fixed by Jason Betts.
38 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
39 configuration). If a CNAME target was not a wellformed name pattern, a
42 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
43 the OS reports them interleaved with other addresses.
45 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
46 used both for input and for a verify callout, both encrypted, SMTP
47 responses being sent by the server could be lost. This resulted in
48 dropped connections and sometimes bounces generated by a peer sending
51 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
52 string. Previously it was possible to cause undefined behaviour in a
53 library routine (usually a crash). Found by "zerons".
55 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
58 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
59 API was removed, so update to use the newer ones.
61 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
62 any timeout set, is taking a long time. Previously we would hang on to a
63 rotated logfile "forever" if the input was arriving with long gaps
64 (a previous attempt to fix addressed lack, for a long time, of initial
67 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
68 shared (NFS) environment. The length of the tempfile name is now
69 4 + 16 ("hdr.$message_exim_id") which might break on file
70 systems which restrict the file name length to lower values.
73 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
74 shared (NFS) environment.
76 HS/02 Bug 2392: exigrep does case sensitive *option* processing (as it
77 did for all versions <4.90). Notably -M, -m, --invert, -I may be
80 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
81 on some platforms for bit 31.
83 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
84 to changes apparently associated with TLS1.3 handling some of the APIs
85 previously used were either nonfunctional or inappropriate. Strings
86 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
87 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
88 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
89 This affects log line X= elements, the $tls_{in,out}_cipher variables,
90 and the use of specific cipher names in the encrypted= ACL condition.
92 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
94 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
95 verification result was not updated unless hosts_require_ocsp applied.
97 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
98 queue_list_requires_admin set to false, non-admin users were denied the
101 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
102 directory-of-certs mode. Previously they were advertised despite the
109 JH/01 Remove code calling the customisable local_scan function, unless a new
110 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
112 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
113 non-signal-safe functions being used.
115 JH/03 Bug 2269: When presented with a received message having a stupidly large
116 number of DKIM-Signature headers, disable DKIM verification to avoid
117 a resource-consumption attack. The limit is set at twenty.
119 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
120 report of oldest_pass in ${authres } in consequence, and separate out
121 some descriptions of reasons for verification fail.
123 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
124 files in the spool were present and unlocked. A queue-runner could spot
125 them, resulting in a duplicate delivery. Fix that by doing the unlock
126 after the unlink. Investigation by Tim Stewart. Take the opportunity to
127 add more error-checking on spoolfile handling while that code is being
130 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
131 No known attacks, no CVE, this is defensive hardening.
133 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
134 a queue-runner could start a delivery while other operations were ongoing.
135 Cutthrough delivery was a common victim, resulting in duplicate delivery.
136 Found and investigated by Tim Stewart. Fix by using the open message data
137 file handle rather than opening another, and not locally closing it (which
138 releases a lock) for that case, while creating the temporary .eml format
139 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
141 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
142 $sender_verify_failure/$recipient_verify_failure to "random".
144 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
147 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
148 Previously this would segfault.
150 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
153 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
154 like zero, since the resolver should be doing this for us, But we need one
155 as a CNAME but no MX presence gets the CNAME returned; we need to check
156 that doesn't point to an MX to declare it "no MX returned" rather than
157 "error, loop". A new main option is added so the older capability of
158 following some limited number of chain links is maintained.
160 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
162 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
163 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
164 now-deprecated earlier definitions used only specified the range up to TLS
165 1.2 (in the older-version library docs).
167 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
169 JH/15 Rework TLS client-side context management. Stop using a global, and
170 explicitly pass a context around. This enables future use of TLS for
171 connections to service-daemons (eg. malware scanning) while a client smtp
172 connection is using TLS; with cutthrough connections this is quite likely.
174 JH/16 Fix ARC verification to do AS checks in reverse order.
176 JH/17 Support a "tls" option on the ${readsocket } expansion item.
178 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
179 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
180 Previously the "utf8" would be re-prepended for every additional message.
182 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
183 Previously thery were accepted, resulting in issues when attempting to
184 forward messages to a non-supporting MTA.
186 PP/02 Let -n work with printing macros too, not just options.
188 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
189 one parent address was copied, and bogus data was used at delivery-logging
190 time. Either a crash (after delivery) or bogus log data could result.
191 Discovery and analysis by Tim Stewart.
193 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
194 Previously if the string ended mid-character, we did not insert the
195 promised '?' replacement.
197 PP/04 Documentation: current string operators work on bytes, not codepoints.
199 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
200 should pack well giving a smaller memory footprint so better caching and
201 therefore performance. Group the declarations where this can't be done so
202 that the byte-sized flag variables are not interspersed among pointer
203 variables, giving a better chance of good packing by the compiler.
205 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
206 non-null, to avoid issues with sites running BATV. Previously reports were
207 sent with an empty envelope sender so looked like bounces.
209 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
210 The ignore_error flag wasn't being returned from the filter subprocess so
211 was not set for later routers. Investigation and fix by Matthias Kurz.
213 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
214 and a msg:complete for the whole, when a message is manually removed using
215 -Mrm. Developement by Matthias Kurz, hacked on by JH.
217 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
218 a "Gnu special" function, asprintf() in the DB utility binary builds; I
219 hope that is portable enough.
221 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
222 requiring a known-CA anchor certificate; make it now rely entirely on the
223 TLSA as an anchor. Checking the name on the leaf cert against the name
224 on the A-record for the host is still done for TA (but not for EE mode).
226 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
227 included in delivery lines for non-proxied connections, when compiled with
228 SUPPORT_SOCKS and running with proxy logging enabled.
230 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
231 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
232 move the existing event to fire before the normal logging of message
233 failure so that custom logging is bracketed by normal logging.
235 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
236 msg:fail:internal event. Developement by Matthias Kurz.
238 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
239 far too small for todays use of crypto signatures stored there. Go all
240 the way to the max DNS message size of 64kB, even though this might be
241 overmuch for IOT constrained device use.
243 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
244 copy a string over itself. The library routine is documented as not
245 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
247 JH/32 For main options check_spool_space and check_inode_space, where the
248 platform supports 64b integers, support more than the previous 2^31 kB
249 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
250 the previous G, M, k.
252 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
253 $authenticated_fail_id variable on authentication failure. Previously
256 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
257 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
258 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
259 GNUTLS_SEC_PARAM_MEDIUM.
261 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
262 side. Previously we would continue as if no SNI had been received.
264 JH/36 Harden the handling of string-lists. When a list consisted of a sole
265 "<" character, which should be a list-separator specification, we walked
266 off past the nul-terimation.
268 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
269 causes) even when the retry time is not yet met. Previously they were
270 not, meaning that when (say) an account was over-quota and temp-rejecting,
271 and multiple senders' messages were queued, only one sender would get
272 notified on each configured delay_warning cycle.
274 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
276 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
277 cipher-suites, an error can be left on the stack even for a succeeding
278 accept; this results in impossible error messages when a later operation
281 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
282 return error codes indicating retry. Under TLS1.3 this becomes required.
284 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
285 it only wrote the new authenticators, resulting in a lack of tracking of
286 peer changes of ESMTP extensions until the next cache flush.
288 JH/41 Fix the loop reading a message header line to check for integer overflow,
289 and more-often against header_maxsize. Previously a crafted message could
290 induce a crash of the recive process; now the message is cleanly rejected.
292 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
293 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
299 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
300 When redis_servers is set to a list of > 1 element, and the Redis servers
301 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
302 case of MOVED into a DEFER case instead, thus moving the query onto the
303 next server in the list. For a cluster of N elements, all N servers must
304 be defined in redis_servers.
306 GF/02 Catch and remove uninitialized value warning in exiqsumm
307 Check for existence of @ARGV before looking at $ARGV[0]
309 JH/01 Replace the store_release() internal interface with store_newblock(),
310 which internalises the check required to safely use the old one, plus
311 the allocate and data copy operations duplicated in both (!) of the
312 extant use locations.
314 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
315 modifier. This matches the restriction on the commandline.
317 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
318 Previously only the last row was returned.
320 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
321 we assumed that tags in the header were well-formed, and parsed the
322 element content after inspecting only the first char of the tag.
323 Assumptions at that stage could crash the receive process on malformed
326 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
327 While running the DKIM ACL we operate on the Permanent memory pool so that
328 variables created with "set" persist to the DATA ACL. Also (at any time)
329 DNS lookups that fail create cache records using the Permanent pool. But
330 expansions release any allocations made on the current pool - so a dnsdb
331 lookup expansion done in the DKIM ACL releases the memory used for the
332 DNS negative-cache, and bad things result. Solution is to switch to the
333 Main pool for expansions.
334 While we're in that code, add checks on the DNS cache during store_reset,
335 active in the testsuite.
336 Problem spotted, and debugging aided, by Wolfgang Breyha.
338 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
339 When none of the hosts presented to a transport match an already-open
340 connection, close it and proceed with the list. Previously we would
341 queue the message. Spotted by Lena with Yahoo, probably involving
344 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
345 Previously a spurious "250 OK id=" response was appended to the proper
348 JH/08 The "support for" informational output now, which built with Content
349 Scanning support, has a line for the malware scanner interfaces compiled
350 in. Interface can be individually included or not at build time.
352 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
353 by the template makefile "src/EDITME". The "STREAM" support for an older
354 ClamAV interface method is removed.
356 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
357 rows affected is given instead).
359 JH/11 The runtime Berkeley DB library version is now additionally output by
360 "exim -d -bV". Previously only the compile-time version was shown.
362 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
363 SMTP connection. Previously, when one had more receipients than the
364 first, an abortive onward connection was made. Move to full support for
365 multiple onward connections in sequence, handling cutthrough connection
366 for all multi-message initiating connections.
368 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
369 routers. Previously, a multi-recipient message would fail to match the
370 onward-connection opened for the first recipient, and cause its closure.
372 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
373 a timeout on read on a GnuTLS initiating connection, resulting in the
374 initiating connection being dropped. This mattered most when the callout
375 was marked defer_ok. Fix to keep the two timeout-detection methods
378 JH/15 Relax results from ACL control request to enable cutthrough, in
379 unsupported situations, from error to silently (except under debug)
380 ignoring. This covers use with PRDR, frozen messages, queue-only and
383 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
385 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
386 metadata, resulting in a crash in free().
388 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
389 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
390 Broken also in d185889f4, with init system revamp.
392 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
393 Previously we abruptly closed the connection after reading a malware-
394 found indication; now we go on to read the "scan ok" response line,
397 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
398 ACL. Previously, a crash would result.
400 JH/19 Speed up macro lookups during configuration file read, by skipping non-
401 macro text after a replacement (previously it was only once per line) and
402 by skipping builtin macros when searching for an uppercase lead character.
404 JH/20 DANE support moved from Experimental to mainline. The Makefile control
405 for the build is renamed.
407 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
408 was allocated for every new TLS startup, meaning one per message. Fix
409 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
411 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
412 reported the original. Fix to report (as far as possible) the ACL
413 result replacing the original.
415 JH/23 Fix memory leak during multi-message connections using STARTTLS under
416 OpenSSL. Certificate information is loaded for every new TLS startup,
417 and the resources needed to be freed.
419 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
421 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
424 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
425 DATA response info to the (existing) per-recipient response info for
426 the "C=" log element. It can have useful tracking info from the
427 destination system. Patch from Simon Arlott.
429 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
430 length value. Previously this would segfault.
432 HS/02 Support Avast multiline protoocol, this allows passing flags to
433 newer versions of the scanner.
435 JH/28 Ensure that variables possibly set during message acceptance are marked
436 dead before release of memory in the daemon loop. This stops complaints
437 about them when the debug_store option is enabled. Discovered specifically
438 for sender_rate_period, but applies to a whole set of variables.
439 Do the same for the queue-runner and queue-list loops, for variables set
440 from spool message files. Do the same for the SMTP per-message loop, for
441 certain variables indirectly set in ACL operations.
443 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
444 as a multi-recipient message from a mailinglist manager). The coding had
445 an arbitrary cutoff number of characters while checking for more input;
446 enforced by writing a NUL into the buffer. This corrupted long / fast
447 input. The problem was exposed more widely when more pipelineing of SMTP
448 responses was introduced, and one Exim system was feeding another.
449 The symptom is log complaints of SMTP syntax error (NUL chars) on the
450 receiving system, and refused recipients seen by the sending system
451 (propating to people being dropped from mailing lists).
452 Discovered and pinpointed by David Carter.
454 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
455 replaced by the ${authresults } expansion.
457 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
459 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
460 allows proper process termination in container environments.
462 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
463 Previously the "final dot" had a newline after it; ensure it is CR,LF.
465 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
466 and "err_perm", deprecated since 4.83 when the RFC-defined words
467 "temperror" and "permerror" were introduced.
469 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
470 transport-filters or DKIM-signing. The restriction was lost in the
471 consolidation of verify-callout and delivery SMTP handling.
472 Extend the restriction to also cover ARC-signing.
474 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
475 in defer=pass mode supply a 450 to the initiator. Previously the message
478 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
479 tls_require_ciphers is used as before.
481 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
482 "pass_unscanned". Only tmpfails from the scanner are written to
483 the paniclog, as they may require admin intervention (permission
484 denied, license issues). Other scanner errors (like decompression
485 bombs) do not cause a paniclog entry.
487 JH/36 Fix reinitialisation of DKIM logging variable between messages.
488 Previously it was possible to log spurious information in receive log
491 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
492 triggered odd behaviour from Outlook Express clients.
494 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
497 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
498 since the IETF WG has not yet settled on that versus the original
499 "bare" representation.
501 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
502 Previously the millisecond value corrupted the output.
503 Fix also for syslog_pid=no and log_selector +pid, for which the pid
504 corrupted the output.
510 JH/01 Rework error string handling in TLS interface so that the caller in
511 more cases is responsible for logging. This permits library-sourced
512 string to be attached to addresses during delivery, and collapses
513 pairs of long lines into single ones.
515 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
516 during configuration. Wildcards are allowed and expanded.
518 JH/02 Rework error string handling in DKIM to pass more info back to callers.
519 This permits better logging.
521 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
522 do not close it down and have the child transport start it up again on
523 the passed-on TCP connection. Instead, proxy the child (and any
524 subsequent ones) for TLS via a unix-domain socket channel. Logging is
525 affected: the continued delivery log lines do not have any DNSSEC, TLS
526 Certificate or OCSP information. TLS cipher information is still logged.
528 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
529 identical IP addresses on different listening ports. Will also affect
532 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
533 add noisy ifdef guards to special-case this sillyness.
534 Patch from Bernd Kuhls.
536 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
537 than 255 are no longer allowed.
539 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
540 Disable the session-cache too, which might reduce our load. Since we
541 currrectly use a new context for every connection, both as server and
542 client, there is no benefit for these.
543 GnuTLS appears to not support tickets server-side by default (we don't
544 call gnutls_session_ticket_enable_server()) but client side is enabled
545 by default on recent versions (3.1.3 +) unless the PFS priority string
548 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
549 <https://reproducible-builds.org/specs/source-date-epoch/>.
551 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
552 the check for any unsuccessful recipients did not notice the limit, and
553 erroneously found still-pending ones.
555 JH/08 Pipeline CHUNKING command and data together, on kernels that support
556 MSG_MORE. Only in-clear (not on TLS connections).
558 JH/09 Avoid using a temporary file during transport using dkim. Unless a
559 transport-filter is involved we can buffer the headers in memory for
560 creating the signature, and read the spool data file once for the
561 signature and again for transmission.
563 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
564 4.77 as the kernel support then wasn't solid, having issues in 64bit
565 mode. Now, it's been long enough. Add support for FreeBSD also.
567 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
568 case where the routing stage had gathered several addresses to send to
569 a host before calling the transport for the first, we previously failed
570 to close down TLS in the old transport process before passing the TCP
571 connection to the new process. The new one sent a STARTTLS command
572 which naturally failed, giving a failed delivery and bloating the retry
573 database. Investigation and fix prototype from Wolfgang Breyha.
575 JH/12 Fix check on SMTP command input synchronisation. Previously there were
576 false-negatives in the check that the sender had not preempted a response
577 or prompt from Exim (running as a server), due to that code's lack of
578 awareness of the SMTP input buffering.
580 PP/04 Add commandline_checks_require_admin option.
581 Exim drops privileges sanely, various checks such as -be aren't a
582 security problem, as long as you trust local users with access to their
583 own account. When invoked by services which pass untrusted data to
584 Exim, this might be an issue. Set this option in main configuration
585 AND make fixes to the calling application, such as using `--` to stop
588 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
589 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
590 responses to those, into a single TLS record each way (this usually means
591 a single packet). As a side issue, smtp_enforce_sync now works on TLS
594 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
595 affects you only if you're dancing at the edge of the param size limits.
596 If you are, and this message makes sense to you, then: raise the
597 configured limit or use OpenSSL 1.1. Nothing we can do for older
600 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
601 cmdline element to get the documented default one. Previously it was
604 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
605 are unsuable for non-message-specific reasons - eg. connection timeout,
606 banner-time rejection.
608 JH/16 Fix logging of delivery remote port, when specified by router, under
611 PP/06 Repair manualroute's ability to take options in any order, even if one
612 is the name of a transport.
615 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
617 JH/17 Change the list-building routines interface to use the expanding-string
618 triplet model, for better allocation and copying behaviour.
620 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
621 Previously it was constructed the first time a possibly-matching string
622 was met in the configuration file input during startup; now it is done
625 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
626 compatible one, to avoid the (poorly documented) possibility of a config
627 file in the working directory redirecting the DB files, possibly correpting
628 some existing file. CVE-2017-10140 assigned for BDB.
630 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
631 cache-hot. Previously, although the result was properly cached, the
632 initial verify call returned a defer.
634 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
635 the main verify for receipient in uncached-mode.
637 JH/22 Retire historical build files to an "unsupported" subdir. These are
638 defined as "ones for which we have no current evidence of testing".
640 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
641 if present. Previously it was ignored.
643 JH/24 Start using specified-initialisers in C structure init coding. This is
644 a C99 feature (it's 2017, so now considered safe).
646 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
647 if was a fixed-sized field and bitmask ops via macros; it is now more
650 PP/07 GitHub PR 56: Apply MariaDB build fix.
651 Patch provided by Jaroslav Škarvada.
653 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
654 during Coverity cleanups [4.87 JH/47]
655 Diagnosis and fix provided by Michael Fischer v. Mollard.
657 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
658 the right size to place the terminating semicolon on its own folded
659 line, the header hash was calculated to an incorrect value thanks to
660 the (relaxed) space the fold became.
662 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
663 and confused the parent.
665 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
666 which could crash as a result. This could lead to undeliverable messages.
668 JH/28 Logging: "next input sent too soon" now shows where input was truncated
671 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
672 matters on fast-turnover and PID-randomising systems, which were getting
673 out-of-order delivery.
675 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
676 a possibly-overlapping copy. The symptom was that "Remote host closed
677 connection in response to HELO" was logged instead of the actual 4xx
680 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
681 Previously only that bufferd was discarded, resulting in SYMTP command
684 JH/32 DKIM: when a message has multiple signatures matching an identity given
685 in dkim_verify_signers, run the dkim acl once for each. Previously only
686 one run was done. Bug 2189.
688 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
689 "panic the current process" to "deliberately defer". The panic log is
690 still written with the problem list name; the mail and reject logs now
691 get a temp-reject line for the message that was being handled, saying
692 something like "domains check lookup or other defer". The SMTP 451
693 message is still "Temporary local problem".
695 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
696 A crafted sequence of BDAT commands could result in in-use memory beeing
697 freed. CVE-2017-16943.
699 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
700 from SMTP input. Previously it was always done; now only done for DATA
701 and not BDAT commands. CVE-2017-16944.
703 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
704 to the message (such as an overlong header line). Previously this was
705 not done and we did not exit BDAT mode. Followon from the previous item
706 though a different problem.
712 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
713 than -2003 did; needs libidn2 in addition to libidn.
715 JH/02 The path option on a pipe transport is now expanded before use.
717 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
718 Patch provided by "Björn", documentation fix added too.
720 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
721 missing a wire-to-host endian conversion.
723 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
724 close after a BDAT command line could be taken as a following command,
725 giving a synch failure. Fix by only checking for synch immediately
726 before acknowledging the chunk.
728 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
729 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
730 macro. Patches provided by Josh Soref.
732 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
733 Previously we did not; the RFC seems ambiguous and VRFY is not listed
734 by IANA as a service extension. However, John Klensin suggests that we
737 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
738 the dkim code may be unix-mode line endings rather than smtp wire-format
739 CRLF, so prepend a CR to any bare LF.
741 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
742 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
744 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
745 the first were themselves being wrongly included in the feed into dkim
746 processing; with most chunk sizes in use this resulted in an incorrect
747 body hash calculated value.
749 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
750 DKIM signature block, for verification. Although advised against by
751 standards it is specifically not ruled illegal.
753 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
755 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
756 missing a body hash (the bh= tag).
758 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
759 It seems that HAProxy sends the Proxy Protocol information in clear and
760 only then does a TLS startup, so do the same.
762 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
763 TCP connections (such as for Spamd) unless the daemon successfully set
764 Fast Open mode on its listening sockets. This fixes breakage seen on
765 too-old kernels or those not configured for Fast Open, at the cost of
766 requiring both directions being enabled for TFO, and TFO never being used
767 by non-daemon-related Exim processes.
769 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
770 endings, at least on the first header line. Try to canonify any that get
771 past that check, despite the cost.
773 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
774 now limited to an arbitrary five deep, while parsing addresses with the
775 strip_excess_angle_brackets option enabled.
777 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
778 instead leave the unprompted TLS handshake in socket buffer for the
779 TLS library to consume.
781 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
783 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
785 JH/16 Drop variables when they go out of scope. Memory management drops a whole
786 region in one operation, for speed, and this leaves assigned pointers
787 dangling. Add checks run only under the testsuite which checks all
788 variables at a store-reset and panics on a dangling pointer; add code
789 explicitly nulling out all the variables discovered. Fixes one known
790 bug: a transport crash, where a dangling pointer for $sending_ip_address
791 originally assigned in a verify callout, is re-used.
793 PP/06 Drop '.' from @INC in various Perl scripts.
795 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
797 PP/08 Reduce a number of compilation warnings under clang; building with
798 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
799 should be warning-free.
801 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
803 HS/01 Fix portability problems introduced by PP/08 for platforms where
804 realloc(NULL) is not equivalent to malloc() [SunOS et al].
806 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
807 chunk. This allows us to accept broken chunked messages. We need a more
808 general solution here.
810 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
811 already-broken messages in the queue.
813 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
815 JH/19 Fix reference counting bug in routing-generated-address tracking.
821 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
822 supports it and a size is available (ie. the sending peer gave us one).
824 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
825 years of being deprecated). The replacements are the ACLs
826 acl_smtp_mime and acl_not_smtp_mime.
828 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
829 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
830 or even an in-clear connection were permitted. Now, if the host lookup
831 was dnssec and dane was requested then the host is only used if the
832 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
833 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
834 if one fails this test.
835 This means that a poorly-configured remote DNS will make it incommunicado;
836 but it protects against a DNS-interception attack on it.
838 JH/04 Bug 1810: make continued-use of an open smtp transport connection
839 non-noisy when a race steals the message being considered.
841 JH/05 If main configuration option tls_certificate is unset, generate a
842 self-signed certificate for inbound TLS connections.
844 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
845 in rewrites and routers.
847 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
848 and logged a warning sing 4.83; now they are a configuration file error.
850 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
851 (lacking @domain). Apply the same qualification processing as RCPT.
853 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
855 JH/10 Support ${sha256:} applied to a string (as well as the previous
858 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
859 a cutthrough deliver is pending, as we always want to make a connection.
860 This also avoids re-routing the message when later placing the cutthrough
861 connection after a verify cache hit.
862 Do not update it with the verify result either.
864 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
865 when routing results in more than one destination address.
867 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
868 signing (which inhibits the cutthrough capability). Previously only
869 the presence of an option was tested; now an expansion evaluating as
870 empty is permissible (obviously it should depend only on data available
871 when the cutthrough connection is made).
873 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
874 the relevant preceding SMTP command did not note the pipelining mode.
876 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
877 Previously they were not counted.
879 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
880 as one having no matching records. Previously we deferred the message
881 that needed the lookup.
883 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
884 distinguished as "(=".
886 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
887 for missing MX records. Previously it only worked for missing A records.
889 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
891 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
892 after the data-go-ahead and data-ack. Patch from Jason Betts.
894 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
895 even for a "none" policy. Patch from Tony Meyer.
897 JH/22 Fix continued use of a connection for further deliveries. If a port was
898 specified by a router, it must also match for the delivery to be
901 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
902 When one of the recipients of a message was unsuitable for the connection
903 (has no matching addresses), we lost track of needing to mark it
904 deferred. As a result mail would be lost.
906 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
908 JH/25 Decoding ACL controls is now done using a binary search; the source code
909 takes up less space and should be simpler to maintain. Merge the ACL
910 condition decode tables also, with similar effect.
912 JH/26 Fix problem with one_time used on a redirect router which returned the
913 parent address unchanged. A retry would see the parent address marked as
914 delivered, so not attempt the (identical) child. As a result mail would
917 JH/27 Fix a possible security hole, wherein a process operating with the Exim
918 UID can gain a root shell. Credit to http://www.halfdog.net/ for
919 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
922 JH/28 Enable {spool,log} filesystem space and inode checks as default.
923 Main config options check_{log,spool}_{inodes,space} are now
924 100 inodes, 10MB unless set otherwise in the configuration.
926 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
927 Previously it only applied to the main-section connection policy
930 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
932 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
933 by me. Added RFC7919 DH primes as an alternative.
935 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
936 are not in the system include path.
938 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
939 GnuTLS, when a session startup failed (eg because the client disconnected)
940 Exim did stdio operations after fclose. This was exposed by a recent
941 change which nulled out the file handle after the fclose.
943 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
944 signed directly by the cert-signing cert, rather than an intermediate
945 OCSP-signing cert. This is the model used by LetsEncrypt.
947 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
949 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
950 an incoming connection.
952 HS/02 Bug 1802: Do not half-close the connection after sending a request
955 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
956 fallback to "prime256v1".
958 JH/34 SECURITY: Use proper copy of DATA command in error message.
959 Could leak key material. Remotely exploitable. CVE-2016-9963.
965 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
966 and 3.4.4 - once the server is enabled to respond to an OCSP request
967 it does even when not requested, resulting in a stapling non-aware
968 client dropping the TLS connection.
970 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
971 support variable-length bit vectors. No functional change.
973 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
974 The I= interface field on outgoing lines is now after the H= remote
975 host field, same as incoming lines. There is a separate
976 outgoing_interface log selector which allows you to disable the
979 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
980 If not running log_selector +smtp_connection the mainlog would be held
981 open indefinitely after a "too many connections" event, including to a
982 deleted file after a log rotate. Leave the per net connection logging
983 leaving it open for efficiency as that will be quickly detected by the
984 check on the next write.
986 HS/01 Bug 1671: Fix post transport crash.
987 Processing the wait-<transport> messages could crash the delivery
988 process if the message IDs didn't exist for some reason. When
989 using 'split_spool_directory=yes' the construction of the spool
990 file name failed already, exposing the same netto behaviour.
992 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
993 mime_regex ACL conditions.
995 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
996 to DSN fail messages (bounces): remote IP, remote greeting, remote response
997 to HELO, local diagnostic string.
999 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1000 log line to debug. Even when configured with a tls authenticator many
1001 client connections are expected to not authenticate in this way, so
1002 an authenticate fail is not an error.
1004 HS/02 Add the Exim version string to the process info. This way exiwhat
1005 gives some more detail about the running daemon.
1007 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1008 matter for fast-change records such as DNSBLs.
1010 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1011 retry record, even if constant. There may be multiple transports with
1012 different interface settings and the retry behaviour needs to be kept
1015 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1017 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1019 JH/10 Bug 840: fix log_defer_output option of pipe transport
1021 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1022 hosts_randomize. This matters a lot when combined with mua_wrapper.
1024 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1025 ${quote_pgsql:<string>} operator.
1027 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1029 JH/14 Change header file name for freeradius-client. Relevant if compiling
1030 with Radius support; from the Gentoo tree and checked under Fedora.
1032 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1034 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1035 option transport_filter, meaning no filtering.
1037 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1039 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1040 defaults to "*" (all hosts). The variable is now available when not built
1041 with TLS, default unset, mainly to enable keeping the testsuite sane.
1042 If a server certificate is not supplied (via tls_certificate) an error is
1043 logged, and clients will find TLS connections fail on startup. Presumably
1044 they will retry in-clear.
1045 Packagers of Exim are strongly encouraged to create a server certificate
1046 at installation time.
1048 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1049 with the $config_file variable.
1051 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1052 in transport context, after the attempt, and per-recipient. The latter type
1053 is per host attempted. The event data is the error message, and the errno
1054 information encodes the lookup type (A vs. MX) used for the (first) host,
1055 and the trailing two digits of the smtp 4xx response.
1057 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1058 to write to mainlog (or rejectlog, paniclog) in the window between file
1059 creation and permissions/ownership being changed. Particularly affects
1060 installations where exicyclog is run as root, rather than exim user;
1061 result is that the running daemon panics and dies.
1063 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1065 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1066 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1067 "pri" and "weight". Note that the previous implicit priority given by the
1068 list order is no longer honoured.
1070 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1071 for DKIM processing.
1073 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1074 by defining SUPPORT_SOCKS.
1076 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1077 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1078 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1079 variables are renamed to proxy_{local,external}_{address,port}.
1081 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1082 for a build by defining SUPPORT_I18N
1084 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1085 of the query string, and make ${quote_redis:} do that quoting.
1087 JH/29 Move Events support from Experimental to mainline, enabled by default
1088 and removable for a build by defining DISABLE_EVENT.
1090 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1092 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1093 cached by the daemon.
1095 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1096 by defining LOOKUP_REDIS. The libhiredis library is required.
1098 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1099 keys are given for lookup.
1101 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1102 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1103 only supported when built with TLS support. The PolarSSL SHA routines
1104 are still used when the TLS library is too old for convenient support.
1106 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1107 openssl_options), for security. OpenSSL forces this from version 1.1.0
1108 server-side so match that on older versions.
1110 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1111 allocation for $value could be released as the expansion processing
1112 concluded, but leaving the global pointer active for it.
1114 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1115 and to use the domains and local_parts ACL conditions.
1117 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1118 incorrectly not doubled on cutthrough transmission, hence seen as a
1119 body-termination at the receiving system - resulting in truncated mails.
1120 Commonly the sender saw a TCP-level error, and retransmitted the message
1121 via the normal store-and-forward channel. This could result in duplicates
1122 received - but deduplicating mailstores were liable to retain only the
1123 initial truncated version.
1125 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1127 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1129 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1130 we're in there, support oversigning also; bug 1309.
1132 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1134 HS/04 Add support for keep_environment and add_environment options.
1136 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1137 either intentional arithmetic overflow during PRNG, or testing config-
1140 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1141 delivery resulted in actual delivery. Cancel cutthrough before DATA
1144 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1145 rejects a recipient: pass the reject to the originator.
1147 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1148 Many were false-positives and ignorable, but it's worth fixing the
1151 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1152 for the new environment-manipulation done at startup. Move the routines
1153 from being local to tls.c to being global via the os.c file.
1155 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1156 an extract embedded as result-arg for a map, the first arg for extract
1157 is unavailable so we cannot tell if this is a numbered or keyed
1158 extraction. Accept either.
1164 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1167 JH/02 The smtp transport option "multi_domain" is now expanded.
1169 JH/03 The smtp transport now requests PRDR by default, if the server offers
1172 JH/04 Certificate name checking on server certificates, when exim is a client,
1173 is now done by default. The transport option tls_verify_cert_hostnames
1174 can be used to disable this per-host. The build option
1175 EXPERIMENTAL_CERTNAMES is withdrawn.
1177 JH/05 The value of the tls_verify_certificates smtp transport and main options
1178 default to the word "system" to access the system default CA bundle.
1179 For GnuTLS, only version 3.0.20 or later.
1181 JH/06 Verification of the server certificate for a TLS connection is now tried
1182 (but not required) by default. The verification status is now logged by
1183 default, for both outbound TLS and client-certificate supplying inbound
1186 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1189 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1190 Status Notification (bounce) messages are now MIME format per RFC 3464.
1191 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1192 under the control of the dsn_advertise_hosts option, and routers may
1193 have a dsn_lasthop option.
1195 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1196 default, modifiable by a malware= option. The list separator for
1197 the options can now be changed in the usual way. Bug 68.
1199 JH/10 The smtp_receive_timeout main option is now expanded before use.
1201 JH/11 The incoming_interface log option now also enables logging of the
1202 local interface on delivery outgoing connections.
1204 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1205 if the interface and destination host and port all match.
1207 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1210 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1211 Patch from Andrew Lewis.
1213 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1214 now supports optional time-restrictions, weighting, and priority
1215 modifiers per server. Patch originally by <rommer@active.by>.
1217 JH/16 The spamd_address main option now supports a mixed list of local
1218 and remote servers. Remote servers can be IPv6 addresses, and
1219 specify a port-range.
1221 JH/17 Bug 68: The spamd_address main option now supports an optional
1222 timeout value per server.
1224 JH/18 Bug 1581: Router and transport options headers_add/remove can
1225 now have the list separator specified.
1227 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1230 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1233 JH/21 Support for the A6 type of dns record is withdrawn.
1235 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1236 rather than the verbs used.
1238 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1239 from 255 to 1024 chars.
1241 JH/24 Verification callouts now attempt to use TLS by default.
1243 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1244 are generic router options now. The defaults didn't change.
1246 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1247 Original patch from Alexander Shikoff, worked over by JH.
1249 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1250 files if ZCAT_COMMAND is not executable.
1252 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1254 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1256 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1257 Normally benign, it bites when the pair was led to by a CNAME;
1258 modern usage is to not canonicalize the domain to a CNAME target
1259 (and we were inconsistent anyway for A-only vs AAAA+A).
1261 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1263 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1264 when evaluating $sender_host_dnssec.
1266 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1267 $sender_helo_dnssec variable.
1269 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1271 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1273 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1275 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1276 documented as working, but never had. Support all but $spam_report.
1278 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1279 added for tls authenticator.
1281 HS/03 Add perl_taintmode main config option
1287 TL/01 When running the test suite, the README says that variables such as
1288 no_msglog_check are global and can be placed anywhere in a specific
1289 test's script, however it was observed that placement needed to be near
1290 the beginning for it to behave that way. Changed the runtest perl
1291 script to read through the entire script once to detect and set these
1292 variables, reset to the beginning of the script, and then run through
1293 the script parsing/test process like normal.
1295 TL/02 The BSD's have an arc4random API. One of the functions to induce
1296 adding randomness was arc4random_stir(), but it has been removed in
1297 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1298 function when detected.
1300 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1301 cause callback expansion.
1303 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1304 syntax errors in an expansion can be treated as a string instead of
1305 logging or causing an error, due to the internal use of bool_lax
1306 instead of bool when processing it.
1308 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1309 server certificates when making smtp deliveries.
1311 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1313 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1315 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1317 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1318 Merged patch from Sebastian Wiedenroth.
1320 JH/05 Fix results-pipe from transport process. Several recipients, combined
1321 with certificate use, exposed issues where response data items split
1322 over buffer boundaries were not parsed properly. This eventually
1323 resulted in duplicates being sent. This issue only became common enough
1324 to notice due to the introduction of connection certificate information,
1325 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1327 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1328 size buffer was used, resulting in syntax errors when an expansion
1331 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1332 version 3.3.6 or later.
1334 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1335 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1336 both become "event_action", the variables become $event_name, $event_data
1337 and $event_defer_errno. There is a new variable $verify_mode, usable in
1338 routers, transports and related events. The tls:cert event is now also
1339 raised for inbound connections, if the main configuration event_action
1342 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1343 early OCSP support, but no stapling (appears to be less than 1.0.0).
1345 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1346 server certificate names available under the smtp transport option
1347 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1350 JH/10 Time-related extraction expansions from certificates now use the main
1351 option "timezone" setting for output formatting, and are consistent
1352 between OpenSSL and GnuTLS compilations. Bug 1541.
1354 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1355 encoded parameter in the incoming message. Bug 1558.
1357 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1358 include certificate info, eximon was claiming there were spoolfile
1361 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1363 JH/14 Log delivery-related information more consistently, using the sequence
1364 "H=<name> [<ip>]" wherever possible.
1366 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1367 are problematic for Debian distribution, omit them from the release
1370 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1372 JH/16 Fix string representation of time values on 64bit time_t architectures.
1375 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1376 output list separator was used.
1381 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1382 checkers that were complaining about end of non-void function with no
1385 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1386 This was a regression introduced in 4.83 by another bugfix.
1388 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1390 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1391 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1397 TF/01 Correctly close the server side of TLS when forking for delivery.
1399 When a message was received over SMTP with TLS, Exim failed to clear up
1400 the incoming connection properly after forking off the child process to
1401 deliver the message. In some situations the subsequent outgoing
1402 delivery connection happened to have the same fd number as the incoming
1403 connection previously had. Exim would try to use TLS and fail, logging
1404 a "Bad file descriptor" error.
1406 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1407 utilities have not been installed.
1409 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1410 temporary space as the ACL may create new global variables.
1412 TL/01 LDAP support uses per connection or global context settings, depending
1413 upon the detected version of the libraries at build time.
1415 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1416 to extract and use the src ip:port in logging and expansions as if it
1417 were a direct connection from the outside internet. PPv2 support was
1418 updated based on HAProxy spec change in May 2014.
1420 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1422 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1423 Properly escape header and check for NULL return.
1425 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1428 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1430 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1431 characters in header names, implemented as a verify condition.
1432 Contributed by Michael Fischer v. Mollard.
1434 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1435 results permerror and temperror. Previous values are deprecated but
1436 still accepted. In a future release, err_perm and err_temp will be
1437 completely removed, which will be a backward incompatibility if the
1438 ACL tests for either of these two old results. Patch contributed by
1439 user bes-internal on the mailing list.
1441 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1443 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1444 selectors, in both main and reject logs.
1446 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1449 JH/07 Add malware type "sock" for talking to simple daemon.
1451 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1453 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1454 routers/transports under cutthrough routing.
1456 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1457 numbers. Touch up "bool" conditional to keep the same definition.
1459 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1461 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1463 JH/12 Expand items in router/transport headers_add or headers_remove lists
1464 individually rather than the list as a whole. Bug 1452.
1466 Required for reasonable handling of multiple headers_ options when
1467 they may be empty; requires that headers_remove items with embedded
1468 colons must have them doubled (or the list-separator changed).
1470 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1471 view the policy declared in the DMARC record. Currently, $dmarc_status
1472 is a combined value of both the record presence and the result of the
1475 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1477 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1478 dnslookup router and the smtp transport (applying to the forward
1481 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1482 of ldap servers used for a specific lookup. Patch provided by Heiko
1485 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1486 New variable $lookup_dnssec_authenticated for observability.
1488 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1489 Patch submitted by Lars Timman.
1491 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1493 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1494 Requires trusted mode and valid format message id, aborts otherwise.
1495 Patch contributed by Heiko Schlichting.
1497 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1498 certextract with support for various fields. Bug 1358.
1500 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1501 is requested by default, modifiable by smtp transport option
1504 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1505 operate on certificate variables to give certificate fingerprints
1506 Also new ${sha256:cert_variable}.
1508 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1510 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1511 Christian Aistleitner.
1513 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1515 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1516 file. Patch from Wolfgang Breyha.
1518 JH/25 Expand the coverage of the delivery $host and $host_address to
1519 client authenticators run in verify callout. Bug 1476.
1521 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1522 align with daemon_smtp_ports. Bug 72.
1524 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1525 support and error reporting did not work properly.
1527 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1528 and is readable. Patch from Andrew Colin Kissa.
1530 TL/14 Enhance documentation of ${run expansion and how it parses the
1531 commandline after expansion, particularly in the case when an
1532 unquoted variable expansion results in an empty value.
1534 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1536 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1537 by renaming away from it. Addresses GH issue 15, reported by
1540 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1542 TL/15 SECURITY: prevent double expansion in math comparison functions
1543 (can expand unsanitized data). Not remotely exploitable.
1550 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1552 PP/02 Make -n do something, by making it not do something.
1553 When combined with -bP, the name of an option is not output.
1555 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1558 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1559 $sender_host_name and config options to manage this, and basic check
1562 PP/05 DSCP support for outbound connections and control modifier for inbound.
1564 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1565 (Only plugin which currently uses this is kerberos4, which nobody should
1566 be using, but we should make it available and other future plugins might
1567 conceivably use it, even though it would break NAT; stuff *should* be
1568 using channel bindings instead).
1570 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1571 name; added for Sendmail compatibility; requires admin caller.
1572 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1573 just ignore it); requires trusted caller.
1574 Also parse but ignore: -Ac -Am -X<logfile>
1577 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1579 TL/02 Add +smtp_confirmation as a default logging option.
1581 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1582 Patch by Magnus Holmgren from 2007-02-20.
1584 TL/04 Bugzilla 1281 - Spec typo.
1585 Bugzilla 1283 - Spec typo.
1586 Bugzilla 1290 - Spec grammar fixes.
1588 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1590 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1592 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1593 mailing list by by Dmitry Isaikin.
1595 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1597 JH/02 Support "G" suffix to numbers in ${if comparisons.
1599 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1601 NM/01 Bugzilla 1197 - Spec typo
1602 Bugzilla 1196 - Spec examples corrections
1604 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1606 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1607 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1610 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1611 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1613 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1614 "acl {{name}{arg}...}", and optional args on acl condition
1617 JH/05 Permit multiple router/transport headers_add/remove lines.
1619 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1621 JH/07 Avoid using a waiting database for a single-message-only transport.
1622 Performance patch from Paul Fisher. Bugzilla 1262.
1624 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1627 JH/09 Add $headers_added variable, with content from use of ACL modifier
1628 add_header (but not yet added to the message). Bugzilla 199.
1630 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1631 Pulled from Bugzilla 817 by Wolfgang Breyha.
1633 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1635 (nb: this is the same fix as in Exim 4.80.1)
1637 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1640 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1643 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1645 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1646 advertise SMTP AUTH mechanism to us, instead of a generic
1647 protocol violation error. Also, make Exim more robust to bad
1648 data from the Dovecot auth socket.
1650 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1652 When a queue runner is handling a message, Exim first routes the
1653 recipient addresses, during which it prunes them based on the retry
1654 hints database. After that it attempts to deliver the message to
1655 any remaining recipients. It then updates the hints database using
1658 So if a recipient address works intermittently, it can get repeatedly
1659 deferred at routing time. The retry hints record remains fresh so the
1660 address never reaches the final cutoff time.
1662 This is a fairly common occurrence when a user is bumping up against
1663 their storage quota. Exim had some logic in its local delivery code
1664 to deal with this. However it did not apply to per-recipient defers
1665 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1667 This change adds a proper retry rule check during routing so that the
1668 final cutoff time is checked against the message's age. We only do
1669 this check if there is an address retry record and there is not a
1670 domain retry record; this implies that previous attempts to handle
1671 the address had the retry_use_local_parts option turned on. We use
1672 this as an approximation for the destination being like a local
1673 delivery, as in LMTP.
1675 I suspect this new check makes the old local delivery cutoff check
1676 redundant, but I have not verified this so I left the code in place.
1678 TF/02 Correct gecos expansion when From: is a prefix of the username.
1680 Test 0254 submits a message to Exim with the header
1684 When I ran the test suite under the user fanf2, Exim expanded
1685 the header to contain my full name, whereas it should have added
1686 a Resent-Sender: header. It erroneously treats any prefix of the
1687 username as equal to the username.
1689 This change corrects that bug.
1691 GF/01 DCC debug and logging tidyup
1692 Error conditions log to paniclog rather than rejectlog.
1693 Debug lines prefixed by "DCC: " to remove any ambiguity.
1695 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1697 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1698 Bug spotted by Jeremy Harris; was flawed since initial commit.
1699 Would have resulted in OCSP responses post-SNI triggering an Exim
1700 NULL dereference and crash.
1702 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1704 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1705 Bug detection, analysis and fix by Samuel Thibault.
1706 Bugzilla 1331, Debian bug #698092.
1708 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1710 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1711 Server implementation by Todd Lyons, client by JH.
1712 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1713 config variable "prdr_enable" controls whether the server
1714 advertises the facility. If the client requests PRDR a new
1715 acl_data_smtp_prdr ACL is called once for each recipient, after
1716 the body content is received and before the acl_smtp_data ACL.
1717 The client is controlled by both of: a hosts_try_prdr option
1718 on the smtp transport, and the server advertisement.
1719 Default client logging of deliveries and rejections involving
1720 PRDR are flagged with the string "PRDR".
1722 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1723 fclose(). Diagnosis by Todd Lyons.
1725 PP/17 Update configure.default to handle IPv6 localhost better.
1726 Patch by Alain Williams (plus minor tweaks).
1729 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1730 This is now consistent with GnuTLS, and is now documented: the
1731 previous undocumented portable approach to treating the option as
1732 unset was to force an expansion failure. That still works, and
1733 an empty string is now equivalent.
1735 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1736 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1737 not performing validation itself.
1739 PP/20 Added force_command boolean option to pipe transport.
1740 Patch from Nick Koston, of cPanel Inc.
1742 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1745 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1747 PP/21 Fix eximon continuous updating with timestamped log-files.
1748 Broken in a format-string cleanup in 4.80, missed when I repaired the
1749 other false fix of the same issue.
1750 Report and fix from Heiko Schlichting.
1753 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1754 Report from Prashanth Katuri.
1756 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1757 It's SecureTransport, so affects any MacOS clients which use the
1758 system-integrated TLS libraries, including email clients.
1760 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1761 using a MIME ACL for non-SMTP local injection.
1762 Report and assistance in diagnosis by Warren Baker.
1764 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1766 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1768 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1769 last id that failed so it may be referenced in subsequent ACL's.
1771 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1774 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1775 ldap library initialization, allowing self-signed CA's to be
1776 used. Also properly sets require_cert option later in code by
1777 using NULL (global ldap config) instead of ldap handle (per
1778 session). Bug diagnosis and testing by alxgomz.
1780 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1781 the src/util/ subdirectory.
1783 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1784 renamed to Transport Post Delivery Action by Jeremy Harris, as
1787 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1788 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1789 redis_servers = needs to be configured which will be used by the redis
1790 lookup. Patch from Warren Baker, of The Packet Hub.
1792 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1794 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1795 hostname or reverse DNS when processing a host list. Used suggestions
1796 from multiple comments on this bug.
1798 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1800 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1801 Missed a few lines, added it to make the runtest require no keyboard
1804 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1805 contains upper case chars. Make router use caseful_local_part.
1807 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1808 support when GnuTLS has been built with p11-kit.
1814 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1816 This, or similar/improved, will also be change PP/11 of 4.82.
1822 PP/01 Handle short writes when writing local log-files.
1823 In practice, only affects FreeBSD (8 onwards).
1824 Bugzilla 1053, with thanks to Dmitry Isaikin.
1826 NM/01 Bugzilla 949 - Documentation tweak
1828 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1831 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1833 PP/02 Implemented gsasl authenticator.
1835 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1837 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1838 `pkg-config foo` for cflags/libs.
1840 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1841 with rest of GSASL and with heimdal_gssapi.
1843 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1844 `pkg-config foo` for cflags/libs for the TLS implementation.
1846 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1847 properties get this fed in as external SSF. A number of robustness
1848 and debugging improvements to the cyrus_sasl authenticator.
1850 PP/08 cyrus_sasl server now expands the server_realm option.
1852 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1853 Patch by Jeremy Harris.
1855 PP/10 Added dbmjz lookup type.
1857 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1859 PP/12 MAIL args handles TAB as well as SP, for better interop with
1860 non-compliant senders.
1861 Analysis and variant patch by Todd Lyons.
1863 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1864 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1865 Patch from Dirk Mueller <dmueller@suse.com>
1867 PP/13 tls_peerdn now print-escaped for spool files.
1868 Observed some $tls_peerdn in wild which contained \n, which resulted
1869 in spool file corruption.
1871 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1872 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1873 or write after TLS renegotiation, which otherwise led to messages
1876 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1877 as a tracking header (ie: a signed header comes before the signature).
1878 Patch from Wolfgang Breyha.
1880 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1881 comma-sep list; embedded commas doubled.
1883 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1885 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1887 Report and patch from Dmitry Banschikov.
1889 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1890 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1891 failures. We appear to now support TLS1.1+ with Exim.
1893 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1894 lets Exim select keys and certificates based upon TLS SNI from client.
1895 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1896 before an outbound SMTP session. New log_selector, +tls_sni.
1898 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1899 NULL dereference. Report and patch from Alun Jones.
1901 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1902 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1903 Not seeing resolver debug output on NetBSD, but suspect this is a
1904 resolver implementation change.
1906 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1907 Left warnings. Added "eximon gdb" invocation mode.
1909 PP/21 Defaulting "accept_8bitmime" to true, not false.
1911 PP/22 Added -bw for inetd wait mode support.
1913 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1914 locate the relevant includes and libraries. Made this the default.
1916 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1917 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1919 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1920 This may cause build issues on older platforms.
1922 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1923 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1924 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1925 Added SNI support via GnuTLS too.
1926 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1928 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1930 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1931 Applied second patch from Janne, implementing suggestion to default
1932 multiple-strings-in-record handling to match SPF spec.
1934 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1936 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1937 read-only, out of scope).
1938 Patch from Wolfgang Breyha, report from Stuart Northfield.
1940 PP/29 Fix three issues highlighted by clang analyser static analysis.
1941 Only crash-plausible issue would require the Cambridge-specific
1942 iplookup router and a misconfiguration.
1943 Report from Marcin Mirosław.
1945 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1947 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1948 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1949 As part of this, removing so much warning spew let me fix some minor
1950 real issues in debug logging.
1952 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1953 assignment on my part. Fixed.
1955 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1956 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1957 Janne Snabb (who went above and beyond: thank you).
1959 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1960 string otherwise requires a connection and a bunch more work and it's
1961 relatively easy to get wrong. Should also expose TLS library linkage
1964 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1965 64-bit ${eval} (JH/03).
1967 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1968 GNU libc to support some of the 64-bit stuff, should not lead to
1969 conflicts. Defined before os.h is pulled in, so if a given platform
1970 needs to override this, it can.
1972 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1973 protection layer was required, which is not implemented.
1974 Bugzilla 1254, patch from Wolfgang Breyha.
1976 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1977 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1978 tls_dhparam take prime identifiers. Also unbreak combination of
1979 OpenSSL+DH_params+TLSSNI.
1981 PP/39 Disable SSLv2 by default in OpenSSL support.
1987 PP/01 Solaris build fix for Oracle's LDAP libraries.
1988 Bugzilla 1109, patch from Stephen Usher.
1990 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1992 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1995 TF/02 Fix a couple more cases where we did not log the error message
1996 when unlink() failed. See also change 4.74-TF/03.
1998 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1999 lock up or crash if it happened to be inside a call to libc when it
2000 got a SIGUSR1 from exiwhat.
2002 The SIGUSR1 handler appends the current process status to the process
2003 log which is later printed by exiwhat. It used to use the general
2004 purpose logging code to do this, but several functions it calls are
2005 not safe for signals.
2007 The new output code in the SIGUSR1 handler is specific to the process
2008 log, and simple enough that it's easy to inspect for signal safety.
2009 Removing some special cases also simplifies the general logging code.
2010 Removing the spurious timestamps from the process log simplifies
2013 TF/04 Improved ratelimit ACL condition.
2015 The /noupdate option has been deprecated in favour of /readonly which
2016 has clearer semantics. The /leaky, /strict, and /readonly update modes
2017 are mutually exclusive. The update mode is no longer included in the
2018 database key; it just determines when the database is updated. (This
2019 means that when you upgrade Exim will forget old rate measurements.)
2021 Exim now checks that the per_* options are used with an update mode that
2022 makes sense for the current ACL. For example, when Exim is processing a
2023 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2024 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2025 must specify per_mail/readonly. If you omit the update mode it defaults to
2026 /leaky where that makes sense (as before) or /readonly where required.
2028 The /noupdate option is now undocumented but still supported for
2029 backwards compatibility. It is equivalent to /readonly except that in
2030 ACLs where /readonly is required you may specify /leaky/noupdate or
2031 /strict/noupdate which are treated the same as /readonly.
2033 A useful new feature is the /count= option. This is a generalization
2034 of the per_byte option, so that you can measure the throughput of other
2035 aggregate values. For example, the per_byte option is now equivalent
2036 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2038 The per_rcpt option has been generalized using the /count= mechanism
2039 (though it's more complicated than the per_byte equivalence). When it is
2040 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2041 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2042 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2043 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2044 Note that using per_rcpt with a non-readonly update mode in more than
2045 one ACL will cause the recipients to be double-counted. (The per_mail
2046 and per_byte options don't have this problem.)
2048 The handling of very low rates has changed slightly. If the computed rate
2049 is less than the event's count (usually one) then this event is the first
2050 after a long gap. In this case the rate is set to the same as this event's
2051 count, so that the first message of a spam run is counted properly.
2053 The major new feature is a mechanism for counting the rate of unique
2054 events. The new per_addr option counts the number of different
2055 recipients that someone has sent messages to in the last time period. It
2056 behaves like per_rcpt if all the recipient addresses are different, but
2057 duplicate recipient addresses do not increase the measured rate. Like
2058 the /count= option this is a general mechanism, so the per_addr option
2059 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2060 example, measure the rate that a client uses different sender addresses
2061 with the options per_mail/unique=$sender_address. There are further
2062 details in the main documentation.
2064 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2066 TF/06 Removed a few PCRE remnants.
2068 TF/07 Automatically extract Exim's version number from tags in the git
2069 repository when doing development or release builds.
2071 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2072 Bugzilla 879. Patch from Paul Fisher.
2074 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2075 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2078 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2080 PP/05 Variable $av_failed, true if the AV scanner deferred.
2081 Bugzilla 1078. Patch from John Horne.
2083 PP/06 Stop make process more reliably on build failure.
2084 Bugzilla 1087. Patch from Heiko Schlittermann.
2086 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2087 Bugzilla 1089. Patch from Heiko Schlittermann.
2089 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2090 Bugzilla 1131. Patch from Holger Weiß.
2092 PP/09 Handle IPv6 addresses with SPF.
2093 Bugzilla 860. Patch from Wolfgang Breyha.
2095 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2097 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2100 PP/11 match_* no longer expand right-hand-side by default.
2101 New compile-time build option, EXPAND_LISTMATCH_RHS.
2102 New expansion conditions, "inlist", "inlisti".
2104 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2106 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2108 PP/14 fix log_write() format string regression from TF/03.
2109 Bugzilla 1152. Patch from Dmitry Isaikin.
2115 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2117 PP/02 Harmonised TLS library version reporting; only show if debugging.
2118 Layout now matches that introduced for other libraries in 4.74 PP/03.
2120 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2122 PP/04 New "dns_use_edns0" global option.
2124 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2127 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2128 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2130 TK/01 Updated PolarSSL code to 0.14.2.
2131 Bugzilla 1097. Patch from Andreas Metzler.
2133 PP/07 Catch divide-by-zero in ${eval:...}.
2134 Fixes bugzilla 1102.
2136 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2139 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2140 format-string attack -- SECURITY: remote arbitrary code execution.
2142 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2143 time unintentionally subject to list matching rules, letting the header
2144 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2145 arbitrary string expansion). This allowed for information disclosure.
2147 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2148 INT_MIN/-1 -- value coerced to INT_MAX.
2154 NM/01 Workaround for PCRE version dependency in version reporting
2157 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2158 This fixes portability to compilers other than gcc, notably
2159 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2161 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2162 makefiles for portability to HP-UX and POSIX correctness.
2164 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2165 Also via indented variable definition in the Makefile.
2166 (Debugging by Oliver Heesakkers).
2168 PP/02 Restore caching of spamd results with expanded spamd_address.
2169 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2171 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2172 Improves build reliability. Fix from: Frank Elsner
2174 NM/02 Fix wide character breakage in the rfc2047 coding
2175 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2177 NM/03 Allow underscore in dnslist lookups
2178 Fixes bug 1026. Patch from Graeme Fowler
2180 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2181 Code patches from Adam Ciarcinski of NetBSD.
2183 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2186 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2187 is logged, to avoid truncation. Patch from John Horne.
2189 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2190 Patch from Jakob Hirsch.
2192 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2193 SQL string expansion failure details.
2194 Patch from Andrey Oktyabrski.
2196 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2197 Patch from Simon Arlott.
2199 PP/09 New lookups functionality failed to compile on old gcc which rejects
2200 extern declarations in function scope.
2201 Patch from Oliver Fleischmann
2203 PP/10 Use sig_atomic_t for flags set from signal handlers.
2204 Check getgroups() return and improve debugging.
2205 Fixed developed for diagnosis in bug 927 (which turned out to be
2208 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2209 Patch from Mark Zealey.
2211 PP/12 Bugzilla 1056: Improved spamd server selection.
2212 Patch from Mark Zealey.
2214 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2215 Based on patch from Heiko Schlittermann.
2217 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2218 Patch from Uwe Doering, sign-off by Michael Haardt.
2220 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2221 variable declaration deep within a block. Bug and patch from
2224 PP/15 lookups-Makefile IRIX compatibility coercion.
2226 PP/16 Make DISABLE_DKIM build knob functional.
2228 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2229 Patch by Simon Arlott
2231 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2232 variable argument macros. Our copy now differs from upstream.
2238 TF/01 Failure to get a lock on a hints database can have serious
2239 consequences so log it to the panic log.
2241 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2242 controlled using the smtp_confirmation log selector.
2244 TF/03 Include the error message when we fail to unlink a spool file.
2246 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2247 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2248 for maintaining out-of-tree patches for some time.
2250 PP/01 Bugzilla 139: Documentation and portability issues.
2251 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2252 Handle per-OS dynamic-module compilation flags.
2254 PP/02 Let /dev/null have normal permissions.
2255 The 4.73 fixes were a little too stringent and complained about the
2256 permissions on /dev/null. Exempt it from some checks.
2257 Reported by Andreas M. Kirchwitz.
2259 PP/03 Report version information for many libraries, including
2260 Exim version information for dynamically loaded libraries. Created
2261 version.h, now support a version extension string for distributors
2262 who patch heavily. Dynamic module ABI change.
2264 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2265 privilege escalation vulnerability whereby the Exim run-time user
2266 can cause root to append content of the attacker's choosing to
2269 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2272 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2273 If dropping privileges for untrusted macros, we disabled normal logging
2274 on the basis that it would fail; for the Exim run-time user, this is not
2275 the case, and it resulted in successful deliveries going unlogged.
2276 Fixed. Reported by Andreas Metzler.
2282 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2283 only prepend for the Resent-* case. Fixes regression introduced in
2284 Exim 4.70 by NM/22 for Bugzilla 607.
2286 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2287 increasing numbers of administrators be bitten by this.
2289 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2291 PP/03 Bugzilla 994: added openssl_options main configuration option.
2293 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2295 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2297 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2299 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2300 without a peer certificate, leading to a segfault because of an
2301 assumption that peers always have certificates. Be a little more
2302 paranoid. Problem reported by Martin Tscholak.
2304 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2305 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2306 NB: ClamAV planning to remove STREAM in "middle of 2010".
2307 CL also introduces -bmalware, various -d+acl logging additions and
2308 more caution in buffer sizes.
2310 PP/09 Implemented reverse_ip expansion operator.
2312 PP/10 Bugzilla 937: provide a "debug" ACL control.
2314 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2316 PP/12 Bugzilla 973: Implement --version.
2318 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2320 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2322 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2324 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2325 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2326 ignore trailing whitespace.
2328 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2330 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2333 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2334 Notification from Dr Andrew Aitchison.
2336 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2337 ExtendedDetectionInfo response format.
2338 Notification from John Horne.
2340 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2343 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2344 XSL and documented dependency on system catalogs, with examples of how
2347 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2350 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2351 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2352 configuration file which is writeable by the Exim user or group.
2354 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2355 of configuration files to cover files specified with the -C option if
2356 they are going to be used with root privileges, not just the default
2359 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2360 option (effectively making it always true).
2362 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2363 files to be used while preserving root privileges.
2365 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2366 that rogue child processes cannot use them.
2368 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2369 run-time user, instead of root.
2371 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2372 Exim run-time user without dropping privileges.
2374 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2375 result string, instead of calling string_vformat() twice with the same
2378 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2379 for other users. Others should always drop root privileges if they use
2380 -C on the command line, even for a whitelisted configure file.
2382 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2384 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2390 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2391 $data_path, and $header_path variables; fixed documentation bugs and
2394 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2395 exipick to access non-standard spools, including the "frozen" queue
2398 NM/01 Bugzilla 965: Support mysql stored procedures.
2399 Patch from Alain Williams
2401 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2403 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2404 Patch from Andreas Metzler
2406 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2407 Patch from Kirill Miazine
2409 NM/05 Bugzilla 671: Added umask to procmail example.
2411 JJ/03 installed exipick 20100323.0, fixing doc bug
2413 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2414 directory. Notification and patch from Dan Rosenberg.
2416 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2418 TK/02 Improve log output when DKIM signing operation fails.
2420 MH/01 Treat the transport option dkim_domain as a colon separated
2421 list, not as a single string, and sign the message with each element,
2422 omitting multiple occurences of the same signer.
2424 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2425 Bugzilla 985, 986. Patch by Simon Arlott
2427 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2428 Patch by Simon Arlott
2430 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2431 MBX locking. Notification from Dan Rosenberg.
2437 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2439 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2441 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2443 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2445 NM/04 Bugzilla 915: Flag broken perl installation during build.
2451 TK/01 Added patch by Johannes Berg that expands the main option
2452 "spamd_address" if it starts with a dollar sign.
2454 TK/02 Write list of recipients to X-Envelope-Sender header when building
2455 the mbox-format spool file for content scanning (suggested by Jakob
2458 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2459 (http://www.dcc-servers.net/) support via dccifd. Activated by
2460 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2462 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2463 by Mark Daniel Reidel <mr@df.eu>.
2465 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2466 When building exim an external PCRE library is now needed -
2467 PCRE is a system library on the majority of modern systems.
2468 See entry on PCRE_LIBS in EDITME file.
2470 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2471 conversation. Added nologin parameter to request.
2472 Patch contributed by Kirill Miazine.
2474 TF/01 Do not log submission mode rewrites if they do not change the address.
2476 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2478 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2479 log files in place. Contributed by Roberto Lima.
2481 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2483 TF/03 Bugzilla 615: When checking the local_parts router precondition
2484 after a local_part_suffix or local_part_prefix option, Exim now
2485 does not use the address's named list lookup cache, since this
2486 contains cached lookups for the whole local part.
2488 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2489 Robert Millan. Documentation is in experimental-spec.txt.
2491 TF/04 Bugzilla 668: Fix parallel build (make -j).
2493 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2495 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2496 Patch provided by Jan Srzednicki.
2498 TF/05 Leading white space used to be stripped from $spam_report which
2499 wrecked the formatting. Now it is preserved.
2501 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2502 that they are available at delivery time.
2504 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2506 TF/08 TLS error reporting now respects the incoming_interface and
2507 incoming_port log selectors.
2509 TF/09 Produce a more useful error message if an SMTP transport's hosts
2510 setting expands to an empty string.
2512 NM/06 Bugzilla 744: EXPN did not work under TLS.
2513 Patch provided by Phil Pennock.
2515 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2516 Patch provided by Richard Godbee.
2518 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2519 acl_smtp_notquit, added index entry.
2521 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2522 Patch provided by Eugene Bujak.
2524 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2525 accept(). Patch provided by Maxim Dounin.
2527 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2528 Patch provided by Phil Pennock.
2530 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2532 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2533 Patch provided by Brad "anomie" Jorsch.
2535 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2536 Patch provided by Dean Brooks.
2538 TK/05 Add native DKIM support (does not depend on external libraries).
2540 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2541 Patch provided by Graeme Fowler.
2543 NM/16 Bugzilla 851: Documentation example syntax fix.
2545 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2547 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2550 NM/19 Bugzilla 745: TLS version reporting.
2551 Patch provided by Phil Pennock.
2553 NM/20 Bugzilla 167: bool: condition support.
2554 Patch provided by Phil Pennock.
2556 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2557 clients. Patch provided by Phil Pennock.
2559 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2560 Patch provided by Brad "anomie" Jorsch.
2562 NM/23 Bugzilla 687: Fix misparses in eximstats.
2563 Patch provided by Heiko Schlittermann.
2565 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2566 Patch provided by Heiko Schlittermann.
2568 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2569 plus update to original patch.
2571 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2573 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2574 Patch provided by David Brownlee.
2576 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2578 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2580 NM/30 Bugzilla 888: TLS documentation bugfixes.
2582 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2584 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2585 Unlike the original bugzilla I have changed all shell scripts in src tree.
2587 NM/33 Bugzilla 898: Transport filter timeout fix.
2588 Patch by Todd Rinaldo.
2590 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2591 Patch by Serge Demonchaux.
2593 NM/35 Bugzilla 39: Base64 decode bug fixes.
2594 Patch by Jakob Hirsch.
2596 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2598 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2600 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2602 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2608 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2609 ALT-N's libdkim that I have put here:
2610 http://duncanthrax.net/exim-experimental/
2612 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2613 were called 'true' and it seems that C99 defines that as a reserved
2614 keyword to be used with 'bool' variable types. That means you could
2615 not include C99-style headers which use bools without triggering
2616 build errors in sieve.c.
2618 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2619 as mailq or other aliases. Changed the --help handling significantly
2620 to do whats expected. exim_usage() emits usage/help information.
2622 SC/01 Added the -bylocaldomain option to eximstats.
2624 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2626 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2628 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2634 PH/01 Another patch from the Sieve maintainer.
2636 PH/02 When an IPv6 address is converted to a string for single-key lookup
2637 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2638 dots are used instead of colons so that keys in lsearch files need not
2639 contain colons. This was done some time before quoting was made available
2640 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2641 (notated using the quote facility) so as to distinguish them from IPv4
2642 keys. This meant that lookups for IP addresses in host lists did not work
2643 for iplsearch lookups.
2645 This has been fixed by arranging for IPv6 addresses to be expressed with
2646 colons if the lookup type is iplsearch. This is not incompatible, because
2647 previously such lookups could never work.
2649 The situation is now rather anomalous, since one *can* have colons in
2650 ordinary lsearch keys. However, making the change in all cases is
2651 incompatible and would probably break a number of configurations.
2653 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2656 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2657 conversion specification without a maximum field width, thereby enabling
2658 a rogue spamd server to cause a buffer overflow. While nobody in their
2659 right mind would setup Exim to query an untrusted spamd server, an
2660 attacker that gains access to a server running spamd could potentially
2661 exploit this vulnerability to run arbitrary code as the Exim user.
2663 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2664 $primary_hostname instead of what libspf2 thinks the hosts name is.
2666 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2667 a directory entry by the name of the lookup key. Previously, if a
2668 symlink pointed to a non-existing file or a file in a directory that
2669 Exim lacked permissions to read, a lookup for a key matching that
2670 symlink would fail. Now it is enough that a matching directory entry
2671 exists, symlink or not. (Bugzilla 503.)
2673 PH/03 The body_linecount and body_zerocount variables are now exported in the
2676 PH/04 Added the $dnslist_matched variable.
2678 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2679 This means they are set thereafter only if the connection becomes
2682 PH/06 Added the client_condition to authenticators so that some can be skipped
2683 by clients under certain conditions.
2685 PH/07 The error message for a badly-placed control=no_multiline_responses left
2686 "_responses" off the end of the name.
2688 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2690 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2691 (without spaces) instead of just copying the configuration text.
2693 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2695 PH/11 Added $max_received_linelength.
2697 PH/12 Added +ignore_defer and +include_defer to host lists.
2699 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2700 way in which PCRE > 7.0 is built.
2702 PH/14 Implemented queue_only_load_latch.
2704 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2705 MAIL command. The effect was to mangle the value on 64-bit systems.
2707 PH/16 Another patch from the Sieve maintainer.
2709 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2711 PH/18 If a system quota error occurred while trying to create the file for
2712 a maildir delivery, the message "Mailbox is full" was not appended to the
2713 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2714 only to a quota excession during the actual writing of the file.
2716 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2717 characters?) which causes problems in log lines. The DN values are now
2718 passed through string_printing() before being added to log lines.
2720 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2721 and InterBase are left for another time.)
2723 PH/21 Added message_body_newlines option.
2725 PH/22 Guard against possible overflow in moan_check_errorcopy().
2727 PH/23 POSIX allows open() to be a macro; guard against that.
2729 PH/24 If the recipient of an error message contained an @ in the local part
2730 (suitably quoted, of course), incorrect values were put in $domain and
2731 $local_part during the evaluation of errors_copy.
2737 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2738 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2741 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2742 issue a MAIL command.
2744 PH/02 In an ACL statement such as
2746 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2748 if a client was not listed at all, or was listed with a value other than
2749 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2750 the condition was not true (as it should be), so access was not denied.
2751 The bug was that the ! inversion was incorrectly passed on to the second
2752 item. This has been fixed.
2754 PH/03 Added additional dnslists conditions == and =& which are different from
2755 = and & when the dns lookup returns more than one IP address.
2757 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2758 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2760 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2761 FSYNC, which compiles an option called disable_fsync that allows for
2762 bypassing fsync(). The documentation is heavily laced with warnings.
2764 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2766 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2767 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2768 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2769 including adding "make clean"; (3) Added -fPIC when compiling the test
2770 dynamically loaded module, to get rid of a warning.
2772 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2773 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2774 = 0s. The bug is otherwise harmless.
2776 PH/07 There was a bug in the dovecot authenticator such that the value of
2777 $auth1 could be overwritten, and so not correctly preserved, after a
2778 successful authentication. This usually meant that the value preserved by
2779 the server_setid option was incorrect.
2781 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2783 PH/09 Installed PCRE release 7.0.
2785 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2786 run for batched SMTP input. It is now run at the start of every message
2787 in the batch. While fixing this I discovered that the process information
2788 (output by running exiwhat) was not always getting set for -bs and -bS
2789 input. This is fixed, and it now also says "batched" for BSMTP.
2791 PH/11 Added control=no_pipelining.
2793 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2794 patch, slightly modified), and move the expansion of helo_data till after
2795 the connection is made in the smtp transport (so it can use these
2798 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2800 PH/14 Added log_selector = +pid.
2802 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2804 PH/16 Add ${if forany and ${if forall.
2806 PH/17 Added dsn_from option to vary the From: line in DSNs.
2808 PH/18 Flush SMTP output before performing a callout, unless control =
2809 no_callout_flush is set.
2811 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2812 was true (the default) a successful delivery failed to delete the retry
2813 item, thus causing premature timeout of the address. The bug is now
2816 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2818 PH/21 Long custom messages for fakedefer and fakereject are now split up
2819 into multiline responses in the same way that messages for "deny" and
2820 other ACL rejections are.
2822 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2823 with slight modification.
2825 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2826 draft, changing the syntax and factoring some duplicate code".
2828 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2829 for deliveries of the second and subsequent messages over the same SMTP
2832 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2833 ${reduce, with only minor "tidies".
2835 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2837 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2838 expansion side effects.
2840 PH/27 When a message times out after an over-quota error from an Exim-imposed
2841 quota, the bounce message says "mailbox is full". This message was not
2842 being given when it was a system quota that was exceeded. It now should
2845 MH/03 Made $recipients available in local_scan(). local_scan() already has
2846 better access to the recipient list through recipients_list[], but
2847 $recipients can be useful in postmaster-provided expansion strings.
2849 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2850 in the case of a MAIL command with additional options following the
2851 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2852 were accidentally chopped off.
2854 PH/29 SMTP synchronization checks are implemented when a command is read -
2855 there is a check that no more input is waiting when there shouldn't be
2856 any. However, for some commands, a delay in an ACL can mean that it is
2857 some time before the response is written. In this time, more input might
2858 arrive, invalidly. So now there are extra checks after an ACL has run for
2859 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2860 pipelining has not been advertised.
2862 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2864 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2865 returned in the SMTP response when smtp_return_error_details was false.
2866 This has been fixed.
2868 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2869 instead of the C I/O that was originally supplied, because problems were
2870 reported on Solaris.
2872 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2873 Exim which did not show up earlier: it was assuming that a call to
2874 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2875 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2876 was a macro that became an assignment, so it seemed to work. This has
2877 changed to a proper function call with a void return, hence the compile
2878 error. Exim's code has been fixed.
2880 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2883 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2885 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2887 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2888 because it turns out to be too big or there is a timeout) there is no
2889 "Completed" line in the log. When some messages of this type were
2890 selected by exigrep, they were listed as "not completed". Others were
2891 picked up by some special patterns. I have improved the selection
2892 criteria to be more general.
2894 PH/38 The host_find_failed option in the manualroute router can now be set
2895 to "ignore", to completely ignore a host whose IP address cannot be
2896 found. If all hosts are ignored, the behaviour is controlled by the new
2897 host_all_ignored option.
2899 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2900 homing or because of multiple MX records with /mx) generated more than
2901 one IP address, and the following item turned out to be the local host,
2902 all the secondary addresses of the first item were incorrectly removed
2903 from the list, along with the local host and any following hosts (which
2904 is what is supposed to happen).
2906 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2907 whoever called Exim into the -H file. In the case of the daemon it was
2908 behaving confusingly. When first started, it used values for whoever
2909 started the daemon, but after a SIGHUP it used the Exim user (because it
2910 calls itself on a restart). I have changed the code so that it now always
2913 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2914 message are rejected with the same error (e.g. no authentication or bad
2915 sender address), and a DATA command is nevertheless sent (as can happen
2916 with PIPELINING or a stupid MUA), the error message that was given to the
2917 RCPT commands is included in the rejection of the DATA command. This is
2918 intended to be helpful for MUAs that show only the final error to their
2921 PH/42 Another patch from the Sieve maintainer.
2923 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2924 Eximstats - Fixed some broken HTML links and added missing column headers
2926 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2927 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2929 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2931 PH/43 Yet another patch from the Sieve maintainer.
2933 PH/44 I found a way to check for a TCP/IP connection going away before sending
2934 the response to the final '.' that terminates a message, but only in the
2935 case where the client has not sent further data following the '.'
2936 (unfortunately, this is allowed). However, in many cases there won't be
2937 any further data because there won't be any more messages to send. A call
2938 to select() can be used: if it shows that the input is "ready", there is
2939 either input waiting, or the socket has been closed. An attempt to read
2940 the next input character can distinguish the two cases. Previously, Exim
2941 would have sent an OK response which the client would never have see.
2942 This could lead to message repetition. This fix should cure that, at
2943 least in a lot of common cases.
2945 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2946 advertised in response to EHLO.
2952 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2953 fixed by 4.65/MH/01 (is this a record?) are fixed:
2955 (i) An empty string was always treated as zero by the numeric comparison
2956 operators. This behaviour has been restored.
2958 (ii) It is documented that the numeric comparison operators always treat
2959 their arguments as decimal numbers. This was broken in that numbers
2960 starting with 0 were being interpreted as octal.
2962 While fixing these problems I realized that there was another issue that
2963 hadn't been noticed. Values of message_size_limit (both the global option
2964 and the transport option) were treated as octal if they started with 0.
2965 The documentation was vague. These values are now always treated as
2966 decimal, and I will make that clear in the documentation.
2972 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2973 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2976 MH/01 Don't check that the operands of numeric comparison operators are
2977 integers when their expansion is in "skipping" mode (fixes bug
2978 introduced by 4.64-PH/07).
2980 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2981 child addresses, Exim now panics and dies. Previously, because the count
2982 is held in a short int, deliveries were likely to be lost. As such a
2983 large number of recipients for a single message is ridiculous
2984 (performance will be very, very poor), I have chosen to impose a limit
2985 rather than extend the field.
2991 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2992 leftover -K file (the existence of which was triggered by #402).
2993 While we were at it, introduced process PID as part of the -K
2994 filename. This should rule out race conditions when creating
2997 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2998 processing considerably. Previous code took too long for large mails,
2999 triggering a timeout which in turn triggers #401.
3001 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3002 in the DK code in transports.c. sendfile() is not really portable,
3003 hence the _LINUX specificness.
3005 TF/01 In the add_headers option to the mail command in an Exim filter,
3006 there was a bug that Exim would claim a syntax error in any
3007 header after the first one which had an odd number of characters
3010 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3011 callout verification, Exim cached a "reject" for the entire domain. This
3012 is correct for most verifications, but it is not correct for a recipient
3013 verification with use_sender or use_postmaster set, because in that case
3014 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3015 case of MAIL FROM:<> rejection from other early rejections (e.g.
3016 rejection of HELO). When verifying a recipient using a non-null MAIL
3017 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3018 Whatever the result of the callout, the value of the domain cache is
3019 left unchanged (for any other kind of callout, getting as far as trying
3020 RCPT means that the domain itself is ok).
3022 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3025 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3026 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3027 session, the daemon ignores ECONNECT errors and logs others; it now
3028 ignores EPIPE as well.
3030 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3031 (quoted-printable decoding).
3033 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3034 later the small subsequent patch to fix an introduced bug.
3036 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3038 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3040 PH/08 An error is now given if message_size_limit is specified negative.
3042 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3043 to be given (somewhat) arbitrary names.
3045 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3048 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3049 miscellaneous code fixes
3051 PH/10 Added the log_reject_target ACL modifier to specify where to log
3054 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3055 hostname. This is wrong, because it relates to the incoming message (and
3056 probably the interface on which it is arriving) and not to the outgoing
3057 callout (which could be using a different interface). This has been
3058 changed to use the value of the helo_data option from the smtp transport
3059 instead - this is what is used when a message is actually being sent. If
3060 there is no remote transport (possible with a router that sets up host
3061 addresses), $smtp_active_hostname is used.
3063 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3064 tweaks were necessary in order to get it to work (see also 21 below):
3065 (a) The code assumed that strncpy() returns a negative number on buffer
3066 overflow, which isn't the case. Replaced with Exim's string_format()
3068 (b) There were several signed/unsigned issues. I just did the minimum
3069 hacking in of casts. There is scope for a larger refactoring.
3070 (c) The code used strcasecmp() which is not a standard C function.
3071 Replaced with Exim's strcmpic() function.
3072 (d) The code set only $1; it now sets $auth1 as well.
3073 (e) A simple test gave the error "authentication client didn't specify
3074 service in request". It would seem that Dovecot has changed its
3075 interface. Fortunately there's a specification; I followed it and
3076 changed what the client sends and it appears to be working now.
3078 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3081 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3082 address A is aliased to B and C, where B exists and C does not. Without
3083 -v the output is "A verified" because verification stops after a
3084 successful redirection if more than one address is generated. However,
3085 with -v the child addresses are also verified. Exim was outputting "A
3086 failed to verify" and then showing the successful verification for C,
3087 with its parentage. It now outputs "B failed to verify", showing B's
3088 parentage before showing the successful verification of C.
3090 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3091 look up a TXT record in a specific list after matching in a combined
3094 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3095 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3096 they consult the DNS. I had assumed they would set it the way they
3097 wanted; and indeed my experiments on Linux seem to show that in some
3098 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3099 To be on the safe side, however, I have now made the interface to
3100 host_find_byname() similar to host_find_bydns(), with an argument
3101 containing the DNS resolver options. The host_find_byname() function now
3102 sets these options at its start, just as host_find_bydns() does. The smtp
3103 transport options dns_qualify_single and dns_search_parents are passed to
3104 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3105 of host_find_byname() use the default settings of RES_DEFNAMES
3106 (qualify_single) but not RES_DNSRCH (search_parents).
3108 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3109 spool_read_header() do less string testing, by means of a preliminary
3110 switch on the second character of optional "-foo" lines. (This is
3111 overdue, caused by the large number of possibilities that now exist.
3112 Originally there were few.) While I was there, I also converted the
3113 str(n)cmp tests so they don't re-test the leading "-" and the first
3114 character, in the hope this might squeeze out yet more improvement.
3116 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3117 flag allowing group syntax was set by the header_syntax check but not
3118 turned off, possible causing trouble later; (2) The flag was not being
3119 set at all for the header_verify test, causing "group"-style headers to
3120 be rejected. I have now set it in this case, and also caused header_
3121 verify to ignore an empty address taken from a group. While doing this, I
3122 came across some other cases where the code for allowing group syntax
3123 while scanning a header line wasn't quite right (mostly, not resetting
3124 the flag correctly in the right place). These bugs could have caused
3125 trouble for malformed header lines. I hope it is now all correct.
3127 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3128 with the "reply" argument non-NULL. The code, however (which originally
3129 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3130 but it didn't always do it. This confused somebody who was copying the
3131 code for some other use. I have removed all the tests.
3133 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3134 feature that was used to support insecure browsers during the U.S. crypto
3135 embargo. It requires special client support, and Exim is probably the
3136 only MTA that supported it -- and would never use it because real RSA is
3137 always available. This code has been removed, because it had the bad
3138 effect of slowing Exim down by computing (never used) parameters for the
3139 RSA_EXPORT functionality.
3141 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3142 authenticator to fail if there's a tab character in the incoming data
3143 (there should never be unless someone is messing about, as it's supposed
3144 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3145 if the connection is using TLS or if the remote IP is the same as the
3146 local IP, and the "valid-client-cert option" if a client certificate has
3149 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3150 authenticators. This can be used for authorization after authentication
3151 succeeds. (In the case of plaintext, it servers for both authentication
3154 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3155 if any retry times were supplied.
3157 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3158 connection, where there is no client IP address to check. In this
3159 situation, the verify now always succeeds.
3161 PH/25 Applied John Jetmore's -Mset patch.
3163 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3165 PH/27 In a string expansion for a processed (not raw) header when multiple
3166 headers of the same name were present, leading whitespace was being
3167 removed from all of them, but trailing whitespace was being removed only
3168 from the last one. Now trailing whitespace is removed from each header
3169 before concatenation. Completely empty headers in a concatenation (as
3170 before) are ignored.
3172 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3173 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3175 PH/29 [Removed. This was a change that I later backed out, and forgot to
3176 correct the ChangeLog entry (that I had efficiently created) before
3177 committing the later change.]
3179 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3180 address errors (4xx response to RCPT) over the same connection as other
3181 messages routed to the same hosts. Such deliveries are always "forced",
3182 so retry times are not inspected. This resulted in far too many retries
3183 for the affected addresses. The effect occurred only when there were more
3184 hosts than the hosts_max_try setting in the smtp transport when it had
3185 the 4xx errors. Those hosts that it had tried were not added to the list
3186 of hosts for which the message was waiting, so if all were tried, there
3187 was no problem. Two fixes have been applied:
3189 (i) If there are any address or message errors in an SMTP delivery, none
3190 of the hosts (tried or untried) are now added to the list of hosts
3191 for which the message is waiting, so the message should not be a
3192 candidate for sending over the same connection that was used for a
3193 successful delivery of some other message. This seems entirely
3194 reasonable: after all the message is NOT "waiting for some host".
3195 This is so "obvious" that I'm not sure why it wasn't done
3196 previously. Hope I haven't missed anything, but it can't do any
3197 harm, as the worst effect is to miss an optimization.
3199 (ii) If, despite (i), such a delivery is accidentally attempted, the
3200 routing retry time is respected, so at least it doesn't keep
3201 hammering the server.
3203 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3204 in ${readsocket because some servers need this prod.
3206 PH/32 Added some extra debug output when updating a wait-xxx database.
3208 PH/33 The hint "could be header name not terminated by colon", which has been
3209 given for certain expansion errors for a long time, was not being given
3210 for the ${if def:h_colon_omitted{... case.
3212 PH/34 The spec says: "With one important exception, whenever a domain list is
3213 being scanned, $domain contains the subject domain." There was at least
3214 one case where this was not true.
3216 PH/35 The error "getsockname() failed: connection reset by peer" was being
3217 written to the panic log as well as the main log, but it isn't really
3218 panic-worthy as it just means the connection died rather early on. I have
3219 removed the panic log writing for the ECONNRESET error when getsockname()
3222 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3223 runs only) independently of the message's sender address. This meant
3224 that, if the 4xx error was in fact related to the sender, a different
3225 message to the same recipient with a different sender could confuse
3226 things. In particular, this can happen when sending to a greylisting
3227 server, but other circumstances could also provoke similar problems.
3228 I have changed the default so that the retry time for these errors is now
3229 based a combination of the sender and recipient addresses. This change
3230 can be overridden by setting address_retry_include_sender=false in the
3233 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3234 remote server are returned as part of bounce messages. This was not
3235 happening for LMTP over a pipe (the lmtp transport), but now it is the
3236 same for both kinds of LMTP.
3238 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3239 in header lines that were in fact CNAMEs. This is no longer the case.
3241 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3242 and queue runs started by the daemon processed all messages. This has
3243 been fixed so that -R and -S can now usefully be given with -q<time>.
3245 PH/40 Import PCRE release 6.7 (fixes some bugs).
3247 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3249 PH/42 Give an error if -q is specified more than once.
3251 PH/43 Renamed the variables $interface_address and $interface_port as
3252 $received_ip_address and $received_port, to make it clear that these
3253 values apply to message reception, and not to the outgoing interface when
3254 a message is delivered. (The old names remain recognized, of course.)
3256 PH/44 There was no timeout on the connect() call when using a Unix domain
3257 socket in the ${readsocket expansion. There now is.
3259 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3260 be meaningful with "accept".
3262 SC/01 Eximstats V1.43
3263 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3265 SC/02 Eximstats V1.44
3266 Use a glob alias rather than an array ref in the generated
3267 parser. This improves both readability and performance.
3269 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3270 Collect SpamAssassin and rejection statistics.
3271 Don't display local sender or destination tables unless
3272 there is data to show.
3273 Added average volumes into the top table text output.
3275 SC/04 Eximstats V1.46
3276 Collect data on the number of addresses (recipients)
3277 as well as the number of messages.
3279 SC/05 Eximstats V1.47
3280 Added 'Message too big' to the list of mail rejection
3281 reasons (thanks to Marco Gaiarin).
3283 SC/06 Eximstats V1.48
3284 Mainlog lines which have GMT offsets and are too short to
3285 have a flag are now skipped.
3287 SC/07 Eximstats V1.49 (Alain Williams)
3288 Added the -emptyok flag.
3290 SC/08 Eximstats V1.50
3291 Fixes for obtaining the IP address from reject messages.
3293 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3294 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3295 whitespace changes from 4.64-PH/27
3297 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3300 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3301 are found, allow negative numbers in numeric criteria)
3303 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3305 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3306 to match changes made in 4.64-PH/43
3308 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3310 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3311 "domain" is actually an IP address, and force a failure. This locks out
3312 those revolvers/nameservers that support "A-for-A" lookups, in
3313 contravention of the specifications.
3315 PH/48 When a host name was looked up from an IP address, and the subsequent
3316 forward lookup of the name timed out, the host name was left in
3317 $sender_host_name, contrary to the specification.
3319 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3320 restricted to single-key lookups, Exim was not diagnosing an error if
3321 * or *@ was used with a query-style lookup.
3323 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3325 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3326 long ago, but noone interested enough thought of it. Let's just say that
3327 the "1.1" means that there are some new functions that weren't there at
3328 some point in the past.
3330 PH/51 Error processing for expansion failure of helo_data from an smtp
3331 transport during callout processing was broken.
3333 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3334 tested/used via the -bh/-bhc/-bs options.
3336 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3337 bug, fixed in subsequent PCRE releases).
3339 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3340 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3342 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3348 SC/01 Use a glob alias rather than an array ref in eximstats generated
3349 parser. This improves both readability and performance.
3351 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3352 Don't display local sender or destination tables in eximstats unless
3353 there is data to show.
3354 Added average volumes into the eximstats top table text output.
3356 SC/03 Collect data on the number of addresses (recipients) as well
3357 as the number of messages in eximstats.
3359 TF/01 Correct an error in the documentation for the redirect router. Exim
3360 does (usually) call initgroups() when daemonizing.
3362 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3363 with consistent privilege compared to when running as a daemon.
3365 TF/03 Note in the spec that $authenticated_id is not set for local
3366 submissions from trusted users.
3368 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3369 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3371 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3372 by adding some example configuration directives to the default
3373 configuration file. A little bit of work is required to uncomment the
3374 directives and define how usernames and passwords are checked, but
3375 there is now a framework to start from.
3377 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3378 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3379 without this. I don't know how relevant this is to other LDAP libraries.
3381 PH/02 Add the verb name to the "unknown ACL verb" error.
3383 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3385 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3387 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3388 directory not expanded when it should be if an expanded home directory
3389 was set for the address (which is overridden by the transport).
3391 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3394 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3395 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3396 because it is too late at that time, and has no effect.
3398 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3399 security issue with \' (bugzilla #107). I could not use the
3400 PQescapeStringConn() function, because it needs a PGconn value as one of
3403 PH/08 When testing addresses using -bt, indicate those final addresses that
3404 are duplicates that would not cause an additional delivery. At least one
3405 person was confused, thinking that -bt output corresponded to deliveries.
3406 (Suppressing duplicates isn't a good idea as you lose the information
3407 about possibly different redirections that led to the duplicates.)
3409 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3410 systems where poll() doesn't work, in particular OS X.
3412 PH/10 Added more information to debugging output for retry time not reached.
3414 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3415 operations in malware.c.
3417 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3420 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3421 syslog_duplication set false, log lines that would normally be written
3422 both the the main log and to the reject log were not written to syslog at
3425 PH/14 In the default configuration, change the use of "message" in ACL warn
3426 statements to "add_header".
3428 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3429 not followed by a command (e.g. "seen endif").
3431 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3432 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3435 PH/17 Added extra conditions to the default value of delay_warning_condition
3439 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3440 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3441 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3444 The Auto-Submitted: and various List- headers are standardised, whereas I
3445 don't think Precedence: ever was.
3447 PH/18 Refactored debugging code in route_finduser() to show more information,
3448 in particular, the error code if getpwnam() issues one.
3450 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3451 This is apparently needed in addition to the PH/07 change above to avoid
3452 any possible encoding problems.
3454 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3455 but not after initializing Perl.
3457 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3458 output them only if debugging. By default they are written stderr,
3459 apparently, which is not desirable.
3461 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3464 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3467 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3469 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3470 authenticated or an ident call has been made. Suppress the default
3471 values for $authenticated_id and $authenticated_sender (but permit -oMai
3472 and -oMas) when testing with -bh.
3474 PH/24 Re-jigged the order of the tests in the default configuration so that the
3475 tests for valid domains and recipients precede the DNS black list and CSA
3476 tests, on the grounds that those ones are more expensive.
3478 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3479 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3480 This bug exists in every version of Exim that I still have, right back to
3483 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3484 However, an attempt to turn on case-sensitivity in a regex key by
3485 including (?-i) didn't work because the subject string was already
3486 lowercased, and the effects were non-intuitive. It turns out that a
3487 one-line patch can be used to allow (?-i) to work as expected.
3493 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3494 other effects) broke the use of negated acl sub-conditions.
3496 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3499 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3500 "Deny" causes Exim to reject the incoming connection with a 554 error.
3501 Unfortunately, if there is a major crisis, such as a disk failure,
3502 tcp-wrappers gives "deny", whereas what one would like would be some
3503 kind of temporary error. A kludge has been added to help with this.
3504 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3505 554 error is used if errno is still zero or contains ENOENT (which occurs
3506 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3509 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3511 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3512 errors. Otherwise a message that provokes a temporary error (when other
3513 messages do not) can cause a whole host to time out.
3515 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3516 addresses were routed directly to files or pipes from a redirect router.
3517 File deliveries just didn't batch; pipe deliveries might have suffered
3520 PH/06 A failure to get a lock for a hints database would erroneously always say
3521 "Failed to get write lock", even when it was really a read lock.
3523 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3524 of 0600. This has been changed to use the value of the lockfile_mode
3525 option (which defaults to 0600).
3527 PH/08 Applied small patch from the Sieve maintainer.
3529 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3530 folder from quota calculations, a direct delivery into this folder messed
3531 up the contents of the maildirsize file. This was because the regex was
3532 used only to exclude .Trash (or whatever) when the size of the mailbox
3533 was calculated. There was no check that a delivery was happening into an
3534 excluded directory. This bug has been fixed by ignoring all quota
3535 processing for deliveries into excluded directories.
3537 PH/10 Added the maildirfolder_create_regex option to appendfile.
3543 PH/01 The code for finding all the local interface addresses on a FreeBSD
3544 system running IPv6 was broken. This may well have applied to all BSD
3545 systems, as well as to others that have similar system calls. The broken
3546 code found IPv4 interfaces correctly, but gave incorrect values for the
3547 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3548 that it would not match correctly against @[] and not recognize the IPv6
3551 PH/02 The ipliteral router was not recognizing addresses of the form user@
3552 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3554 PH/03 Added disable_ipv6.
3556 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3557 decoded form, because it is most often used to construct To: headers
3558 lines in autoreplies, and the decoded form may well be syntactically
3559 invalid. However, $reply_address has leading white space removed, and all
3560 newlines turned into spaces so that the autoreply transport does not
3563 PH/05 If group was specified without a user on a router, and no group or user
3564 was specified on a transport, the group from the router was ignored.
3566 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3567 for visible compile-time settings that can be used to change these
3568 numbers, for those that want even more. Backwards compatibility with old
3569 spool files has been maintained. However, going back to a previous Exim
3570 release will lost any variables that are in spool files.
3572 PH/07 Two small changes when running in the test harness: increase delay when
3573 passing a TCP/IP connection to a new process, in case the original
3574 process has to generate a bounce, and remove special handling of
3575 127.0.0.2 (sic), which is no longer necessary.
3577 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3578 be the same on different OS.
3580 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3583 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3584 whether --show-vars was specified or not
3586 JJ/02 exipick: Added support for new ACL variable spool format introduced
3589 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3590 syntactically invalid From: or Reply-to: line, and a filter used this to
3591 generate an autoreply, and therefore failed to obtain an address for the
3592 autoreply, Exim could try to deliver to a non-existent relative file
3593 name, causing unrelated and misleading errors. What now happens is that
3594 it logs this as a hard delivery error, but does not attempt to create a
3597 PH/11 The exinext utility has a -C option for testing purposes, but although
3598 the given file was scanned by exinext itself; it wasn't being passed on
3599 when Exim was called.
3601 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3602 an end-of-file indication when reading a command response.
3604 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3605 compiled. In many other places in Exim, IPv6 addresses are always
3606 recognized, so I have changed this. It also means that IPv4 domain
3607 literals of the form [IPV4:n.n.n.n] are now always recognized.
3609 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3610 used if the router is not running as root, for example, when verifying at
3611 ACL time, or when using -bh. The debugging output from this situation was
3612 non-existent - all you got was a failure to exec. I have made two
3615 (a) Failures to set uid/gid, the current directory, or a process leader
3616 in a subprocess such as that created by queryprogram now generate
3617 suitable debugging output when -d is set.
3619 (b) The queryprogram router detects when it is not running as root,
3620 outputs suitable debugging information if -d is set, and then runs
3621 the subprocess without attempting to change uid/gid.
3623 PH/15 Minor change to Makefile for building test_host (undocumented testing
3626 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3627 additional section of a DNS packet that returns MX or SRV records.
3628 Instead, it always explicitly searches for A/AAAA records. This avoids
3629 major problems that occur when a DNS server includes only records of one
3630 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3631 fixed another bug: if SRV records were looked up and the corresponding
3632 address records were *not* found in the additional section, the port
3633 values from the SRV records were lost.
3635 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3636 using the correct key (the original address) when searching the retry
3637 rules in order to find which one to use for generating the retry hint.
3639 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3640 adding the default one. Similarly, if it contains a Reply-To: header, the
3641 errors_reply_to option, if set, is not used.
3643 PH/19 When calculating a retry time, Exim used to measure the "time since
3644 failure" by looking at the "first failed" field in the retry record. Now
3645 it does not use this if it is later than than the arrival time of the
3646 message. Instead it uses the arrival time. This makes for better
3647 behaviour in cases where some deliveries succeed, thus re-setting the
3648 "first failed" field. An example is a quota failure for a huge message
3649 when small messages continue to be delivered. Without this change, the
3650 "time since failure" will always be short, possible causing more frequent
3651 delivery attempts for the huge message than are intended.
3652 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3654 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3655 $1, $2, $3) because the numerical variables can be reset during some
3656 expansion items (e.g. "match"), thereby losing the authentication data.
3658 PH/21 Make -bV show the size of off_t variables so that the test suite can
3659 decide whether to run tests for quotas > 2G.
3661 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3662 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3663 filecount value is greater than 2G or if a quota value is greater than 2G
3664 on a system where the size of off_t is not greater than 4, a panic error
3667 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3668 never match. The debug and -bh output now contains an explicit error
3669 message indicating a malformed IPv4 address or mask.
3671 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3672 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3673 PH/23 above applies.
3675 PH/25 Do not write to syslog when running in the test harness. The only
3676 occasion when this arises is a failure to open the main or panic logs
3677 (for which there is an explicit test).
3679 PH/26 Added the /no_tell option to "control=freeze".
3681 PH/27 If a host name lookup failed very early in a connection, for example, if
3682 the IP address matched host_lookup and the reverse lookup yielded a name
3683 that did not have a forward lookup, an error message of the form "no IP
3684 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3685 could be logged. Now it outputs the IP address instead of "NULL".
3687 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3688 allows the sender and the authenticated sender to be set when
3689 submitting a message from within Exim. Since child_open_exim() is
3690 documented for local_scan(), the new function should be too.
3692 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3693 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3694 results in an empty string is now treated as unset.
3696 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3698 PH/31 Added sender_verify_fail logging option.
3700 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3701 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3702 tidied the source and removed it altogether.
3704 PH/33 When a queue run was abandoned because the load average was too high, a
3705 log line was always written; now it is written only if the queue_run log
3706 selector is set. In addition, the log line for abandonment now contains
3707 information about the queue run such as the pid. This is always present
3708 in "start" and "stop" lines but was omitted from the "abandon" line.
3710 PH/34 Omit spaces between a header name and the colon in the error message that
3711 is given when verify = headers_syntax fails (if there are lots of them,
3712 the message gets confusing).
3714 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3715 names, as there are now some PTR records that contain slashes. This check
3716 is only to protect against broken name servers that fall over on strange
3717 characters, so the fact that it applies to all lookups doesn't matter.
3719 PH/36 Now that the new test suite is complete, we can remove some of the
3720 special code in Exim that was needed for the old test suite. For example,
3721 sorting DNS records because real resolvers return them in an arbitrary
3722 order. The new test suite's fake resolver always returns records in the
3725 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3726 bounces) except when queue_only is set, to avoid logging races between
3727 the different processes.
3729 PH/38 Panic-die if .include specifies a non-absolute path.
3731 PH/39 A tweak to the "H" retry rule from its user.
3733 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3734 a label. They prevented compilation on older perls.
3736 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3737 a warning to be raised on newish perls.
3739 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3740 on queue. Changes to match documented behaviour of showing count of
3741 messages matching specified criteria.
3743 PH/40 Changed the default ident timeout from 30s to 5s.
3745 PH/41 Added support for the use of login_cap features, on those BSD systems
3746 that have them, for controlling the resources used by pipe deliveries.
3748 PH/42 The content-scanning code uses fopen() to create files in which to put
3749 message data. Previously it was not paying any attention to the mode of
3750 the files. Exim runs with umask(0) because the rest of the code creates
3751 files with open(), and sets the required mode explicitly. Thus, these
3752 files were ending up world-writeable. This was not a big issue, because,
3753 being within the spool directory, they were not world-accessible. I have
3754 created a function called modefopen, which takes an additional mode
3755 argument. It sets umask(777), creates the file, chmods it to the required
3756 mode, then resets the umask. All the relevant calls to fopen() in the
3757 content scanning code have been changed to use this function.
3759 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3760 to 24 hours. This avoids potential overflow problems when processing G
3761 and H retry rules. I suspect nobody ever tinkers with this value.
3763 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3765 PH/45 When the plaintext authenticator is running as a client, the server's
3766 challenges are checked to ensure they are valid base64 strings. By
3767 default, the authentication attempt is cancelled if an invalid string is
3768 received. Setting client_ignore_invalid_base64 true ignores these errors.
3769 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3770 they are received. Thus, the responses can be made to depend on the
3771 challenges. If an invalid string is ignored, an empty string is placed in
3774 PH/46 Messages that are created by the autoreply transport now contains a
3775 References: header, in accordance with RFCs 2822 and 3834.
3777 PH/47 Added authenticated_sender_force to the smtp transport.
3779 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3781 PH/49 Installed latest patch from the Sieve maintainer.
3783 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3784 was also set, the appendfile transport was unnecessarily scanning a
3785 directory of message files (e.g. for maildir delivery) to find the count
3786 of files (along with the size), even though it did not need this
3787 information. It now does the scan only if it needs to find either the
3788 size of the count of files.
3790 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3792 PH/52 Two bugs concerned with error handling when the smtp transport is
3795 (i) Exim was not creating retry information for temporary errors given
3796 for individual recipients after the DATA command when the smtp transport
3797 was used in LMTP mode. This meant that they could be retried too
3798 frequently, and not timed out correctly.
3800 (ii) Exim was setting the flag that allows error details to be returned
3801 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3802 recipients that were returned after the DATA command.
3804 PH/53 This is related to PH/52, but is more general: for any failing address,
3805 when detailed error information was permitted to be returned to the
3806 sender, but the error was temporary, then after the final timeout, only
3807 "retry timeout exceeded" was returned. Now it returns the full error as
3808 well as "retry timeout exceeded".
3810 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3811 do this, and (what is worse) MTAs that accept it.
3813 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3814 will now be deprecated.
3816 PH/56 New os.c-cygwin from the Cygwin maintainer.
3818 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3819 formats (previously only available in exim formats via -bpr, -bpru,
3820 and -bpra. Now also available in native and exiqgrep formats)
3822 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3823 with very large, slow to parse queues
3825 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3827 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3829 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3830 responses to authentication challenges, though it was showing the
3831 challenges; (ii) I've removed the CR characters from the debug output for
3834 PH/58 Allow for the insertion of a newline as well as a space when a string
3835 is turned into more than one encoded-word during RFC 2047 encoding. The
3836 Sieve code now uses this.
3838 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3839 data_4xx, lost_connection, tls_required.
3841 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3842 message was being sent as an SMTP response.
3844 PH/61 Add -l and -k options to exicyclog.
3846 PH/62 When verifying, if an address was redirected to one new address, so that
3847 verification continued, and the new address failed or deferred after
3848 having set something in $address_data, the value of $address_data was not
3849 passed back to the ACL. This was different to the case when no
3850 redirection occurred. The value is now passed back in both cases.
3852 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3853 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3854 use login_cap.h, so on its own it isn't the distinguishing feature. The
3855 new name refers directly to the setclassresources() function.
3857 PH/65 Added configuration files for NetBSD3.
3859 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3861 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3862 is preferred over IPv4.
3864 PH/68 The bounce_return_message and bounce_return_body options were not being
3865 honoured for bounces generated during the reception of non-SMTP messages.
3866 In particular, this applied to messages rejected by the ACL. This bug has
3867 been fixed. However, if bounce_return_message is true and bounce_return_
3868 body is false, the headers that are returned for a non-SMTP message
3869 include only those that have been read before the error was detected.
3870 (In the case of an ACL rejection, they have all been read.)
3872 PH/69 The HTML version of the specification is now built in a directory called
3873 spec_html instead of spec.html, because the latter looks like a path with
3874 a MIME-type, and this confuses some software.
3876 PH/70 Catch two compiler warnings in sieve.c.
3878 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3879 function verify_get_ident() calls ip_connect() to connect a socket, but
3880 if the "connect()" function timed out, ip_connect() used to close the
3881 socket. However, verify_get_ident() also closes the socket later, and in
3882 between Exim writes to the log, which may get opened at this point. When
3883 the socket was closed in ip_connect(), the log could get the same file
3884 descriptor number as the socket. This naturally causes chaos. The fix is
3885 not to close the socket in ip_connect(); the socket should be closed by
3886 the function that creates it. There was only one place in the code where
3887 this was missing, in the iplookup router, which I don't think anybody now
3888 uses, but I've fixed it anyway.
3890 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3891 well as to direct DNS lookups. Otherwise the handling of names in host
3892 lists is inconsistent and therefore confusing.
3898 PH/01 Two changes to the default runtime configuration:
3900 (1) Move the checks for relay_from_hosts and authenticated clients from
3901 after to before the (commented out) DNS black list checks.
3903 (2) Add control=submission to the relay_from_hosts and authenticated
3904 clients checks, on the grounds that messages accepted by these
3905 statements are most likely to be submissions.
3907 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3909 (1) Generate an error if the third argument for the ${prvs expansion is
3912 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3915 (3) Reset the variables that are obtained from the first argument of
3916 ${prvscheck and used in the second argument before leaving the code,
3917 because their memory is reclaimed, so using them afterwards may do
3920 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3921 one (it's much easier than Tom thought :-).
3923 (5) Because of (4), we can now allow for the use of $prvscheck_result
3924 inside the third argument.
3926 PH/03 For some reason, the default setting of PATH when running a command from
3927 a pipe transport was just "/usr/bin". I have changed it to
3930 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3931 anything to be listed in the output from -bV.
3933 PH/05 When a filter generated an autoreply, the entire To: header line was
3934 quoted in the delivery log line, like this:
3936 => >A.N.Other <ano@some.domain> <original@ddress> ...
3938 This has been changed so that it extracts the operative address. There
3939 may be more than one such address. If so, they are comma-separated, like
3942 => >ano@some.domain,ona@other.domain <original@ddress> ...
3944 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3945 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3946 not being looked up in the rDNS to get a host name, Exim was showing the
3947 IP address twice in Received: lines, even though the IP addresses were
3948 identical. For example:
3950 Received: from [1.2.3.4] (helo=[1.2.3.4])
3952 However, if the real host name was known, it was omitting the HELO data
3953 if it matched the actual IP address. This has been tidied up so that it
3954 doesn't show the same IP address twice.
3956 PH/07 When both +timestamp and +memory debugging was on, the value given by
3957 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3958 called by the debug printing, thereby overwriting the timestamp buffer.
3959 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3961 PH/08 When the original message was included in an autoreply transport, it
3962 always said "this is a copy of the message, including all the headers",
3963 even if body_only or headers_only was set. It now gives an appropriate
3966 PH/09 Applied a patch from the Sieve maintainer which:
3968 o fixes some comments
3969 o adds the (disabled) notify extension core
3970 o adds some debug output for the result of if/elsif tests
3971 o points to the current vacation draft in the documentation
3972 and documents the missing references header update
3976 o fixes a bug in processing the envelope test (when testing
3977 multiple envelope elements, the last element determined the
3980 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3981 Electronic Mail") by including:
3983 Auto-submitted: auto-generated
3985 in the messages that it generates (bounce messages and others, such as
3986 warnings). In the case of bounce messages for non-SMTP messages, there was
3987 also a typo: it was using "Auto_submitted" (underscore instead of
3988 hyphen). Since every message generated by Exim is necessarily in response
3989 to another message, thes have all been changed to:
3991 Auto-Submitted: auto-replied
3993 in accordance with these statements in the RFC:
3995 The auto-replied keyword:
3997 - SHOULD be used on messages sent in direct response to another
3998 message by an automatic process,
4000 - MUST NOT be used on manually-generated messages,
4002 - MAY be used on Delivery Status Notifications (DSNs) and Message
4003 Disposition Notifications (MDNs),
4005 - MUST NOT be used on messages generated by automatic or periodic
4006 processes, except for messages which are automatic responses to
4009 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4010 to the default Received: header definition.
4012 PH/12 Added log selector acl_warn_skipped (default on).
4014 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4015 variables because (a) they are in the wrong storage pool and (b) even if
4016 they were copied, it wouldn't work properly because of the caching.
4018 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4019 checking when decoding. Apparently there are clients that generate
4020 overlong encoded strings. Why am I not surprised?
4022 PH/15 If the first argument of "${if match_address" was not empty, but did not
4023 contain an "@" character, Exim crashed. Now it writes a panic log message
4024 and treats the condition as false.
4026 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4028 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4029 extension "envelope-auth". The code is finished and in agreement with
4030 other implementations, but there is no documentation so far and in fact,
4031 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4032 not changing the active code.
4034 Print executed "if" and "elsif" statements when debugging is used. This
4035 helps a great deal to understand what a filter does.
4037 Document more things not specified clearly in RFC3028. I had all this
4038 sorted out, when out of a sudden new issues came to my mind. Oops."
4040 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4043 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4044 canonical form (as documented). However, after a host name lookup from
4045 the IP address, check_host() was doing a simple string comparison with
4046 addresses acquired from the DNS when checking that the found name did
4047 have the original IP as one of its addresses. Since any found IPv6
4048 addresses are likely to be in abbreviated form, the comparison could
4049 fail. Luckily, there already exists a function for doing the comparison
4050 by converting both addresses to binary, so now that is used instead of
4051 the text comparison.
4053 PH/20 There was another similar case to PH/19, when a complete host name was
4054 given in a host list; looking up its IP address could give an abbreviated
4055 form, whereas the current host's name might or might not be abbreviated.
4056 The same fix has been applied.
4062 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4063 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4066 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4067 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4069 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4071 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4072 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4073 Also, for "Auto-Submitted", treat anything other than "no" as
4074 non-personal, in accordance with RFC 3834. (Previously it treated
4075 anything starting "auto-" as non-personal.)
4077 TF/01 The control=submission/name=... option had a problem with syntax
4078 errors if the name included a slash character. The /name= option
4079 now slurps the rest of the string, so it can include any characters
4080 but it must come last in the list of options (after /sender_retain
4083 PH/05 Some modifications to the interface to the fake nameserver for the new
4091 TK/01 Added the "success_on_redirect" address verification option. See
4092 NewStuff for rationale and an example.
4094 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4096 PH/02 Patch to exigrep to allow it to work on syslog lines.
4098 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4099 fread() to skip over the body file's header line, because in Cygwin the
4100 header line is locked and is inaccessible.
4102 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4103 co-exist for some time) to make it clear that it is the Exim ID that is
4104 referenced, not the Message-ID: header line.
4106 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4107 string_format() function, because snprintf() does not exist on all
4110 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4111 ${stat: expansion item.
4113 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4114 protocol synchronization error", to keep the pedants happy.
4116 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4117 well as for IRIX systems, when gcc is being used. See the host.c source
4120 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4122 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4125 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4126 transport and to the smtp transport in LMTP mode.
4128 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4130 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4131 run a filter in a subprocess. This could lead to confusion in subsequent
4132 lookups in the parent process. There should also be a search_tidyup() at
4133 the end of the subprocess.
4135 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4136 only if the host matched helo_try_verify_hosts, which caused the
4137 verification to occur when the EHLO/HELO command was issued. The ACL just
4138 tested the remembered result. Now, if a previous verification attempt has
4139 not happened, "verify = helo" does it there and then.
4141 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4143 TK/03 Fix log output including CR from clamd.
4145 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4146 exist provoked a memory error which could cause a segfault.
4148 PH/15 Installed PCRE 6.2
4150 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4152 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4153 of the problem. Specifically, suggested +O2 rather than +O1 for the
4156 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4158 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4159 record, the port was not correctly logged when the outgoing_port log
4160 selector was set (it logged the transort's default port).
4162 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4163 fallback_hosts, and "hosts" in the smtp transport.
4165 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4166 host errors such as "Connection refused".
4168 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4169 authentication with radiusclient 0.4.9:
4171 - Error returned from rc_read_config was caught wrongly
4172 - Username/password not passed on to radius server due to wrong length.
4174 The presumption is that some radiusclient API changes for 4.51/PH/17
4175 were not taken care of correctly. The code is still untested by me (my
4176 Linux distribution still has 0.3.2 of radiusclient), but it was
4177 contributed by a Radius user.
4179 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4180 expanding the "port" option of the smtp transport.
4182 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4183 while reading a MIME header. Thanks to Tom Hughes for a patch.
4185 PH/24 Include config.h inside local_scan.h so that configuration settings are
4188 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4189 that in an ACL for RCPT (for example), you can examine exactly what was
4192 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4193 commands, but it was not correctly comparing the address with the actual
4194 client host address. Thus, it would show the EHLO address in Received:
4195 header lines when this was not necessary.
4197 PH/27 Added the % operator to ${eval:}.
4199 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4200 it should be ignoring failures (because with -C, for example, it has lost
4201 privilege). It wasn't ignoring creation failures other than "already
4204 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4207 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4208 because an input line was too long, either on its own, or by virtue of
4209 too many continuations, the temporary file was not being removed, and the
4210 return code was incorrect.
4212 PH/31 Missing "BOOL" in function definition in filtertest.c.
4214 PH/32 Applied Sieve patches from the maintainer.
4216 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4218 PH/33 Added "verify = not_blind".
4220 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4221 Local/Makefile (with some defaults set). These are used in built scripts
4222 such as exicyclog, but they have never been used in the exim_install
4223 script (though there are many overriding facilities there). I have
4224 arranged that the exim_install script now takes note of these two
4227 PH/35 Installed configuration files for Dragonfly.
4229 PH/36 When a locally submitted message by a trusted user did not contain a
4230 From: header, and the sender address was obtained from -f or from an SMTP
4231 MAIL command, and the trusted user did not use -F to supply a sender
4232 name, $originator_name was incorrectly used when constructing a From:
4233 header. Furthermore, $originator_name was used for submission mode
4234 messages from external hosts without From: headers in a similar way,
4235 which is clearly wrong.
4237 PH/37 Added control=suppress_local_fixups.
4239 PH/38 When log_selector = +received_sender was set, and the addition of the
4240 sender made the log line's construction buffer exactly full, or one byte
4241 less than full, an overflow happened when the terminating "\n" was
4244 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4245 when the result of a list match is failure because a DNS lookup failed.
4247 PH/40 RM_COMMAND is now used in the building process.
4249 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4250 the "build-* directories that it finds.
4252 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4253 address in a domain literal was a prefix of an interface address.
4255 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4256 when verifying a sender address, unless rewrite_headers is false.
4258 PH/44 Wrote a long comment about why errors_to addresses are verified as
4259 recipients, not senders.
4261 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4262 the ratelimit ACL was added.
4264 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4266 PH/46 Added extra information about PostgreSQL errors to the error string.
4268 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4269 avoiding the need to install special zones in a real server. This is
4270 backwards compatible; if it can't find the fake resolver, it drops back.
4271 Thus, both old and new test suites can be run.
4273 TF/02 Added util/ratelimit.pl
4275 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4276 clock is set back in time.
4278 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4279 Candler <B.Candler@pobox.com>.
4281 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4282 OK for addresses that are the result of redirections.
4284 PH/48 A number of further additions for the benefit of the new test suite,
4285 including a fake gethostbyname() that interfaces to the fake DNS resolver
4288 TF/06 The fix for widen_domains has also been applied to qualify_single and
4289 search_parents which are the other dnslookup options that can cause
4292 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4295 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4297 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4298 service" if it encountered an MX record with an empty target hostname.
4299 The message is now "an MX or SRV record indicated no SMTP service".
4301 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4302 if the DNS of the sending site is misconfigured. This is quite a
4303 common situation. This change restores the behaviour of treating a
4304 helo verification defer as a failure.
4306 PH/51 If self=fail was set on a router, the bounce message did not include the
4307 actual error message.
4313 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4315 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4316 command itself ended in error, the underlying message about the transport
4317 filter timeout was being overwritten with the pipe command error. Now the
4318 underlying error message should be appended to the second error message.
4320 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4322 PH/02 Reduce the amount of output that "make" produces by default. Full output
4323 can still be requested.
4325 PH/03 The warning log line about a condition test deferring for a "warn" verb
4326 was being output only once per connection, rather than after each
4327 occurrence (because it was using the same function as for successful
4328 "warn" verbs). This seems wrong, so I have changed it.
4330 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4331 it should not have, which might have caused a crash in the right
4332 circumstances, but probably never did.
4334 PH/04 Installed a modified version of Tony Finch's patch to make submission
4335 mode fix the return path as well as the Sender: header line, and to
4336 add a /name= option so that you can make the user's friendly name appear
4339 TF/03 Added the control = fakedefer ACL modifier.
4341 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4342 Mark Lowes for thorough testing.
4344 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4346 TK/03 Merged latest SRS patch from Miles Wilton.
4348 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4349 with the definition in sysexits.h (which is #included earlier).
4350 Fortunately, Exim does not actually use EX_OK. The code used to try to
4351 preserve the sysexits.h value, by assuming that macro definitions were
4352 scanned for macro replacements. I have been disabused of this notion,
4353 so now the code just undefines EX_OK before #including unistd.h.
4355 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4356 in the smtp transport. When a block could not be written in a single
4357 write() function, the timeout was being re-applied to each part-write.
4358 This seems wrong - if the receiver was accepting one byte at a time it
4359 would take for ever. The timeout is now adjusted when this happens. It
4360 doesn't have to be particularly precise.
4362 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4363 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4365 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4366 without a domain if the check to <postmaster@domain> fails.
4368 SC/01 Eximstats: added -xls and the ability to specify output files
4369 (patch written by Frank Heydlauf).
4371 SC/02 Eximstats: use FileHandles for outputting results.
4373 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4375 SC/04 Eximstats: fixed display of large numbers with -nvr option
4377 SC/05 Eximstats: fixed merging of reports with empty tables.
4379 SC/06 Eximstats: added the -include_original_destination flag
4381 SC/07 Eximstats: removed tabs and trailing whitespace.
4383 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4385 TK/06 MBOX spool code: Add real "From " MBOX separator line
4386 so the .eml file is really in mbox format (even though
4387 most programs do not really care). Patch from Alex Miller.
4389 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4390 The latter is generated from $received_to and is only set if the
4391 message has one envelope recipient. SA can use these headers,
4392 obviously out-of-the-box. Patch from Alex Miller.
4394 PH/08 The ${def test on a variable was returning false if the variable's
4395 value was "0", contrary to what the specification has always said!
4396 The result should be true unless the variable is empty.
4398 PH/09 The syntax error of a character other than { following "${if
4399 def:variable_name" (after optional whitespace) was not being diagnosed.
4400 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4401 accidental colon was present, for example, could give incorrect results.
4403 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4404 result is used (not including appendfile, where other changes are about
4407 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4408 This involved changing a lot of size variables from int to off_t. It
4409 should work with maildirs and everything.
4411 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4412 spamd dying while we are connected to it.
4414 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4417 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4418 function for BDB 4.3.
4420 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4422 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4423 the sources for PCRE have been re-arranged and more files are now
4426 PH/15 The code I had for printing potentially long long variables in PH/11
4427 above was not the best (it lost precision). The length of off_t variables
4428 is now inspected at build time, and an appropriate printing format (%ld
4429 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4430 to be "long long int" or "long int". This is needed for the internal
4431 formatting function string_vformat().
4433 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4434 the configuration file to be ":syslog", then the script "guesses" where
4435 the logs files are, rather than using the compiled in default. In our
4436 case the guess is not the same as the compiled default, so the script
4437 suddenly stopped working when I started to use syslog. The patch checks
4438 to see if log_file_path is "". If so, it attempts to read it from exim
4439 with no configuration file to get the compiled in version, before it
4440 falls back to the previous guessing code."
4442 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4443 implementing BATV in an Exim configuration. See NewStuff for the gory
4446 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4447 Makefile that are specific to HP-UX.
4449 PH/18 If the "use_postmaster" option was set for a recipient callout together
4450 with the "random" option, the postmaster address was used as the MAIL
4451 FROM address for the random test, but not for the subsequent recipient
4452 test. It is now used for both.
4454 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4455 patch removes a few documentation additions to RFC 3028, because the
4456 latest draft now contains them. It adds the new en;ascii-case comparator
4457 and a new error check for 8bit text in MIME parts. Comparator and
4458 require names are now matched exactly. I enabled the subaddress
4459 extension, but it is not well tested yet (read: it works for me)."
4461 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4462 rework some of the code of TK/09 above to avoid the hardwired use of
4463 "%lld" and "long long". Replaced the call to snprintf() with a call to
4466 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4467 records point to non-existent hosts", "retry timeout exceeded", and
4468 "retry time not reached for any host after a long failure period".
4470 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4471 experimental DomainKeys support:
4473 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4474 (2) On an error such as an illegally used "control", the wrong name for
4475 the control was given.
4477 These problems did NOT occur unless DomainKeys support was compiled.
4479 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4481 PH/24 Added ${if match_ip condition.
4483 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4484 databases so that it will be absolutely obvious if a crash occurs in the
4485 DB library. This is a regular occurrence (often caused by mis-matched
4488 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4489 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4490 picked up on a user's system that detects such things. There doesn't seem
4491 to be a gcc warning option for this - only an attribute that has to be
4492 put on the function's prototype. It seems that in Fedora Core 4 they have
4493 set this on a number of new functions. No doubt there will be more in due
4496 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4497 specify a transport. However, if an address that was verified by such a
4498 router was the subject of a callout, Exim crashed because it tried to
4499 read the rcpt_include_affixes from the non-existent transport. Now it
4500 just assumes that the setting of that option is false. This bug was
4501 introduced by 4.51/PH/31.
4503 PH/28 Changed -d+all to exclude +memory, because that information is very
4504 rarely of interest, but it makes the output a lot bigger. People tend to
4505 do -d+all out of habit.
4507 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4508 code in os-type was giving problems when libc.so lives in lib64, like on
4511 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4512 aren't the modern standard, and it seems that some systems' include files
4513 don't always have them. Exim was already checking for some of the newer
4514 ones like T_AAAA, and defining it itself. I've added checks for all the
4515 record types that Exim uses.
4517 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4518 not automatically generating a new one, as it is supposed to. This
4519 prevented TLS from working. If the file did exist, but contained invalid
4520 data, a new version was generated, as expected. It was only the case of a
4521 non-existent file that was broken.
4523 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4524 with a change in libdomainkeys > 0.64.
4526 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4527 from DNS. If the selector record carries the flag, it now has
4528 precedence over the domain-wide flag.
4530 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4532 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4533 the use of a port name that isn't defined in /etc/services) occurred, the
4534 message was deferred as in a normal delivery, and thus remained on the
4535 spool, instead of being failed because of the mua_wrapper setting. This
4536 is now fixed, and I tidied up some of the mua_wrapper messages at the
4539 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4540 the messages in a hash of arrays rather than using individual hashes.
4541 This is a bit cleaner and results in dramatic memory savings, albeit
4542 at a slight CPU cost.
4544 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4545 as requested by Marc Sherman.
4547 SC/10 Eximstats: added histograms for user specified patterns as requested
4550 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4552 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4553 fopen() in the content-scanning modules that did not already have it.
4559 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4560 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4562 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4564 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4566 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4567 to test Sieve filters that use "vacation".
4569 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4570 that changes the way the GnuTLS parameters are stored in the cache file.
4571 The new format can be generated externally. For backward compatibility,
4572 if the data in the cache doesn't make sense, Exim assumes it has read an
4573 old-format file, and it generates new data and writes a new file. This
4574 means that you can't go back to an older release without removing the
4577 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4578 work if there are any delivery delays because "one_time" forces the
4579 parent to be marked "delivered", so its unseen clone is never tried
4580 again. For this reason, Exim now forbids the simultaneous setting of
4583 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4584 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4585 entry if you want to know the details. The fix, however, neglected to
4586 consider the case when local delivery batching is involved. The test for
4587 "previously delivered" was not happening when checking to see if an
4588 address could be batched with a previous (undelivered) one; under
4589 certain circumstances this could lead to multiple deliveries to the same
4592 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4593 in its include files, and this causes problems building Exim.
4595 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4596 header_syntax) but Exim was just ignoring anything given after a slash.
4597 In particular, this caused confusion with an attempt to use "verify =
4598 reverse_host_lookup/defer_ok". An error is now given when options are
4599 supplied for verify items that do not have them. (Maybe reverse_host_
4600 lookup should have a defer_ok option, but that's a different point.)
4602 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4603 defined by RFC 821) to 2048, because there were problems with some AUTH
4604 commands, and RFC 1869 says the size should be increased for extended
4605 SMTP commands that take arguments.
4607 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4610 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4611 "unknown" error; now it says that the functionality isn't in the binary.
4613 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4614 an address' error message when a string expansion fails (syntax or
4615 whatever). Otherwise the password may appear in the log. Following change
4616 PH/42 below, there is no longer a chance of it appearing in a bounce
4619 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4621 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4622 of its addresses was ever tried. (Bugzilla bug #2.)
4624 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4625 the result incorrectly in the debug output. (It correctly added a newline
4626 to what was transported.)
4628 TF/01 Added $received_time.
4630 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4631 commented out examples of how to interface to a virus scanner and to
4632 SpamAssassin. Also added commented examples of av_scanner and
4633 spamd_address settings.
4635 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4636 and controls are allowed in which ACLs. There were a couple of minor
4637 errors. Some of the entries in the conditions table (which is a table of
4638 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4639 negation of where the condition IS allowed.
4641 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4643 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4644 header file does not have a version number, so I've had to invent a new
4645 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4646 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4647 radiusclient), but it was contributed by a Radius user.
4649 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4650 files or directories, for OpenSSL.
4652 PH/19 When an Exim process that is running as root has to create an Exim log
4653 file, it does so in a subprocess that runs as exim:exim so as to get the
4654 ownership right at creation (otherwise, other Exim processes might see
4655 the file with the wrong ownership). There was no test for failure of this
4656 fork() call, which would lead to the process getting stuck as it waited
4657 for a non-existent subprocess. Forks do occasionally fail when resources
4658 run out. I reviewed all the other calls to fork(); they all seem to check
4661 PH/20 When checking for unexpected SMTP input at connect time (before writing
4662 the banner), Exim was not dealing correctly with a non-positive return
4663 from the read() function. If the client had disconnected by this time,
4664 the result was a log entry for a synchronization error with an empty
4665 string after "input=" when read() returned zero. If read() returned -1
4666 (an event I could not check), uninitialized data bytes were printed.
4667 There were reports of junk text (parts of files, etc) appearing after
4670 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4672 PH/22 Added support for macro redefinition, and (re)definition in between
4673 driver and ACL definitions.
4675 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4676 forgetting to use the resulting value; it was using the unexpanded value.
4678 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4679 hadn't been configured. The fix is from Juergen Kreileder, who
4680 understands it better than I do:
4682 "Here's what I see happening with three configured cyrus_sasl
4683 authenticators configured (plain, login, cram-md5):
4685 On startup auth_cyrus_sasl_init() gets called for each of these.
4686 This means three calls to sasl_listmech() without a specified mech_list.
4687 => SASL tests which mechs of all available mechs actually work
4688 => three warnings about OTP not working
4689 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4691 With the patch, sasl_listmech() also gets called three times. But now
4692 SASL's mech_list option is set to the server_mech specified in the the
4693 authenticator. Or in other words, the answer from sasl_listmech()
4694 gets limited to just the mech you're testing for (which is different
4696 => the return list contains just 'plain' or 'login', 'cram-md5' or
4697 nothing depending on the value of ob->server_mech.
4699 I've just tested the patch: Authentication still works fine,
4700 unavailable mechs specified in the exim configuration are still
4701 caught, and the auth.log warnings about OTP are gone."
4703 PH/25 When debugging is enabled, the contents of the command line are added
4704 to the debugging output, even when log_selector=+arguments is not
4707 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4708 answer is "GNU", and only if the return is "GNU/something" is the answer
4711 PH/27 $acl_verify_message is now set immediately after the failure of a
4712 verification in an ACL, and so is available in subsequent modifiers. In
4713 particular, the message can be preserved by coding like this:
4715 warn !verify = sender
4716 set acl_m0 = $acl_verify_message
4718 Previously, $acl_verify_message was set only while expanding "message"
4719 and "log_message" when a very denied access.
4721 PH/28 Modified OS/os.c-Linux with
4723 -#ifndef OS_LOAD_AVERAGE
4724 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4726 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4727 nomenclature these days.)
4729 PH/29 Installed patch from the Sieve maintainer that adds the options
4730 sieve_useraddress and sieve_subaddress to the redirect router.
4732 PH/30 In these circumstances:
4733 . Two addresses routed to the same list of hosts;
4734 . First host does not offer TLS;
4735 . First host accepts first address;
4736 . First host gives temporary error to second address;
4737 . Second host offers TLS and a TLS session is established;
4738 . Second host accepts second address.
4739 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4740 and peerdn, if requested) that were in fact used only for the second
4743 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4744 attention to any local part prefix or suffix that was matched by the
4745 router that accepted the address. It now behaves in the same way as it
4746 does for delivery: the affixes are removed from the local part unless
4747 rcpt_include_affixes is set on the transport.
4749 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4750 timeout during the DATA phase of an incoming message.
4752 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4753 applied a patch sanctioned by the Sieve maintainer.
4755 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4756 the uid or gid is negative. A case of a negative gid caused this to be
4757 noticed. The fix allows for either to be negative.
4759 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4760 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4763 PH/36 The change PH/12 above was broken. Fixed it.
4765 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4766 the grounds that routing the same address twice would always produce the
4767 same answer. This might have been true once, but it is certainly no
4768 longer true now. Routing a child address may depend on the previous
4769 routing that produced that child. Some complicated redirection strategies
4770 went wrong when messages had multiple recipients, and made Exim's
4771 behaviour dependent on the order in which the addresses were given.
4773 I have moved the duplicate checking until after the routing is complete.
4774 Exim scans the addresses that are assigned to local and remote
4775 transports, and removes any duplicates. This means that more work will be
4776 done, as duplicates will always all be routed, but duplicates are
4777 presumably rare, so I don't expect this is of any significance.
4779 For deliveries to pipes, files, and autoreplies, the duplicate checking
4780 still happens during the routing process, since they are not going to be
4783 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4784 It corrects a timeout issue with spamd. This is Ian's comment: "The
4785 background is that sometimes spamd either never reads data from a
4786 connection it has accepted, or it never writes response data. The exiscan
4787 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4788 blindly assumes that writes won't block so it may never time out."
4790 PH/39 Allow G after quota size as well as K and M.
4792 PH/40 The value set for $authenticated_id in an authenticator may not contain
4793 binary zeroes or newlines because the value is written to log lines and
4794 to spool files. There was no check on this. Now the value is run through
4795 the string_printing() function so that such characters are converted to
4796 printable escape sequences.
4798 PH/41 $message_linecount is a new variable that contains the total number of
4799 lines in the message. Compare $body_linecount, which is the count for the
4802 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4803 bounce and delay warning messages, except in certain special cases, which
4806 (a) An SMTP error message from a remote host;
4807 (b) A message specified in a :fail: redirection;
4808 (c) A message specified in a "fail" command in a system filter;
4809 (d) A message specified in a FAIL return from the queryprogram router;
4810 (e) A message specified by the cannot_route_message router option.
4812 In these cases only, Exim does include the error details in bounce and
4813 warning messages. There are also a few cases where bland messages such
4814 as "unrouteable address" or "local delivery error" are given.
4816 PH/43 $value is now also set for the "else" part of a ${run expansion.
4818 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4819 being worked on, but at least Exim now implements the latest version to
4822 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4823 process to complete was treated as a delivery failure, a timeout while
4824 writing the message to the pipe was logged, but erroneously treated as a
4825 successful delivery. Such timeouts include transport filter timeouts. For
4826 consistency with the overall process timeout, these timeouts are now
4827 treated as errors, giving rise to delivery failures by default. However,
4828 there is now a new Boolean option for the pipe transport called
4829 timeout_defer, which, if set TRUE, converts the failures into defers for
4830 both kinds of timeout. A transport filter timeout is now identified in
4833 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4834 systems where "make" and "gmake" are different, calling "gmake" at top
4835 level broke things. I've arranged for the value of $(MAKE) to be passed
4836 from the Makefile to this script so that it can call the same version of
4840 A note about Exim versions 4.44 and 4.50
4841 ----------------------------------------
4843 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4844 changes of various kinds. As a consequence, a big documentation update was
4845 needed. This delayed the release for rather longer than seemed good, especially
4846 in the light of a couple of (minor) security issues. Therefore, the changes
4847 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4848 release. So 4.44 and 4.50 are in effect two different branches that both start
4851 I have left the 4.50 change log unchanged; it contains all the changes since
4852 4.43. The change log for 4.44 is below; many of its items are identical to
4853 those for 4.50. This seems to be the most sensible way to preserve the
4854 historical information.
4860 1. Minor wording change to the doc/README.SIEVE file.
4862 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4863 computation of the current number of files was incorrect.
4865 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4866 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4869 4. Give more explanation in the error message when the command for a transport
4870 filter fails to execute.
4872 5. There are several places where Exim runs a non-Exim command in a
4873 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4874 was being done only for the command run by the queryprogram router. It is
4875 now done for all such subprocesses. The other cases are: ${run, transport
4876 filters, and the commands run by the lmtp and pipe transports.
4878 6. Added CONFIGURE_GROUP build-time option.
4880 7. Some older OS have a limit of 256 on the maximum number of file
4881 descriptors. Exim was using setrlimit() to set 1000 as a large value
4882 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4883 systems. I've change it so that if it can't get 1000, it tries for 256.
4885 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4886 was an oversight, and furthermore, ever since the addition of extra
4887 controls (e.g. 4.43/32), the checks on when to allow different forms of
4888 "control" were broken. There should now be diagnostics for all cases when a
4889 control that does not make sense is encountered.
4891 9. Added the /retain_sender option to "control=submission".
4893 10. $recipients is now available in the predata ACL (oversight).
4895 11. Tidy the search cache before the fork to do a delivery from a message
4896 received from the command line. Otherwise the child will trigger a lookup
4897 failure and thereby defer the delivery if it tries to use (for example) a
4898 cached ldap connection that the parent has called unbind on.
4900 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4901 of $address_data from the recipient verification was clobbered by the
4902 sender verification.
4904 13. The value of address_data from a sender verification is now available in
4905 $sender_address_data in subsequent conditions in the ACL statement.
4907 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4909 15. Added a new option "connect=<time>" to callout options, to set a different
4912 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4913 was its contents. (It was OK if the option was not defined at all.)
4915 17. A "Completed" log line is now written for messages that are removed from
4916 the spool by the -Mrm option.
4918 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4919 information about exactly what failed.
4921 19. Added -dd to debug only the daemon process.
4923 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4924 handles timeouts, both on the server side and network timeouts. Renamed the
4925 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4927 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4928 It is now set to "smtps".
4930 22. $host_address is now set to the target address during the checking of
4931 ignore_target_hosts.
4933 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4934 being passed; this would have caused $sender_host_name to have been used if
4935 matching the list had actually called for a host name (not very likely,
4936 since this list is usually IP addresses). A host name is now passed as
4939 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4940 code that specifies a non-restarting handler (typically sigaction() in
4941 modern systems) in an attempt to fix a rare and obscure crash bug.
4943 25. Narrowed the window for a race in the daemon that could cause it to ignore
4944 SIGCHLD signals. This is not a major problem, because they are used only to
4945 wake it up if nothing else does.
4947 26. A malformed maildirsize file could cause Exim to calculate negative values
4948 for the mailbox size or file count. Odd effects could occur as a result.
4949 The maildirsize information is now recalculated if the size or filecount
4952 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4953 support for a long time. Removed HAVE_SYS_VFS_H.
4955 28. Installed the latest version of exipick from John Jetmore.
4957 29. In an address list, if the pattern was not a regular expression, an empty
4958 subject address (from a bounce message) matched only if the pattern was an
4959 empty string. Non-empty patterns were not even tested. This was the wrong
4960 because it is perfectly reasonable to use an empty address as part of a
4961 database query. An empty address is now tested by patterns that are
4962 lookups. However, all the other forms of pattern expect the subject to
4963 contain a local part and a domain, and therefore, for them, an empty
4964 address still always fails if the pattern is not itself empty.
4966 30. Exim went into a mad DNS loop when attempting to do a callout where the
4967 host was specified on an smtp transport, and looking it up yielded more
4968 than one IP address.
4970 31. Re-factored the code for checking spool and log partition space into a
4971 function that finds that data and another that does the check. The former
4972 is then used to implement four new variables: $spool_space, $log_space,
4973 $spool_inodes, and $log_inodes.
4975 32. The RFC2047 encoding function was originally intended for short strings
4976 such as real names; it was not keeping to the 75-character limit for
4977 encoded words that the RFC imposes. It now respects the limit, and
4978 generates multiple encoded words if necessary. To be on the safe side, I
4979 have increased the buffer size for the ${rfc2047: expansion operator from
4982 33. It is now permitted to omit both strings after an "if" condition; if the
4983 condition is true, the result is "true". As before, when the second string
4984 is omitted, a false condition yields an empty string. This makes it less
4985 cumbersome to write custom ACL and router conditions.
4987 34. Failure to deliver a bounce message always caused it to be frozen, even if
4988 there was an errors_to setting on the router. The errors_to setting is now
4991 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4992 canonical form (fully expanded) before being placed in
4993 $sender_host_address.
4995 36. The table in the code that translates DNS record types into text (T_A to
4996 "A" for instance) was missing entries for NS and CNAME. It is just possible
4997 that this could have caused confusion if both these types were looked up
4998 for the same domain, because the text type is used as part of Exim's
4999 per-process caching. But the chance of anyone hitting this buglet seems
5002 37. The dnsdb lookup has been extended in a number of ways.
5004 (1) There is a new type, "zns", which walks up the domain tree until it
5005 finds some nameserver records. It should be used with care.
5007 (2) There is a new type, "mxh", which is like "mx" except that it returns
5008 just the host names, not the priorities.
5010 (3) It is now possible to give a list of domains (or IP addresses) to be
5011 looked up. The behaviour when one of the lookups defers can be
5012 controlled by a keyword.
5014 (4) It is now possible to specify the separator character for use when
5015 multiple records are returned.
5017 38. The dnslists ACL condition has been extended: it is now possible to supply
5018 a list of IP addresses and/or domains to be looked up in a particular DNS
5021 39. Added log_selector=+queue_time_overall.
5023 40. When running the queue in the test harness, wait just a tad after forking a
5024 delivery process, to get repeatability of debugging output.
5026 41. Include certificate and key file names in error message when GnuTLS fails
5027 to set them up, because the GnuTLS error message doesn't include the name
5028 of the failing file when there is a problem reading it.
5030 42. Allow both -bf and -bF in the same test run.
5032 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5034 44. The "Exiscan patch" is now merged into the mainline Exim source.
5036 45. Sometimes the final signoff response after QUIT could fail to get
5037 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5038 < 0 before doing a fflush(). This bug looks as though it goes back to the
5039 introduction of TLS in release 3.20, but "sometimes" must have been rare
5040 because the tests only now provoked it.
5042 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5043 (this can affect the format of dates).
5045 47. exim_tidydb, when checking for the continued existence of a message for
5046 which it has found a message-specific retry record, was not finding
5047 messages that were in split spool directories. Consequently, it was
5048 deleting retry records that should have stayed in existence.
5050 48. Steve fixed some bugs in eximstats.
5052 49. The SPA authentication driver was not abandoning authentication and moving
5053 on to the next authenticator when an expansion was forced to fail,
5054 contradicting the general specification for all authenticators. Instead it
5055 was generating a temporary error. It now behaves as specified.
5057 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5058 (the order specifies the preference for clients). The order is now AES256,
5059 AES128, 3DES, ARCFOUR128.
5061 51. Small patch to Sieve code - explicitly set From: when generating an
5064 52. Exim crashed if a remote delivery caused a very long error message to be
5065 recorded - for instance if somebody sent an entire SpamAssassin report back
5066 as a large number of 550 error lines. This bug was coincidentally fixed by
5067 increasing the size of one of Exim's internal buffers (big_buffer) that
5068 happened as part of the Exiscan merge. However, to be on the safe side, I
5069 have made the code more robust (and fixed the comments that describe what
5072 53. Now that there can be additional text after "Completed" in log lines (if
5073 the queue_time_overall log selector is set), a one-byte patch to exigrep
5074 was needed to allow it to recognize "Completed" as not the last thing in
5077 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5078 patch that reportedly fixes this has been added. I am not expert enough to
5079 create a test for it. This is what the patch creator wrote:
5081 "I found a little strange behaviour of ldap code when working with
5082 Windows 2003 AD Domain, where users was placed in more than one
5083 Organization Units. When I tried to give exim partial DN, the exit code
5084 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5085 But simultaneously result of request was absolutely normal ldap result,
5086 so I produce this patch..."
5088 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5089 REFERENCE, so I have modified the code to exclude the patch when that macro
5092 55. Some experimental protocols are using DNS PTR records for new purposes. The
5093 keys for these records are domain names, not reversed IP addresses. The
5094 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5095 leaves it alone. Component reversal etc. now happens only for IP addresses.
5098 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5100 57. Double the size of the debug message buffer (to 2048) so that more of very
5101 long debug lines gets shown.
5103 58. The exicyclog utility now does better if the number of log files to keep
5104 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5106 59. Two changes related to the smtp_active_hostname option:
5108 (1) $smtp_active_hostname is now available as a variable.
5109 (2) The default for smtp_banner uses $smtp_active_hostname instead
5110 of $primary_hostname.
5112 60. The host_aton() function is supposed to be passed a string that is known
5113 to be a valid IP address. However, in the case of IPv6 addresses, it was
5114 not checking this. This is a hostage to fortune. Exim now panics and dies
5115 if the condition is not met. A case was found where this could be provoked
5116 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5117 components; fortuitously, this particular loophole had already been fixed
5118 by change 4.50/55 above.
5120 If there are any other similar loopholes, the new check in host_aton()
5121 itself should stop them being exploited. The report I received stated that
5122 data on the command line could provoke the exploit when Exim was running as
5123 exim, but did not say which command line option was involved. All I could
5124 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5125 running as the user.
5128 61. There was a buffer overflow vulnerability in the SPA authentication code
5129 (which came originally from the Samba project). I have added a test to the
5130 spa_base64_to_bits() function which I hope fixes it.
5133 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5134 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5136 63. The daemon start-up calls getloadavg() while still root for those OS that
5137 need the first call to be done as root, but it missed one case: when
5138 deliver_queue_load_max is set with deliver_drop_privilege. This is
5139 necessary for the benefit of the queue runner, because there is no re-exec
5140 when deliver_drop_privilege is set.
5142 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5143 This has been fixed.
5145 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5146 was in use, was not putting the data itself into the right store pool;
5147 consequently, it could be overwritten for a subsequent message in the same
5148 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5151 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5153 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5154 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5155 the function were treating the return as a boolean value, which happened to
5156 work because 0=false and not-0=true, but is not correct code.
5158 68. The host_aton() function was not handling scoped IPv6 addresses (those
5159 with, for example, "%eth0" on the end) correctly.
5161 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5162 negated items (that is, ~something) in unsigned ints. Some compilers
5163 apparently mutter when there is no cast.
5165 70. If an address verification called from an ACL failed, and did not produce a
5166 user-specific message (i.e. there was only a "system" message), nothing was
5167 put in $acl_verify_message. In this situation, it now puts the system
5170 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5171 session; change 4.31/43 added the unwanted input to the log line - except
5172 that it did not do this in the start of session case. It now does.
5174 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5175 This is wrong and can cause the other end to generate a synchronization
5176 error if it is another Exim or anything else that does the synchronization
5177 check. A QUIT command is no longer sent after a timeout.
5179 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5180 during host lookups.
5182 74. The defer_ok option of callout verification was not working if it was used
5183 when verifying addresses in header lines, that is, for this case:
5185 verify = header_sender/callout=defer_ok
5187 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5188 those file descriptors could be used for SMTP connections. If anything
5189 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5190 could be sent to the SMTP client, causing chaos. The daemon now opens
5191 stdin, stdout, and stderr to /dev/null when it puts itself into the
5194 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5195 log by default. The user can override this with suitable Perl magic.
5197 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5198 the log message when discard triggers, was not working for the DATA ACL or
5199 for the non-SMTP ACL.
5201 78. Error message wording change in sieve.c.
5203 79. If smtp_accept_max_per_host was set, the number of connections could be
5204 restricted to fewer than expected, because the daemon was trying to set up
5205 a new connection before checking whether the processes handling previous
5206 connections had finished. The check for completed processes is now done
5207 earlier. On busy systems, this bug wouldn't be noticed because something
5208 else would have woken the daemon, and it would have reaped the completed
5211 80. If a message was submitted locally by a user whose login name contained one
5212 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5213 It caused a spool format error. I have fixed the spool reading code. A
5214 related problem was that the "from" clause in the Received: line became
5215 illegal because of the space(s). It is now covered by ${quote_local_part.
5217 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5220 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5222 83. Patch from Sieve maintainer to fix unterminated string problem in
5223 "vacation" handling.
5225 84. Some minor changes to the Linux configuration files to help with other
5226 OS variants using glibc.
5228 85. One more patch for Sieve to update vacation handling to latest spec.
5231 ----------------------------------------------------
5232 See the note above about the 4.44 and 4.50 releases.
5233 ----------------------------------------------------
5239 1. Change 4.43/35 introduced a bug that caused file counts to be
5240 incorrectly computed when quota_filecount was set in an appendfile
5243 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5244 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5247 3. Give more explanation in the error message when the command for a transport
5248 filter fails to execute.
5250 4. There are several places where Exim runs a non-Exim command in a
5251 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5252 was being done only for the command run by the queryprogram router. It is
5253 now done for all such subprocesses. The other cases are: ${run, transport
5254 filters, and the commands run by the lmtp and pipe transports.
5256 5. Some older OS have a limit of 256 on the maximum number of file
5257 descriptors. Exim was using setrlimit() to set 1000 as a large value
5258 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5259 systems. I've change it so that if it can't get 1000, it tries for 256.
5261 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5262 was an oversight, and furthermore, ever since the addition of extra
5263 controls (e.g. 4.43/32), the checks on when to allow different forms of
5264 "control" were broken. There should now be diagnostics for all cases when a
5265 control that does not make sense is encountered.
5267 7. $recipients is now available in the predata ACL (oversight).
5269 8. Tidy the search cache before the fork to do a delivery from a message
5270 received from the command line. Otherwise the child will trigger a lookup
5271 failure and thereby defer the delivery if it tries to use (for example) a
5272 cached ldap connection that the parent has called unbind on.
5274 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5275 of $address_data from the recipient verification was clobbered by the
5276 sender verification.
5278 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5279 was its contents. (It was OK if the option was not defined at all.)
5281 11. A "Completed" log line is now written for messages that are removed from
5282 the spool by the -Mrm option.
5284 12. $host_address is now set to the target address during the checking of
5285 ignore_target_hosts.
5287 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5288 being passed; this would have caused $sender_host_name to have been used if
5289 matching the list had actually called for a host name (not very likely,
5290 since this list is usually IP addresses). A host name is now passed as
5293 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5294 code that specifies a non-restarting handler (typically sigaction() in
5295 modern systems) in an attempt to fix a rare and obscure crash bug.
5297 15. Narrowed the window for a race in the daemon that could cause it to ignore
5298 SIGCHLD signals. This is not a major problem, because they are used only to
5299 wake it up if nothing else does.
5301 16. A malformed maildirsize file could cause Exim to calculate negative values
5302 for the mailbox size or file count. Odd effects could occur as a result.
5303 The maildirsize information is now recalculated if the size or filecount
5306 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5307 support for a long time. Removed HAVE_SYS_VFS_H.
5309 18. Updated exipick to current release from John Jetmore.
5311 19. Allow an empty sender to be matched against a lookup in an address list.
5312 Previously the only cases considered were a regular expression, or an
5315 20. Exim went into a mad DNS lookup loop when doing a callout where the
5316 host was specified on the transport, if the DNS lookup yielded more than
5319 21. The RFC2047 encoding function was originally intended for short strings
5320 such as real names; it was not keeping to the 75-character limit for
5321 encoded words that the RFC imposes. It now respects the limit, and
5322 generates multiple encoded words if necessary. To be on the safe side, I
5323 have increased the buffer size for the ${rfc2047: expansion operator from
5326 22. Failure to deliver a bounce message always caused it to be frozen, even if
5327 there was an errors_to setting on the router. The errors_to setting is now
5330 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5331 canonical form (fully expanded) before being placed in
5332 $sender_host_address.
5334 24. Updated eximstats to version 1.33
5336 25. Include certificate and key file names in error message when GnuTLS fails
5337 to set them up, because the GnuTLS error message doesn't include the name
5338 of the failing file when there is a problem reading it.
5340 26. Expand error message when OpenSSL has problems setting up cert/key files.
5343 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5344 (this can affect the format of dates).
5346 28. exim_tidydb, when checking for the continued existence of a message for
5347 which it has found a message-specific retry record, was not finding
5348 messages that were in split spool directories. Consequently, it was
5349 deleting retry records that should have stayed in existence.
5351 29. eximstats updated to version 1.35
5352 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5353 1.35 - bugfix such that pie charts by volume are generated correctly
5355 30. The SPA authentication driver was not abandoning authentication and moving
5356 on to the next authenticator when an expansion was forced to fail,
5357 contradicting the general specification for all authenticators. Instead it
5358 was generating a temporary error. It now behaves as specified.
5360 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5361 (the order specifies the preference for clients). The order is now AES256,
5362 AES128, 3DES, ARCFOUR128.
5364 31. Small patch to Sieve code - explicitly set From: when generating an
5367 32. Exim crashed if a remote delivery caused a very long error message to be
5368 recorded - for instance if somebody sent an entire SpamAssassin report back
5369 as a large number of 550 error lines. This bug was coincidentally fixed by
5370 increasing the size of one of Exim's internal buffers (big_buffer) that
5371 happened as part of the Exiscan merge. However, to be on the safe side, I
5372 have made the code more robust (and fixed the comments that describe what
5375 33. Some experimental protocols are using DNS PTR records for new purposes. The
5376 keys for these records are domain names, not reversed IP addresses. The
5377 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5378 leaves it alone. Component reversal etc. now happens only for IP addresses.
5381 34. The host_aton() function is supposed to be passed a string that is known
5382 to be a valid IP address. However, in the case of IPv6 addresses, it was
5383 not checking this. This is a hostage to fortune. Exim now panics and dies
5384 if the condition is not met. A case was found where this could be provoked
5385 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5386 components; fortuitously, this particular loophole had already been fixed
5387 by change 4.50/55 or 4.44/33 above.
5389 If there are any other similar loopholes, the new check in host_aton()
5390 itself should stop them being exploited. The report I received stated that
5391 data on the command line could provoke the exploit when Exim was running as
5392 exim, but did not say which command line option was involved. All I could
5393 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5394 running as the user.
5397 35. There was a buffer overflow vulnerability in the SPA authentication code
5398 (which came originally from the Samba project). I have added a test to the
5399 spa_base64_to_bits() function which I hope fixes it.
5402 36. The daemon start-up calls getloadavg() while still root for those OS that
5403 need the first call to be done as root, but it missed one case: when
5404 deliver_queue_load_max is set with deliver_drop_privilege. This is
5405 necessary for the benefit of the queue runner, because there is no re-exec
5406 when deliver_drop_privilege is set.
5408 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5409 was in use, was not putting the data itself into the right store pool;
5410 consequently, it could be overwritten for a subsequent message in the same
5411 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5414 38. Sometimes the final signoff response after QUIT could fail to get
5415 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5416 < 0 before doing a fflush(). This bug looks as though it goes back to the
5417 introduction of TLS in release 3.20, but "sometimes" must have been rare
5418 because the tests only now provoked it.
5424 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5425 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5426 Now it returns an integer. A number of calls were still expecting a T/F
5427 return. Fortuitously, in all cases, the tests worked in OK situations,
5428 which is the norm. However, things would have gone wrong on any write
5429 failures on the smtp file descriptor. This function is used when sending
5430 messages over SMTP and also when doing verify callouts.
5432 2. When Exim is called to do synchronous delivery of a locally submitted
5433 message (the -odf or -odi options), it no longer closes stderr before doing
5436 3. Implemented the mua_wrapper option.
5438 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5440 5. Implemented the functions header_remove(), header_testname(),
5441 header_add_at_position(), and receive_remove_recipient(), and exported them
5444 6. If an ACL "warn" statement specified the addition of headers, Exim already
5445 inserted X-ACL-Warn: at the start if there was no header name. However, it
5446 was not making this test for the second and subsequent header lines if
5447 there were newlines in the string. This meant that an invalid header could
5448 be inserted if Exim was badly configured.
5450 7. Allow an ACL "warn" statement to add header lines at the start or after all
5451 the Received: headers, as well as at the end.
5453 8. Added the rcpt_4xx retry error code.
5455 9. Added postmaster_mailfrom=xxx to callout verification option.
5457 10. Added mailfrom=xxxx to the callout verification option, for verify=
5460 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5461 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5463 12. Inserted some casts to stop certain compilers warning when using pointer
5464 differences as field lengths or precisions in printf-type calls (mostly
5465 affecting debugging statements).
5467 13. Added optional readline() support for -be (dynamically loaded).
5469 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5470 same clock tick as a message's arrival, so that its received time was the
5471 same as the "first fail" time on the retry record, and that message
5472 remained on the queue past the ultimate address timeout, every queue runner
5473 would try a delivery (because it was past the ultimate address timeout) but
5474 after another failure, the ultimate address timeout, which should have then
5475 bounced the address, did not kick in. This was a "< instead of <=" error;
5476 in most cases the first failure would have been in the next clock tick
5477 after the received time, and all would be well.
5479 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5480 being recognized when the domain list was tested by the match_domain
5481 condition in an expansion string.
5483 16. Added the ${str2b64: operator.
5485 17. Exim was always calling setrlimit() to set a large limit for the number of
5486 processes, without checking whether the existing limit was already
5487 adequate. (It did check for the limit on file descriptors.) Furthermore,
5488 errors from getrlimit() and setrlimit() were being ignored. Now they are
5489 logged to the main and panic logs, but Exim does carry on, to try to do its
5490 job under whatever limits there are.
5492 18. Imported PCRE 5.0.
5494 19. Trivial typo in log message " temporarily refused connection" (the leading
5497 20. If the log selector return_path_on_delivery was set and an address was
5498 redirected to /dev/null, the delivery process crashed because it assumed
5499 that a return path would always be set for a "successful" delivery. In this
5500 case, the whole delivery is bypassed as an optimization, and therefore no
5503 21. Internal re-arrangement: the function for sending a challenge and reading
5504 a response while authentication was assuming a zero-terminated challenge
5505 string. It's now changed to take a pointer and a length, to allow for
5506 binary data in such strings.
5508 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5510 23. Exim was not respecting finduser_retries when seeking the login of the
5511 uid under which it was called; it was always trying 10 times. (The default
5512 setting of finduser_retries is zero.) Also, it was sleeping after the final
5513 failure, which is pointless.
5515 24. Implemented tls_on_connect_ports.
5517 25. Implemented acl_smtp_predata.
5519 26. If the domain in control=submission is set empty, Exim assumes that the
5520 authenticated id is a complete email address when it generates From: or
5521 Sender: header lines.
5523 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5524 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5525 chown and chgrp in /bin and hostname in /usr/bin.
5527 28. Exim was keeping the "process log" file open after each use, just as it
5528 does for the main log. This opens the possibility of it remaining open for
5529 long periods when the USR1 signal hits a daemon. Occasional processlog
5530 errors were reported, that could have been caused by this. Anyway, it seems
5531 much more sensible not to leave this file open at all, so that is what now
5534 29. The long-running daemon process does not normally write to the log once it
5535 has entered its main loop, and it closes the log before doing so. This is
5536 so that log files can straightforwardly be renamed and moved. However,
5537 there are a couple of unusual error situations where the daemon does write
5538 log entries, and I had neglected to close the log afterwards.
5540 30. The text of an SMTP error response that was received during a remote
5541 delivery was being truncated at 512 bytes. This is too short for some of
5542 the long messages that one sometimes sees. I've increased the limit to
5545 31. It is now possible to make retry rules that apply only when a message has a
5546 specific sender, in particular, an empty sender.
5548 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5549 it possible to be selective about when SMTP synchronization is enforced.
5551 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5553 32. Implemented hosts_connection_nolog.
5555 33. Added an ACL for QUIT.
5557 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5560 35. Added mailbox_size and mailbox_filecount to appendfile.
5562 36. Added control = no_multiline_responses to ACLs.
5564 37. There was a bug in the logic of the code that waits for the clock to tick
5565 in the case where the clock went backwards by a substantial amount such
5566 that the microsecond fraction of "now" was more than the microsecond
5567 fraction of "then" (but the whole seconds number was less).
5569 38. Added support for the libradius Radius client library this is found on
5570 FreeBSD (previously only the radiusclient library was supported).
5576 1. When certain lookups returned multiple values in the form name=value, the
5577 quoting of the values was not always being done properly. Specifically:
5578 (a) If the value started with a double quote, but contained no whitespace,
5580 (b) If the value contained whitespace other than a space character (i.e.
5581 tabs or newlines or carriage returns) it was not quoted.
5582 This fix has been applied to the mysql and pgsql lookups by writing a
5583 separate quoting function and calling it from the lookup code. The fix
5584 should probably also be applied to nisplus, ibase and oracle lookups, but
5585 since I cannot test any of those, I have not disturbed their existing code.
5587 2. A hit in the callout cache for a specific address caused a log line with no
5588 reason for rejecting RCPT. Now it says "Previous (cached) callout
5589 verification failure".
5591 3. There was an off-by-one bug in the queryprogram router. An over-long
5592 return line was truncated at 256 instead of 255 characters, thereby
5593 overflowing its buffer with the terminating zero. As well as fixing this, I
5594 have increased the buffer size to 1024 (and made a note to document this).
5596 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5597 when Exim is waiting for an SMTP response from a remote server, Exim
5598 restarts its select() call on the socket, thereby resetting its timeout.
5599 This is not a problem when such interrupts are rare. Somebody set up a cron
5600 job to run exiwhat every 2 minutes, which is less than the normal select()
5601 timeout (5 or 10 minutes). This meant that the select() timeout never
5602 kicked in because it was always reset. I have fixed this by comparing the
5603 time when an interrupt arrives with the time at the start of the first call
5604 to select(). If more time than the timeout has elapsed, the interrupt is
5605 treated as a timeout.
5607 5. Some internal re-factoring in preparation for the addition of Sieve
5608 extensions (by MH). In particular, the "personal" test is moved to a
5609 separate function, and given an option for scanning Cc: and Bcc: (which is
5610 not set for Exim filters).
5612 6. When Exim created an email address using the login of the caller as the
5613 local part (e.g. when creating a From: or Sender: header line), it was not
5614 quoting the local part when it contained special characters such as @.
5616 7. Installed new OpenBSD configuration files.
5618 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5619 try to make them clearer.
5621 9. Callout options, other than the timeout value, were being ignored when
5622 verifying sender addresses in header lines. For example, when using
5624 verify = header_sender/callout=no_cache
5626 the cache was (incorrectly) being used.
5628 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5629 only the Cygwin environment.
5631 11. When return_path_on_delivery was set as a log selector, if different remote
5632 addresses in the same message used different return paths and parallel
5633 remote delivery occurred, the wrong values would sometimes be logged.
5634 (Whenever a remote delivery process finished, the return path value from
5635 the most recently started remote delivery process was logged.)
5637 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5638 lines when AUTH and/or TLS are in use. This is the "received protocol"
5639 field. Exim used to use "asmtp" for authenticated SMTP, without any
5640 indication (in the protocol name) for TLS use. Now it follows the RFC and
5641 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5642 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5643 names appear in log lines as well as in Received: header lines.
5645 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5646 extensions, and comparison tests, and to fix some bugs.
5648 14. Changes to the "personal" filter test:
5650 (1) The test was buggy in that it was just doing the equivalent of
5651 "contains" tests on header lines. For example, if a user's address was
5652 anne@some.where, the "personal" test would incorrectly be true for
5654 To: susanne@some.where
5656 This test is now done by extracting each address from the header in turn,
5657 and checking the entire address. Other tests that are part of "personal"
5658 are now done using regular expressions (for example, to check local parts
5659 of addresses in From: header lines).
5661 (2) The list of non-personal local parts in From: addresses has been
5662 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5663 taken from the Sieve specification recommendations.
5665 (3) If the message contains any header line starting with "List-" it is
5666 treated as non-personal.
5668 (4) The test for "circular" in the Subject: header line has been removed
5669 because it now seems ill-conceived.
5671 15. Minor typos in src/EDITME comments corrected.
5673 16. Installed latest exipick from John Jetmore.
5675 17. If headers_add on a router specified a text string that was too long for
5676 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5677 of string_sprintf() is now avoided.
5679 18. $message_body_size was not set (it was always zero) when running the DATA
5680 ACL and the local_scan() function.
5682 19. For the "mail" command in an Exim filter, no default was being set for
5683 the once_repeat time, causing a random time value to be used if "once" was
5684 specified. (If the value happened to be <= 0, no repeat happened.) The
5685 default is now 0s, meaning "never repeat". The "vacation" command was OK
5686 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5687 (I found it when inspecting the code).
5689 20. There is now an overall timeout for performing a callout verification. It
5690 defaults to 4 times the callout timeout, which applies to individual SMTP
5691 commands during the callout. The overall timeout applies when there is more
5692 than one host that can be tried. The timeout is checked before trying the
5693 next host. This prevents very long delays if there are a large number of
5694 hosts and all are timing out (e.g. when the network connections are timing
5695 out). The value of the overall timeout can be changed by specifying an
5696 additional sub-option for "callout", called "maxwait". For example:
5698 verify = sender/callout=5s,maxwait=20s
5700 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5701 to the end before writing, but this should make it even safer).
5703 22. Exim was forgetting that it had advertised PIPELINING for the second and
5704 subsequent messages on an SMTP connection. It was also not resetting its
5705 memory on STARTTLS and an internal HELO.
5707 23. When Exim logs an SMTP synchronization error within a session, it now
5708 records whether PIPELINING has been advertised or not.
5710 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5711 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5712 rather than long int.
5714 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5716 26. Added the never_mail option to autoreply.
5722 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5723 crash if the getsockname() call failed; this can happen if a connection is
5724 closed very soon after it is established. The problem was simply in the
5725 order in which certain operations were done, causing Exim to try to write
5726 to the SMTP stream before it had set up the file descriptor. The bug has
5727 been fixed by making things happen in the correct order.
5733 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5734 before the connection was closed, thus losing the rejection response.
5736 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5737 some early Solaris releases, but causes trouble in current releases where
5738 socklen_t is defined.
5740 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5743 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5746 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5747 that releasing the top of it at the end releases what was used for sub-
5748 expansions (unless the block got too big). However, discard this block if
5749 the first thing is a variable or header, so that we can use its block when
5750 it is dynamic (useful for very large $message_headers, for example).
5752 6. Lookups now cache *every* query, not just the most recent. A new, separate
5753 store pool is used for this. It can be recovered when all lookup caches are
5754 flushed. Lookups now release memory at the end of their result strings.
5755 This has involved some general refactoring of the lookup sources.
5757 7. Some code has been added to the store_xxx() functions to reduce the amount
5758 of flapping under certain conditions.
5760 8. log_incoming_interface used to affect only the <= reception log lines. Now
5761 it causes the local interface and port to be added to several more SMTP log
5762 lines, for example "SMTP connection from", and rejection lines.
5764 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5766 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5768 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5769 could be overwritten at the end of the current message (or the start of a
5770 new message if it was set in a HELO ACL). The value is now preserved for
5771 the duration of the SMTP connection.
5773 12. If a transport had a headers_rewrite setting, and a matching header line
5774 contained an unqualified address, that address was qualified, even if it
5775 did not match any rewriting rules. The underlying bug was that the values
5776 of the flags that permit the existence of unqualified sender and recipient
5777 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5778 non-local messages, and by -bnq for local messages) were not being
5779 preserved with the message after it was received.
5781 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5782 "next input=" as part of the text comprising the host identity instead of
5783 the correct text. The code was using the same buffer for two different
5784 strings. However, depending on which order the printing function evaluated
5785 its arguments, the bug did not always show up. Under Linux, for example, my
5786 test suite worked just fine.
5788 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5789 doesn't work with some older versions of Perl. It has been changed to "my",
5790 which in any case is probably the better facility to use.
5792 15. A really picky compiler found some instances of statements for creating
5793 error messages that either had too many or two few arguments for the format
5796 16. The size of the buffer for calls to the DNS resolver has been increased
5797 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5798 for addresses that have a lot of PTR records. This alleviates a problem; it
5799 does not fully solve it.
5801 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5802 buffer now truncates the list and logs the incident, which is the same
5803 action as happens when Exim is looking up a host name and its aliases.
5804 Previously in this situation something unpredictable would happen;
5805 sometimes it was "internal error: store_reset failed".
5807 18. If a server dropped the connection unexpectedly when an Exim client was
5808 using GnuTLS and trying to read a response, the client delivery process
5809 crashed while trying to generate an error log message.
5811 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5812 string, for example:
5814 warn message = H1: something\nH2: something
5816 the text was added as a single header line from Exim's point of view
5817 though it ended up OK in the delivered message. However, searching for the
5818 second and subsequent header lines using $h_h2: did not work. This has been
5819 fixed. Similarly, if a system filter added multiple headers in this way,
5820 the routers could not see them.
5822 20. Expanded the error message when iplsearch is called with an invalid key to
5823 suggest using net-iplsearch in a host list.
5825 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5826 ACLs are no longer actually imposed (and a message to that effect is
5829 22. If a "gecos" field in a passwd entry contained escaped characters, in
5830 particular, if it contained a \" sequence, Exim got it wrong when building
5831 a From: or a Sender: header from that name. A second bug also caused
5832 incorrect handling when an unquoted " was present following a character
5833 that needed quoting.
5835 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5836 was not being matched caselessly.
5838 24. Arranged for all hyphens in the exim.8 source to be escaped with
5841 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5842 properly. Recipient callout cache records were still being keyed to include
5843 the sender, even when use_sender was set false. This led to far more
5844 callouts that were necessary. The sender is no longer included in the key
5845 when use_sender is false.
5847 26. Added "control = submission" modifier to ACLs.
5849 27. Added the ${base62d: operator to decode base 62 numbers.
5851 28. dnsdb lookups can now access SRV records.
5853 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5854 the configuration file.
5856 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5857 (-v) mode. This makes the output for a verbose queue run more intelligible.
5859 31. Added a use_postmaster feature to recipient callouts.
5861 32. Added the $body_zerocount variable, containing the number of binary zero
5862 bytes in the message body.
5864 33. The time of last modification of the "new" subdirectory is now used as the
5865 "mailbox time last read" when there is a quota error for a maildir
5868 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5870 35. Added +ignore_unknown as a special item in host lists.
5872 36. Code for decoding IPv6 addresses in host lists is now included, even if
5873 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5874 address was recognized as an IP address, but was then not correctly decoded
5875 into binary, causing unexpected and incorrect effects when compared with
5882 1. Very minor rewording of debugging text in manualroute to say "list of
5883 hosts" instead of "hostlist".
5885 2. If verify=header_syntax was set, and a header line with an unqualified
5886 address (no domain) and a large number of spaces between the end of the
5887 name and the colon was received, the reception process suffered a buffer
5888 overflow, and (when I tested it) crashed. This was caused by some obsolete
5889 code that should have been removed. The fix is to remove it!
5891 3. When running in the test harness, delay a bit after writing a bounce
5892 message to get a bit more predictability in the log output.
5894 4. Added a call to search_tidyup() just before forking a reception process. In
5895 theory, someone could use a lookup in the expansion of smtp_accept_max_
5896 per_host which, without the tidyup, could leave open a database connection.
5898 5. Added the variables $recipient_data and $sender_data which get set from a
5899 lookup success in an ACL "recipients" or "senders" condition, or a router
5900 "senders" option, similar to $domain_data and $local_part_data.
5902 6. Moved the writing of debug_print from before to after the "senders" test
5905 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5906 problems for message scanning, either using a data ACL, or using
5907 local_scan() because the Received: header was not generated till after they
5908 were called (in order to set the time as the time of reception completion).
5909 I have revised the way this works. The header is now generated after the
5910 body is received, but before the ACL or local_scan() are called. After they
5911 are run, the timestamp in the header is updated.
5917 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5918 before starting a queue runner without re-exec. This happened only when
5919 deliver_drop_privilege was set or when the Exim user was set to root. The
5920 effect of the bug was that timeouts during subsequent deliveries caused
5921 crashes instead of being properly handled. The handler is now left at its
5922 default (and expected) setting.
5924 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5925 message, again when deliver_drop_privilege is set or Exim is run as root.
5926 The bug described in (1) was not present in this case, but the tidying up
5927 of the other signals was missing. I have made the two cases consistent.
5929 3. The ignore_target_hosts setting on a manualroute router was being ignored
5930 for hosts that were looked up using the /MX notation.
5932 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5935 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5936 operated on the sender address. After changing the $sender_address to <>
5937 for the sender address verify, Exim was re-instated it as the original
5938 (before rewriting) address, but remembering that it had rewritten it, so it
5939 wasn't rewriting it again. This bug also had the effect of breaking the
5940 sender address verification caching when the sender address was rewritten.
5942 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5943 This has been changed so that if the ip literal address matches
5944 ignore_target_hosts, the router declines.
5946 7. Added expansion conditions match_domain, match_address, and match_local_
5947 part (NOT match_host).
5949 8. The placeholder for the Received: header didn't have a length field set.
5951 9. Added code to Exim itself and to exim_lock to test for a specific race
5952 condition that could lead to file corruption when using MBX delivery. The
5953 issue is with the lockfile that is created in /tmp. If this file is removed
5954 after a process has opened it but before that process has acquired a lock,
5955 there is the potential for a second process to recreate the file and also
5956 acquire a lock. This could lead to two Exim processes writing to the file
5957 at the same time. The added code performs the same test as UW imapd; it
5958 checks after acquiring the lock that its file descriptor still refers to
5959 the same named file.
5961 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5962 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5965 11. Added the smtp_active_hostname option. If used, this will typically be made
5966 to depend on the incoming interface address. Because $interface_address is
5967 not set up until the daemon has forked a reception process, error responses
5968 that can happen earlier (such as "too many connections") no longer contain
5971 12. If an expansion in a condition on a "warn" statement fails because a lookup
5972 defers, the "warn" statement is abandoned, and the next ACL statement is
5973 processed. Previously this caused the whole ACL to be aborted.
5975 13. Added the iplsearch lookup type.
5977 14. Added ident_timeout as a log selector.
5979 15. Added tls_certificate_verified as a log selector.
5981 16. Added a global option tls_require_ciphers (compare the smtp transport
5982 option of the same name). This controls incoming TLS connections.
5984 17. I finally figured out how to make tls_require_ciphers do a similar thing
5985 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5986 before starting the TLS session.
5988 18. Tabs are now shown as \t in -bP output.
5990 19. If the log selector return_path_on_delivery was set, Exim crashed when
5991 bouncing a message because it had too many Received: header lines.
5993 20. If two routers both had headers_remove settings, and the first one included
5994 a superfluous trailing colon, the final name in the first list and the
5995 first name in the second list were incorrectly joined into one item (with a
5996 colon in the middle).
6002 1. Added -C and -D options to the exinext utility, mainly to make it easier
6003 to include in the automated testing, but these could be helpful when
6004 multiple configurations are in use.
6006 2. The exinext utility was not formatting the output nicely when there was
6007 an alternate port involved in the retry record key, nor when there was a
6008 message id as well (for retries that were specific to a specific message
6009 and a specific host). It was also confused by IPv6 addresses, because of
6010 the additional colons they contain. I have fixed the IPv4 problem, and
6011 patched it up to do a reasonable job for IPv6.
6013 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6014 delivery, the log line now contains "pipelined" if PIPELINING was used.
6016 4. An SMTP transport process used to panic and die if the bind() call to set
6017 an explicit outgoing interface failed. This has been changed; it is now
6018 treated in the same way as a connect() failure.
6020 5. A reference to $sender_host_name in the part of a conditional expansion
6021 that was being skipped was still causing a DNS lookup. This no longer
6024 6. The def: expansion condition was not recognizing references to header lines
6025 that used bh_ and bheader_.
6027 7. Added the _cache feature to named lists.
6029 8. The code for checking quota_filecount in the appendfile transport was
6030 allowing one more file than it should have been.
6032 9. For compatibility with Sendmail, the command line option
6040 and sets the incoming protocol and host name (for trusted callers). The
6041 host name and its colon can be omitted when only the protocol is to be set.
6042 Note the Exim already has two private options, -pd and -ps, that refer to
6043 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6044 "s", but I don't think that's a major issue.
6046 10. A number of refactoring changes to the code, none of which should affect
6049 (a) The number of logging options was getting close to filling up the
6050 32-bit word that was used as a bit map. I have split them into two classes:
6051 those that are passed in the argument to log_write(), and those that are
6052 only ever tested independently outside of that function. These are now in
6053 separate 32-bit words, so there is plenty of room for expansion again.
6054 There is no change in the user interface or the logging behaviour.
6056 (b) When building, for example, log lines, the code previously used a
6057 macro that called string_cat() twice, in order to add two strings. This is
6058 not really sufficiently general. Furthermore, there was one instance where
6059 it was actually wrong because one of the argument was used twice, and in
6060 one call a function was used. (As it happened, calling the function twice
6061 did not affect the overall behaviour.) The macro has been replaced by a
6062 function that can join an arbitrary number of extra strings onto a growing
6065 (c) The code for expansion conditions now uses a table and a binary chop
6066 instead of a serial search (which was left over from when there were very
6067 few conditions). Also, it now recognizes conditions like "pam" even when
6068 the relevant support is not compiled in: a suitably worded error message is
6069 given if an attempt is made to use such a condition.
6071 11. Added ${time_interval:xxxxx}.
6073 12. A bug was causing one of the ddress fields not to be passed back correctly
6074 from remote delivery subprocesses. The field in question was not being
6075 subsequently used, so this caused to problems in practice.
6077 13. Added new log selectors queue_time and deliver_time.
6079 14. Might have fixed a bug in maildirsizefile handling that threw up
6080 "unexpected character" debug warnings, and recalculated the data
6081 unnecessarily. In any case, I expanded the warning message to give more
6084 15. Added the message "Restricted characters in address" to the statements in
6085 the default ACL that block characters like @ and % in local parts.
6087 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6088 Three changes have been made:
6090 (a) There was a serious bug; a negative response to MAIL caused the whole
6091 recipient domain to be cached as invalid, thereby blocking all messages
6092 to all local parts at the same domain, from all senders. This bug has
6093 been fixed. The domain is no longer cached after a negative response to
6094 MAIL if the sender used is not empty.
6096 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6099 (c) A new callout option, "use_sender" has been added for people who want
6100 the modified behaviour.
6106 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6109 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6110 indeed breaks things for older releases.
6112 3. Added additional logging to the case where there is a problem reading data
6113 from a filter that is running in a subprocess using a pipe, in order to
6114 try to track down a specific problem.
6116 4. Testing facility fudge: when running in the test harness and attempting
6117 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6118 getting "No route to host". Convert this to a timeout.
6120 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6123 6. Some OS don't have socklen_t but use size_t instead. This affects the
6124 fifth argument of getsockopt() amongst other things. This is now
6125 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6126 can be set for individual OS. I have set it for SunOS5, OSF1, and
6127 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6128 some earlier ones do not.
6130 7. Change 4.30/15 was not doing the test caselessly.
6132 8. The standard form for an IPv6 address literal was being rejected by address
6133 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6134 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6135 this, as well as the form without the "IPv6" on the front (but only when
6136 address literals are enabled, of course).
6138 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6140 10. Exim crashed if a message with an empty sender address specified by -f
6141 encountered a router with an errors_to setting. This could be provoked only
6142 by a command such as
6146 where an empty string was supplied; "<>" did not hit this bug.
6148 11. Installed PCRE release 4.5.
6150 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6151 remained set. It is now erased.
6153 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6154 times from message ids (which are base 36 rather than the normal 62).
6156 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6157 were being counted as actual protocol errors, and logged if the log
6158 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6159 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6160 and DATA following a set of rejected RCPTs do not count as protocol errors.
6161 In other words, Exim assumes they were pipelined, though this may not
6162 actually be the case. Of course, in all cases the client gets an
6163 appropriate error code.
6165 15. If a lookup fails in an ACL condition, a message about the failure may
6166 be available; it is used if testing the ACL cannot continue, because most
6167 such messages specify what the cause of the deferral is. However, some
6168 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6169 that caused an old message to be retained and used if a later statement
6170 caused a defer, replacing the real cause of the deferral.
6172 16. If an IP address had so many PTR records that the DNS lookup buffer
6173 was not large enough to hold them, Exim could crash while trying to process
6174 the truncated data. It now detects and logs this case.
6176 17. Further to 4.21/58, another change has been made: if (and only if) the
6177 first line of a message (the first header line) ends with CRLF, a bare LF
6178 in a subsequent header line has a space inserted after it, so as not to
6179 terminate the header.
6181 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6182 unnecessarily, used atoi() instead of strtol(), and didn't check the
6183 termination when getting file sizes from file names by regex.
6185 19. Completely re-implemented the support for maildirsize files, in the light
6186 of a number of problems with the previous contributed implementation
6187 (4.30/29). In particular:
6189 . If the quota is zero, the maildirsize file is maintained, but no quota is
6192 . If the maildir directory does not exist, it is created before any attempt
6193 to write a maildirsize file.
6195 . The quota value in the file is just a cache; if the quota is changed in
6196 the transport, the new value overrides.
6198 . A regular expression is available for excluding directories from the
6201 20. The autoreply transport checks the characters in options that define the
6202 message's headers; it allows continued headers, but it was checking with
6203 isspace() after an embedded newline instead of explicitly looking for a
6206 21. If all the "regular" hosts to which an address was routed had passed their
6207 expiry times, and had not reached their retry times, the address was
6208 bounced, even if fallback hosts were defined. Now Exim should go on to try
6211 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6212 equivalent code in the SMTP transport. Some hosts send humungous responses
6213 to HELO/EHLO, more than 1024 it seems.
6215 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6216 gives compiler warnings in some environments. I've now done it "properly",
6219 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6220 (because of problems with the built-in one) was declared to return uschar *
6221 instead of char *, causing compiler failure.
6223 25. Fixed a file descriptor leak when processing alias/forward files.
6225 26. Fixed a minor format string issue in dbfn.c.
6227 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6229 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6230 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6233 29. When a sender address is verified, it is cached, to save repeating the test
6234 when there is more than one recipient in a message. However, when the
6235 verification involves a callout, it is possible for different callout
6236 options to be set for different recipients. It is too complicated to keep
6237 track of this in the cache, so now Exim always runs a verification when a
6238 callout is required, relying on the callout cache for the optimization.
6239 The overhead is duplication of the address routing, but this should not be
6242 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6243 to be verified with callout=postmaster, and the main callout worked but the
6244 postmaster check failed, the verification correctly failed. However, if a
6245 subsequent RCPT command asked for sender verification *without* the
6246 postmaster check, incorrect caching caused this verification also to fail,
6249 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6250 it was not caching the DNS options (qualify_single, search_parents) that
6251 were used when the lookup failed. A subsequent lookup with different
6252 options therefore always gave the same answer, though there were cases
6253 where it should not have. (Example: a "domains = !$mx_any" option on a
6254 dnslookup router: the "domains" option is always processed without any
6255 widening, but the router might have qualify_single set.) Now Exim uses the
6256 cached value only when the same options are set.
6258 32. Added John Jetmore's "exipick" utility to the distribution.
6260 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6261 than a timeout (e.g. a certificate is required, and is not provided), an
6262 Exim server now closes the connection immediately. Previously it waited for
6263 the client to close - but if the client is SSL, it seems that they each
6264 wait for each other, leading to a delay before one of them times out.
6266 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6267 maintained 0.8.x compatibility because I don't think many are using it, and
6268 it is clearly obsolete.
6270 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6273 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6274 client certificate was expired. A simple patch fixes this, though I don't
6275 understand the full logic of why the verify callback is called multiple
6278 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6279 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6280 which causes problems with some clients (such as the Certicom SSL Plus
6281 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6282 disables the coutermeasure allowing Eudora to connect."
6284 38. Exim was not checking that a write() to a log file succeeded. This could
6285 lead to Bad Things if a log got too big, in particular if it hit a file
6286 size limit. Exim now panics and dies if it cannot write to a log file, just
6287 as it does if it cannot open a log file.
6289 39. Modified OS/Makefile-Linux so that it now contains
6291 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6293 The two -D definitions ensure that Exim is compiled with large file
6294 support, which makes it possible to handle log files that are bigger than
6297 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6298 instance) a domain was checked against a named list that involved a lookup,
6299 causing $domain_data to be set, then another domain was checked against the
6300 same list, then the first domain was re-checked, the value of $domain_data
6301 after the final check could be wrong. In particular, if the second check
6302 failed, it could be set empty. This bug probably also applied to
6305 41. The strip_trailing_dot option was not being applied to the address given
6306 with the -f command-line option.
6308 42. The code for reading a message's header from the spool was incrementing
6309 $received_count, but never initializing it. This meant that the value was
6310 incorrect (doubled) while delivering a message in the same process in which
6311 it was received. In the most common configuration of Exim, this never
6312 happens - a fresh exec is done - but it can happen when
6313 deliver_drop_privilege is set.
6315 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6316 it now includes up to 150 characters of the unexpected data in the log
6319 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6320 and building data strings. The size of both of these buffers was 10 000
6321 bytes - far larger than anybody would *ever* want, thought I. Needless to
6322 say, somebody hit the limit. I have increased the maximum line length to
6323 20 000 and the maximum data length of concatenated lines to 100 000. I have
6324 also fixed two bugs, because there was no checking on these buffers. Tsk,
6325 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6326 buffer is too small.
6328 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6329 lsearch lookups. Now it does.
6331 46. When parsing a route_list item in a manualroute router, a fixed-length
6332 buffer was used for the list of hosts. I made this 1024 bytes long,
6333 thinking that nobody would ever have a list of hosts that long. Wrong.
6334 Somebody had a whole pile of complicated expansion conditions, and the
6335 string was silently truncated, leading to an expansion error. It turns out
6336 that it is easier to change to an unlimited length (owing to other changes
6337 that have happened since this code was originally written) than to build
6338 structure for giving a limitation error. The length of the item that
6339 expands into the list of hosts is now unlimited.
6341 47. The lsearch lookup could not handle data where the length of text line was
6342 more than 4095 characters. Such lines were truncated, leading to shortened
6343 data being returned. It should now handle lines of any length.
6345 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6346 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6349 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6350 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6351 no longer happens in comment lines. A list of replacements is now placed
6352 at the head of all of the source files, except those whose only change is
6353 to replace PERL_COMMAND in the very first #! line.
6355 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6356 messages on the queue, with a bottom-up merge sort, using code contributed
6357 by Michael Haardt. This should make operations like -bp somewhat faster on
6358 large queues. It won't affect queue runners, except when queue_run_in_order
6361 51. Installed eximstats 1.31 in the distribution.
6363 52. Added support for SRV lookups to the dnslookup router.
6365 53. If an ACL referred to $message_body or $message_body_end, the value was not
6366 reset for any messages that followed in the same SMTP session.
6368 54. The store-handling optimization for building very long strings was not
6369 differentiating between the different store pools. I don't think this
6370 actually made any difference in practice, but I've tidied it.
6372 55. While running the routers to verify a sender address, $sender_address
6373 was still set to the sender address. This is wrong, because when routing to
6374 send a bounce to the sender, it would be empty. Therefore, I have changed
6375 it so that, while verifying a sender address, $sender_address is set to <>.
6376 (There is no change to what happens when verifying a recipient address.)
6378 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6379 target A or AAAA records (if not already returned) without resetting the
6380 qualify_single or search_parents options of the DNS resolver. These are
6381 inappropriate in this case because the targets of MX and SRV records must
6382 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6383 target that, when qualified, matched something in the local domain. These
6384 two options are now turned off when doing these lookups.
6386 57. It seems that at least some releases of Reiserfs (which does not have the
6387 concept of a fixed number of inodes) returns zero and not -1 for the
6388 number of available inodes. This interacted badly with check_spool_inodes,
6389 which assumed that -1 was the "no such thing" setting. What I have done is
6390 to check that the total number of inodes is greater than zero before doing
6391 the test of how many are available.
6393 58. When a "warn" ACL statement has a log_message modifier, the message is
6394 remembered, and not repeated. This is to avoid a lot of repetition when a
6395 message has many recipients that cause the same warning to be written.
6396 However, Exim was preserving the list of already written lines for an
6397 entire SMTP session, which doesn't seem right. The memory is now reset if a
6398 new message is started.
6400 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6401 debugging output unless log_rewrite was also set.
6403 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6404 of (int)(handle) when we know that handle contains (void *)(-1).
6406 61. The Exim daemon panic-logs an error return when it closes the incoming
6407 connection. However "connection reset by peer" seems to be common, and
6408 isn't really an error worthy of noting specially, so that particular error
6411 62. When Exim is trying to find all the local interfaces, it used to panic and
6412 die if the ioctl to get the interface flags failed. However, it seems that
6413 on at least one OS (Solaris 9) it is possible to have an interface that is
6414 included in the list of interfaces, but for which you get a failure error
6415 for this call. This happens when the interface is not "plumbed" into a
6416 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6417 failure of the "get flags" call assumes that the interface is down.
6419 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6420 makes life easier for people who are doing arithmetic on fields extracted
6421 from dates, where you often get leading zeros that should not be
6422 interpreted as octal.
6424 64. Added qualify_domain to the redirect router, to override the global
6427 65. If a pathologically long header line contained very many addresses (the
6428 report of this problem mentioned 10 000) and each of them was rewritten,
6429 Exim could use up a very large amount of memory. (It kept on making new
6430 copies of the header line as it rewrote, and never released the old ones.)
6431 At the expense of a bit more processing, the header rewriting function has
6432 been changed so that it no longer eats memory in this way.
6434 66. The generation of the Received: header has been moved from the time that a
6435 message starts to be received, to the time that it finishes. The timestamp
6436 in the Received: header should now be very close to that of the <= log
6437 line. There are two side-effects of this change:
6439 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6440 logged header lines no longer include the local Received: line, because
6441 it has not yet been created. The same applies to a copy of the message
6442 that is returned to a non-SMTP sender when a message is rejected.
6444 (b) When a filter file is tested using -bf, no additional Received: header
6445 is added to the test message. After some thought, I decided that this
6448 This change does not affect the value of $received_for. It is still set
6449 after address rewriting, but before local_scan() is called.
6451 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6453 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6454 gave an unhelpful panic error message, and a defer error. I have managed to
6455 change this behaviour so that it now rejects any supplied certificate,
6456 which seems right, as the list of acceptable certificates is empty.
6458 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6459 gave an unhelpful defer error. I have not managed to make this reject any
6460 supplied certificates, but the error message it gives is "no certificate
6461 supplied", which is not helpful.
6463 70. exigrep's output now also includes lines that are not associated with any
6464 message, but which match the given pattern. Implemented by a patch from
6465 Martin Sluka, which also tidied up the Perl a bit.
6467 71. Recipient callout verification, like sender verification, was using <> in
6468 the MAIL FROM command. This isn't really the right thing, since the actual
6469 sender may affect whether the remote host accepts the recipient or not. I
6470 have changed it to use the actual sender in the callout; this means that
6471 the cache record is now keyed on a recipient/sender pair, not just the
6472 recipient address. There doesn't seem to be a real danger of callout loops,
6473 since a callout by the remote host to check the sender would use <>.
6474 [SEE ABOVE: changed after hitting problems.]
6476 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6477 temporary errors. However, in the case of such a code being given after
6478 the end of a data transmission (i.e. after ".") Exim was failing to write
6479 a retry record for the message. (Yes, there was some broken host that was
6480 actually sending 8xx at this point.)
6482 73. An unknown lookup type in a host list could cause Exim to panic-die when
6483 the list was checked. (An example that provoked this was putting <; in the
6484 middle of a list instead of at the start.) If this happened during a DATA
6485 ACL check, a -D file could be left lying around. This kind of configuration
6486 error no longer causes Exim to die; instead it causes a defer error. The
6487 incident is still logged to the main and panic logs.
6489 74. Buglet left over from Exim 3 conversion. The message "too many messages
6490 in one connection" was written to the rejectlog but not the mainlog, except
6491 when address rewriting (yes!) was being logged.
6493 75. Added write_rejectlog option.
6495 76. When a system filter was run not as root (that is, when system_filter_user
6496 was set), the values of the $n variables were not being returned to the
6497 main process; thus, they were not subsequently available in the $sn
6500 77. Added +return_path_on_delivery log selector.
6502 78. A connection timeout was being treated differently from recipients deferred
6503 when testing hosts_max_try with a message that was older than the host's
6504 retry timeout. (The host should not be counted, thus allowing all hosts to
6505 be tried at least once before bouncing.) This may have been the cause of an
6506 occasionally reported bug whereby a message would remain on the queue
6507 longer than the retry timeout, but would be bounced if a delivery was
6508 forced. I say "may" because I never totally pinned down the problem;
6509 setting up timeout/retry tests is difficult. See also the next item.
6511 79. The ultimate address timeout was not being applied to errors that involved
6512 a combination of host plus message (for example, a timeout on a MAIL
6513 command). When an address resolved to a number of possible hosts, and they
6514 were not all tried for each delivery (e.g. because of hosts_max_try), a
6515 message could remain on the queue longer than the retry timeout.
6517 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6520 81. Fixed an obscure SMTP outgoing bug which required at least the following
6521 conditions: (a) there was another message waiting for the same server;
6522 (b) the server returned 5xx to all RCPT commands in the first message so
6523 that the message was not completed; (c) the server dropped the connection
6524 or gave a negative response to the RSET that Exim sends to abort the
6525 transaction. The observed case was a dropped connection after DATA that had
6526 been sent in pipelining mode. That is, the server had advertised PIPELINING
6527 but was not implementing it correctly. The effect of the bug was incorrect
6528 behaviour, such as trying another host, and this could lead to a crash.
6534 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6535 and daemon.c were passed as pointers to ints; they should have been
6536 pointers to socklen_t variables (which are typically unsigned ints).
6538 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6541 3. Fixed a really odd bug that affected only the testing scheme; patching a
6542 certain fixed string in the binary changed the value of another string that
6543 happened to be identical to the end of the original first string.
6545 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6546 name", it returns that address as the IP address. On some operating
6547 systems (e.g. Solaris), it also passes back the IP address string as the
6548 "host name". However, on others (e.g. Linux), it passes back an empty
6549 string. Exim wasn't checking for this, and was changing the host name to an
6550 empty string, assuming it had been canonicalized.
6552 5. Although rare, it is permitted to have more than one PTR record for a given
6553 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6554 all the names associated with an address, because they do in Solaris.
6555 However, it seems that they do not in Linux for data that comes from the
6556 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6557 I found this out when I moved to a new Linux workstation and tried to run
6558 the Exim test suite.
6560 To get round this problem I have changed the code so that it now does its
6561 own call to the DNS to look up PTR records when searching for a host name.
6562 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6563 addresses that are only in /etc/hosts are still found.
6565 This behaviour is, however, controlled by an option called host_lookup_
6566 order, which defaults to "bydns:byaddr". If people want to use the other
6567 order, or indeed, just use one or the other means of lookup, they can
6568 specify it in this variable.
6570 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6571 some operating systems, this comes back from gethostbyaddr() as an empty
6572 string, and this is what Exim used to test for. However, it seems that in
6573 other systems, "." is yielded. Exim now tests for this case too.
6575 7. The values of check_spool_space and check_log_space are now held internally
6576 as a number of kilobytes instead of an absolute number of bytes. If a
6577 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6578 kilobyte. This means that much larger values can be stored.
6580 8. Exim monitor: an attempt to get the action menu when not actually pointing
6581 at a message produces an empty menu entitled "No message selected". This
6582 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6583 no entries in it ("Shell widget menu has zero width and/or height"). So I
6584 have added a single, blank menu entry in this case.
6586 9. Added ${quote_local_part.
6588 10. MIME decoding is now applied to the contents of Subject: header lines when
6591 11. Now that a reference to $sender_host_address automatically causes a reverse
6592 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6593 host lookup before query-style lookups in lists that might use this
6594 variable. This has therefore been abolished, and the "net-" prefix is no
6595 longer necessary for query-style lookups.
6597 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6598 have been a typo for LFLAGS, so it has been changed.
6600 13. The install script calls Exim with "-C /dev/null" in order to find the
6601 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6602 to be output. However, since Exim outputs its version number before the
6603 error, it didn't break the script. It just looked ugly. I fixed this by
6604 always allowing "-C /dev/null" if the caller is root.
6606 14. Ignore overlarge ACL variable number when reading spool file - insurance
6607 against a later release with more variables having written the file.
6609 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6610 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6611 this, as well as the form without the "IPv6" on the front.
6613 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6614 OS/Makefile-Darwin file.
6616 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6617 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6619 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6620 variable. However, some OS use ints for inodes. I've added cast to long int
6621 to get rid of the compiler warning.
6623 19. I had forgotten to lock out "/../" in configuration file names when
6624 ALT_CONFIG_PREFIX was set.
6626 20. Routers used for verification do not need to specify transports. However,
6627 if such a router generated a host list, and callout was configured, Exim
6628 crashed, because it could not find a port number from the (non-existent)
6629 transport. It now assumes port 25 in this circumstance.
6631 21. Added the -t option to exigrep.
6633 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6634 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6635 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6636 them is set, LOOKUP_LSEARCH is forced.
6638 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6640 24. Added sender and host information to the "rejected by local_scan()" log
6641 line; previously there was no indication of these.
6643 25. Added .include_if_exists.
6645 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6646 Linux. It turns out that not all file systems support this. Apparently some
6647 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6648 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6649 not supported on the file descriptor, is now ignored when Exim is trying to
6650 sync a directory. This applies only to Linux.
6652 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6654 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6655 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6658 29. Added support for maildirsize files from supplied patch (modified a bit).
6660 30. The use of :fail: followed by an empty string could lead Exim to respond to
6661 sender verification failures with (e.g.):
6663 550 Verification failed for <xxx>
6664 550 Sender verify failed
6666 where the first response line was missing the '-' that indicates it is not
6667 the final line of the response.
6669 31. The loop for finding the name of the user that called Exim had a hardwired
6670 limit of 10; it now uses the value of finduser_retries, which is used for
6671 all other user lookups.
6673 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6676 33. Exim was neglecting to zero errno before one call of strtol() when
6677 expanding a string and expecting an integer value. On some systems this
6678 resulted in spurious "integer overflow" errors. Also, it was casting the
6679 result into an int without checking.
6681 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6682 did not work. The code looks as if it has *never* worked, though it appears
6683 to have been documented since at least release 1.62. I have made it work.
6685 35. The "timeout_DNS" error in retry rules, also documented since at least
6686 1.62, also never worked. As it isn't clear exactly what this means, and
6687 clearly it isn't a major issue, I have abolished the feature by treating it
6688 as "timeout", and writing a warning to the main and panic logs.
6690 36. The display of retry rules for -brt wasn't always showing the error code
6693 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6694 timeout_connect_A, timeout_connect_MX.
6696 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6697 to the empty sender.
6699 39. The daemon was not analysing the content of -oX till after it had closed
6700 stderr and disconnected from the controlling terminal. This meant that any
6701 syntax errors were only noted on the panic log, and the return code from
6702 the command was 0. By re-arranging the code a little, I've made the
6703 decoding happen first, so such errors now appear on stderr, and the return
6704 code is 1. However, the actual setting up of the sockets still happens in
6705 the disconnected process, so errors there are still only recorded on the
6708 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6709 connections (as happens on some IP stacks) was logged at start up time as
6710 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6711 it from "IPv6 and IPv4", which means that two separate sockets are being
6714 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6715 says whether AF_INET or AF_INET6 was passed as an argument.
6717 42. Exiwhat output was messed up when time zones were included in log
6720 43. Exiwhat now gives more information about the daemon's listening ports,
6721 and whether -tls-on-connect was used.
6723 44. The "port" option of the smtp transport is now expanded.
6725 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6726 silently ignored. Now an error message is written to the main and panic
6729 46. There's a new ACL modifier called "logwrite" which writes to a log file
6730 as soon as it is encountered.
6732 47. Added $local_user_uid and $local_user_gid at routing time.
6734 48. Exim crashed when trying to verify a sender address that was being
6737 49. Exim was recognizing only a space character after ".include". It now also
6738 recognizes a tab character.
6740 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6741 extracting the relevant information from the specification. The man page no
6742 longer contains scrambled data for the -d option, and I've added a section
6743 at the front about calling Exim under different names.
6745 51. Added "extra_headers" argument to the "mail" command in filter files.
6747 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6750 53. Installed eximstats 1.29.
6752 54. Added transport_filter_timeout as a generic transport option.
6754 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6755 Cc: header lines. This was required by RFC 822, but it not required by RFC
6758 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6759 incoming messages that did not have them. Now it does so only if the
6760 message originates locally, that is, if there is no associated remote host
6761 address. When Resent- header lines are present, this applies to the Resent-
6762 lines rather than the non-Resent- lines.
6764 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6765 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6767 58. Messages for configuration errors now include the name of the main
6768 configuration file - useful now that there may be more than one file in a
6769 list (.included file names were always shown).
6771 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6772 for those rare installations that do not start the daemon as root or run it
6773 setuid root. I've cut out the call to initgroups() if the daemon is not
6776 60. The Exim user and group can now be bound into the binary as text strings
6777 that are looked up at the start of Exim's processing.
6779 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6781 62. Added $mailstore_basename variable.
6783 63. Installed patch to sieve.c from Michael Haardt.
6785 64. When Exim failed to open the panic log after failing to open the main log,
6786 the original message it was trying to log was written to stderr and debug
6787 output, but if they were not available (the usual case in production), it
6788 was lost. Now it is written to syslog before the two lines that record the
6789 failures to open the logs.
6791 65. Users' Exim filters run in subprocesses under the user's uid. It is
6792 possible for a "deliver" command or an alias in a "personal" command to
6793 provoke an address rewrite. If logging of address rewriting is configured,
6794 this fails because the process is not running as root or exim. There may be
6795 a better way of dealing with this, but for the moment (because 4.30 needs
6796 to be released), I have disabled address rewrite logging when running a
6797 filter in a non-root, non-exim process.
6803 1. The buildconfig auxiliary program wasn't quoting the value set for
6804 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6805 not defined. This bug was masked in 4.22 by the effect that was fixed in
6808 2. Some messages that were rejected after a message id was allocated were
6809 shown as "incomplete" by exigrep. It no longer does this for messages that
6810 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6812 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6813 have allow_domain_literals set, the ID did not get logged in the <= line.
6814 Domain literals are now always recognized in Message-ID: header lines.
6816 4. The first argument for a ${extract expansion item is the key name or field
6817 number. Leading and trailing spaces in this item were not being ignored,
6818 causing some misleading effects.
6820 5. When deliver_drop_privilege was set, single queue runner processes started
6821 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6822 same command in the process it spins off) were not dropping privilege.
6824 6. When the daemon running as "exim" started a queue runner, it always
6825 re-executed Exim in the spun-off process. This is a waste of effort when
6826 deliver_drop_privilege is set. The new process now just calls the
6827 queue-runner function directly.
6833 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6836 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6837 ignored. Though the use of -oP was forcing the writing of a pid file, it
6838 was always written to the default place.
6840 3. If the message "no IP address found for host xxxx" is generated during
6841 incoming verification, it is now followed by identification of the incoming
6842 connection (so you can more easily find what provoked it).
6844 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6846 5. Added some features to "harden" Exim a bit more against certain attacks:
6848 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6849 be put in Local/Makefile. This is like the never_users runtime option,
6850 but it cannot be overridden. The default setting is "root".
6852 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6853 prefix string with which any file named in a -C command line option
6856 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6857 is retained for -C and -D only if the caller of Exim is root. Without
6858 it, the exim user may also use -C and -D and retain privilege.
6860 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6861 command line option is disabled.
6863 6. Macro names set by the -D option must start with an upper case letter, just
6864 like macro names defined in the configuration file.
6866 7. Added "dereference=" facility to LDAP.
6868 8. Two instances of the typo "uknown" in the source files are fixed.
6870 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6871 the Configure-Makefile script screwed up while processing it.
6873 10. Incorporated PCRE 4.4.
6875 11. The SMTP synchronization check was not operating right at the start of an
6876 SMTP session. For example, it could not catch a HELO sent before the client
6877 waited for the greeting. There is now a check for outstanding input at the
6878 point when the greeting is written. Because of the duplex, asynchronous
6879 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6880 way, but not yet received, when the check is performed.
6882 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6883 on TCP/IP sockets, because this apparently causes some broken clients to
6886 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6887 unchanged) from the Cygwin maintainer.
6889 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6890 was supported instead of testing for mbx. Effectively a typo.
6892 15. The spa authenticator server code was not checking that the input it
6893 received was valid base64.
6895 16. The debug output line for the "set" modifier in ACLs was not showing the
6896 name of the variable that was being set.
6898 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6900 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6901 lookup on its own. Something else was needed to trigger the lookup. For
6902 example, a match in host_lookup or the need for a host name in a host list.
6903 Now, if $sender_host_name is referenced and the host name has not yet been
6904 looked up, a lookup is performed. If the lookup fails, the variable remains
6905 empty, and $host_lookup_failed is set to "1".
6907 19. Added "eqi" as a case-independent comparison operator.
6909 20. The saslauthd authentication condition could segfault if neither service
6910 nor realm was specified.
6912 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6913 error message that was logged was misleading, and incoming SMTP
6914 connections were dropped. The message is now more accurate, and temporary
6915 errors are given to SMTP connections.
6917 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6918 (except RSET) with a 421 error, until QUIT is received. However, it was
6919 failing to send a response to QUIT.
6921 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6922 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6923 ACL is now run after the helo_try_verify_hosts code.
6925 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6926 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6927 case-independent, so other case variants are also recognized). Apparently
6928 some systems use these upper case variants.
6930 25. If more than two messages were waiting for the same host, and a transport
6931 filter was specified for the transport, Exim sent two messages over the
6932 same TCP/IP connection, and then failed with "socket operation on non-
6933 socket" when it tried to send the third.
6935 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6937 27. The extern definition of crypt16() in expand.c was not being excluded when
6938 the OS had its own crypt16() function.
6940 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6941 as a preferred synonym for return_size_limit, both as an option and as an
6944 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6946 30. Changed the default configuration ACL to relax the local part checking rule
6947 for addresses that are not in any local domains. For these addresses,
6948 slashes and pipe symbols are allowed within local parts, but the sequence
6949 /../ is explicitly forbidden.
6951 31. SPA server authentication was not clearing the challenge buffer before
6954 32. log_message in a "warn" ACL statement was writing to the reject log as
6955 well as to the main log, which contradicts the documentation and doesn't
6956 seem right (because no rejection is happening). So I have stopped it.
6958 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6959 However, I am unable to do any testing of this.
6961 34. Fixed an infelicity in the appendfile transport. When checking directories
6962 for a mailbox, to see if any needed to be created, it was accidentally
6963 using path names with one or more superfluous leading slashes; tracing
6964 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6966 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6967 contents are added to the log line that is written for every discarded
6968 recipient. (Previously a log_message setting was ignored.)
6970 36. The ${quote: operator now quotes the string if it is empty.
6972 37. The install script runs exim in order to find its version number. If for
6973 some reason other than non-existence or emptiness, which it checks, it
6974 could not run './exim', it was installing it with an empty version number,
6975 i.e. as "exim-". This error state is now caught, and the installation is
6978 38. An argument was missing from the function that creates an error message
6979 when Exim fails to connect to the socket for saslauthd authentication.
6980 This could cause Exim to crash, or give a corrupted message.
6982 39. Added isip, isip4, and isip6 to ${if conditions.
6984 40. The ACL variables $acl_xx are now saved with the message, and can be
6985 accessed later in routers, transports, and filters.
6987 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6988 strings in the file are not string-expanded.
6990 42. If a MAIL command specified a SIZE value that was too large to fit into an
6991 int variable, the check against message_size_limit failed. Such values are
6992 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6993 day this will have to be increased, but I don't think I want to be around
6994 when emails are that large.
7001 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7002 iconv() is not standard in FreeBSD.
7004 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7005 IPv6 enabled. The observed symptom was a segmentation fault on return from
7006 the function os_common_find_running_interfaces() in src/os.c.
7008 3. In the check_special_case() function in daemon.c I had used "errno" as an
7009 argument name, which causes warnings on some systems. This was basically a
7010 typo, since it was named "eno" in the comments!
7012 4. The code that waits for the clock to tick (at a resolution of some fraction
7013 of a second) so as to ensure message-id uniqueness was always waiting for
7014 at least one whole tick, when it could have waited for less. [This is
7015 almost certainly not relevant at current processor speeds, where it is
7016 unlikely to ever wait at all. But we try to future-proof.]
7018 5. The function that sleeps for a time interval that includes fractions of a
7019 second contained a race. It did not block SIGALRM between setting the
7020 timer, and suspending (a couple of lines later). If the interval was short
7021 and the sigsuspend() was delayed until after it had expired, the suspension
7022 never ended. On busy systems this could lead to processes getting stuck for
7025 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7026 process, before it forks any delivery processes. The open lookup caching
7027 mechanism meant that the open file or database connection was passed into
7028 the delivery process. The problem was that delivery processes always tidy
7029 up cached lookup data. This could cause a problem for the next delivery
7030 process started by the queue runner, because the external queue runner
7031 process does not know about the closure. So the next delivery process
7032 still has data in the lookup cache. In the case of a file lookup, there was
7033 no problem because closing a file descriptor in a subprocess doesn't affect
7034 the parent. However, if the lookup was caching a connection to a database,
7035 the connection was closed, and the second delivery process was likely to
7036 see errors such as "PGSQL: query failed: server closed the connection
7037 unexpectedly". The problem has been fixed by closing all cached lookups
7038 in a queue runner before running a delivery process.
7040 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7041 seem to have the "const" qualifier which it has on other OS. I've
7044 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7045 *of the same type* (client or server) with the same public name that an
7046 error should be diagnosed.
7048 9. When Exim looked up a host name for an IP address, but failed to find the
7049 original IP address when looking up the host name (a safety check), it
7050 output the message "<ip address> does not match any IP for NULL", which was
7051 confusing, to say the least. The bug was that the host name should have
7052 appeared instead of "NULL".
7054 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7055 user that is built into the binary, and the -C or -D options is used, root
7056 privilege is dropped before the configuration file is read. In addition,
7057 logging is switched to stderr instead of the normal log files. If the
7058 configuration then re-defines the Exim user, the unprivileged environment
7059 is probably not what is expected, so Exim logs a panic warning message (but
7062 However, if deliver_drop_privilege is set, the unprivileged state may well
7063 be exactly what is intended, so the warning has been cut out in that case,
7064 and Exim is allowed to try to write to its normal log files.
7070 1. smtp_return_error_details was not giving details for temporary sender
7071 or receiver verification errors.
7073 2. Diagnose a configuration error if two authenticators have the same public
7076 3. Exim used not to create the message log file for a message until the first
7077 delivery attempt. This could be confusing when incoming messages were held
7078 for policy or load reasons. The message log file is now created at the time
7079 the message is received, and an initial "Received" line is written to it.
7081 4. The automatically generated man page for command line options had a minor
7082 bug that caused no ill effects; however, a more serious problem was that
7083 the procedure for building the man page automatically didn't always
7084 operate. Consequently, release 4.20 contains an out-of-date version. This
7085 shouldn't happen again.
7087 5. When building Exim with embedded Perl support, the script that builds the
7088 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7089 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7091 6. The freeze_tell option was not being used for messages that were frozen on
7092 arrival, either by an ACL or by local_scan().
7094 7. Added the smtp_incomplete_transaction log selector.
7096 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7097 was accepting AUTH without a new EHLO.
7099 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7100 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7103 10. Logging of TCP/IP connections (when configured) now happens in the main
7104 daemon process instead of the child process, so that the TCP/IP connection
7105 count is more accurate (but it can never be perfect).
7107 11. The use of "drop" in a nested ACL was not being handled correctly in the
7108 outer ACL. Now, if condition failure induced by the nested "drop" causes
7109 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7110 or "require"), the connection is dropped.
7112 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7113 that yield "discard" can now be used with an "accept" or a "discard" verb,
7114 but an error is generated for any others (because I can't see a useful way
7115 to define what should happen).
7117 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7118 are now processed in the same way as lines in the Exim configuration file.
7119 In particular, continuation lines are supported.
7121 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7123 15. Added -ti meaning -t -i.
7125 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7126 domains, and warn by logging if others are found.
7128 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7129 returned from GIFCONF when Exim is trying to find the list of interfaces on
7130 a host. The code in os.c has been modified to copy each ifreq to an aligned
7131 structure in all cases.
7133 Also, in some cases, the returned ifreq's were being copied to a 'struct
7134 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7135 means the last couple of bytes of an IPv6 address could be chopped if the
7136 ifreq contained only a normal sockaddr (14 bytes storage).
7138 18. Named domain lists were not supported in the hosts_treat_as_local option.
7139 An entry such as +xxxx was not recognized, and was treated as a literal
7142 19. Ensure that header lines added by a DATA ACL are included in the reject log
7143 if the ACL subsequently rejects the message.
7145 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7146 MD5 (which is deprecated).
7148 21. When testing a filter file using -bf, Exim was writing a message when it
7149 took the sender from a "From " line in the message, but it was not doing so
7150 when it took $return_path from a Return-Path: header line. It now does.
7152 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7153 with a valid header line field name (a series of printing characters
7154 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7156 23. Changed "disc" in the source to "disk" to conform to the documentation and
7157 the book and for uniformity.
7159 24. Ignore Sendmail's -Ooption=value command line item.
7161 25. When execve() failed while trying to run a command in a pipe transport,
7162 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7163 could be confused with a return value of 69 from the command itself. This
7164 has been changed to 127, the value the shell returns if it is asked to run
7165 a non-existent command. The wording for the related log line suggests a
7166 non-existent command as the problem.
7168 26. If received_header_text expands to an empty string, do not add a Received:
7169 header line to the message. (Well, it adds a token one on the spool, but
7170 marks it "old" so that it doesn't get used or transmitted.)
7172 27. Installed eximstats 1.28 (addition of -nt option).
7174 28. There was no check for failure on the call to getsockname() in the daemon
7175 code. This can fail if there is a shortage of resources on the system, with
7176 ENOMEM, for example. A temporary error is now given on failure.
7178 29. Contrary to the C standard, it seems that in some environments, the
7179 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7180 program. Exim now does this explicitly; it affects the formatting of
7181 timestamps using strftime().
7183 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7184 complaints. I've now initialized all the variables, to avoid this.
7186 32. Header lines added by a system filter were not being "seen" during
7187 transport-time rewrites.
7189 33. The info_callback() function passed to OpenSSL is set up with type void
7190 (*)(SSL *, int, int), as described somewhere. However, when calling the
7191 function (actually a macro) that sets it up, the type void(*)() is
7192 expected. I've put in a cast to prevent warnings from picky compilers.
7194 34. If a DNS black list lookup found a CNAME record, but there were no A
7195 records associated with the domain it pointed at, Exim crashed.
7197 35. If a DNS black list lookup returned more than one A record, Exim ignored
7198 all but the first. It now scans all returned addresses if a particular IP
7199 value is being sought. In this situation, the contents of the
7200 $dnslist_value variable are a list of all the addresses, separated by a
7203 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7204 to accept a host name and all its aliases if the forward lookup for any of
7205 them yielded the IP address of the incoming connection. Now it accepts only
7206 those names whose forward lookup yields the correct IP address. Any other
7207 names are discarded. This closes a loophole whereby a rogue DNS
7208 administrator could create reverse DNS records to break through a
7209 wildcarded host restriction in an ACL.
7211 37. If a user filter or a system filter that ran in a subprocess used any of
7212 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7213 the wrong values were passed to the pipe command ($thisaddress had the
7214 value of $0, $0 had the value of $1, etc). This bug was introduced by
7215 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7217 38. Improved the line breaking for long SMTP error messages from ACLs.
7218 Previously, if there was no break point between 40 and 75 characters, Exim
7219 left the rest of the message alone. Two changes have been made: (a) I've
7220 reduced the minimum length to 35 characters; (b) if it can't find a break
7221 point between 35 and 75 characters, it looks ahead and uses the first one
7222 that it finds. This may give the occasional overlong line, but at least the
7223 remaining text gets split now.
7225 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7226 file descriptors might be low, and that setting 1000 would always raise it.
7227 It turns out that in some environments, the limit is already over 1000 and
7228 that lowering it causes trouble. So now Exim takes care not to decrease it.
7230 40. When delivering a message, the value of $return_path is set to $sender_
7231 address at the start of routing (routers may change the value). By an
7232 oversight, this default was not being set up when an address was tested by
7233 -bt or -bv, which affected the outcome if any router or filter referred to
7236 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7237 log only when "message" or "log_message" are set. However, if one of the
7238 conditions was an address verification, or a call to a nested ACL, the
7239 messages generated by the underlying test were being passed through. This
7240 no longer happens. The underlying message is available in $acl_verify_
7241 message for both "message" and "log_message" expansions, so it can be
7242 passed through if needed.
7244 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7245 new expansion $bh_ to give the encoded byte string without charset
7246 translation. Translation happens only if iconv() is available; HAVE_ICONV
7247 indicates this at build time. HEADERS_CHARSET gives the charset to
7248 translate to; headers_charset can change it in the configuration, and
7249 "headers charset" can change it in an individual filter file.
7251 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7252 that creates RFC 2047 encoded "words" labels them as that charset instead
7253 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7254 expansion operator; (ii) when Exim creates a From: line for a local
7255 message; (iii) when a header line is rewritten to include a "phrase" part.
7257 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7258 buggy, causing it to skip the first lines of messages whose message ID
7259 ended in 'D'. This would not have bitten before Exim release 4.14, because
7260 message IDs were unlikely to end in 'D' before then. The effect was to have
7261 incorrect size information for certain domains.
7263 45. #include "config.h" was missing at the start of the crypt16.c module. This
7264 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7267 46. If there was a timeout during a "random" callout check, Exim treated it as
7268 a failure of the random address, and carried on sending RSET and the real
7269 address. If the delay was just some slowness somewhere, the response to the
7270 original RCPT would be taken as a response to RSET and so on, causing
7271 mayhem of various kinds.
7273 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7274 when I implemented it. It didn't allow for the fact that some option values
7275 may legitimately be negative (e.g. size_addition), and it didn't even do
7276 the right test for positive values.
7278 48. Domain names in DNS records are case-independent. Exim always looks them up
7279 in lower case. Some resolvers return domain names in exactly the case they
7280 appear in the zone file, that is, they may contain uppercase letters. Not
7281 all resolvers do this - some return always lower case. Exim was treating a
7282 change of case by a resolver as a change of domain, similar to a widening
7283 of a domain abbreviation. This triggered its re-routing code and so it was
7284 trying to route what was effectively the same domain again. This normally
7285 caused routing to fail (because the router wouldn't handle the domain
7286 twice). Now Exim checks for this case specially, and just changes the
7287 casing of the domain that it ultimately uses when it transmits the message
7290 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7293 50. If a filter generated a file delivery with a non-absolute name (possible if
7294 no home directory exists for the router), the forbid_file option was not
7297 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7298 the existing equality matching.
7300 52. Exim was using ints instead of ino_t variables in some places where it was
7301 dealing with inode numbers.
7303 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7304 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7305 called TMPDIR, and if it finds it is different, it changes its value.
7307 54. The smtp_printf() function is now made available to local_scan() so
7308 additional output lines can be written before returning. There is also an
7309 smtp_fflush() function to enable the detection of a dropped connection.
7310 The variables smtp_input and smtp_batched_input are exported to
7313 55. Changed the default runtime configuration: the message "Unknown user"
7314 has been removed from the ACL, and instead placed on the localuser router,
7315 using the cannot_route_message feature. This means that any verification
7316 failures that generate their own messages won't get overridden. Similarly,
7317 the "Unrouteable address" message that was in the ACL for unverifiable
7318 relay addresses has also been removed.
7320 56. Added hosts_avoid_esmtp to the smtp transport.
7322 57. The exicyclog script was not checking for the esoteric option
7323 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7324 will work only if exicyclog is run under the appropriate euid.
7326 58. Following a discussion on the list, the rules by which Exim recognises line
7327 endings on incoming messages have been changed. The -dropcr and drop_cr
7328 options are now no-ops, retained only for backwards compatibility. The
7329 following line terminators are recognized: LF CRLF CR. However, special
7330 processing applies to CR:
7332 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7333 nor a local message in the state where . is a terminator.
7335 (ii) If a bare CR is encountered in a header line, an extra space is added
7336 after the line terminator so as not to end the header. The reasoning
7337 behind this is that bare CRs in header lines are most likely either
7338 to be mistakes, or people trying to play silly games.
7340 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7341 was being incorrectly given as 18 bytes larger than it should have been.
7342 This is a VOB (very old bug).
7344 60. This may never have affected anything current, but just in case it has:
7345 When the local host is found other than at the start of a list of hosts,
7346 the local host, those with the same MX, and any that follow, are discarded.
7347 When the list in question was part of a longer list of hosts, the following
7348 hosts (not currently being processed) were also being discarded. This no
7349 longer happens. I'm not sure if this situation could ever has previously
7352 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7355 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7356 header. This is something that is recommended in a new Internet Draft, and
7357 is something that is documented as being done by Sendmail. There are two
7358 possible values. For messages generated by the autoreply transport, Exim
7361 Auto-Submitted: auto-replied
7363 whereas for all other generated messages (e.g. bounces) it adds
7365 Auto-Submitted: auto-generated
7367 63. The "personal" condition in filters now includes a test for the
7368 Auto-Submitted: header. If it contains the string "auto-" the message it
7369 not considered personal.
7371 64. Added rcpt_include_affixes as a generic transport option.
7373 65. Added queue_only_override (default true).
7375 66. Added the syslog_duplication option.
7377 67. If what should have been the first header line of a message consisted of
7378 a space followed by a colon, Exim was mis-interpreting it as a header line.
7379 It isn't of course - it is syntactically invalid and should therefore be
7380 treated as the start of the message body. The misbehaviour could have
7381 caused a number of strange effects, including loss of data in subsequent
7382 header lines, and spool format errors.
7384 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7385 client host had authenticated. This control can now be exercised by an ACL
7386 for more flexibility.
7388 69. By default, callouts do not happen when testing with -bh. There is now a
7389 variant, -bhc, which does actually run the callout code, including
7390 consulting and updating the callout cache.
7392 70. Added support for saslauthd authentication, courtesy of Alexander
7395 71. If statvfs() failed on the spool or log directories while checking their
7396 size for availability, Exim confusingly gave the error "space shortage".
7397 Furthermore, in debugging mode it crashed with a floating point exception.
7398 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7399 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7400 serious problem, Exim now writes to the main and panic logs when this
7401 happens, with details of the failure. It then refuses to accept the
7402 incoming message, giving the message "spool directory problem" or "log
7403 directory problem" with a 421 code for SMTP messages.
7405 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7406 0, 1, and 2 exist, because some OS complain for execs without them (see
7407 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7408 descriptors. However, the code omitted to check that the open succeeded,
7409 causing mysterious errors if for some reason the permissions on /dev/null
7410 got screwed. Now Exim writes a message to the main and panic logs, and
7411 bombs out if it can't open /dev/null.
7413 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7414 interact so that it is all more flexible. It is supposed to remain
7415 backwards compatible. Also added extra_local_interfaces.
7417 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7418 to bomb out with an assertion failure - to the client this appears as a
7419 connection drop. This problem occurs in the part of the code that was taken
7420 from the Samba project. Fortunately, the assertion is in a very simple
7421 function, so I have fixed this by reproducing the function inline in the
7422 one place where it is called, and arranging for authentication to fail
7423 instead of killing the process with assert().
7425 75. The SPA client code was not working when the server requested OEM rather
7426 than Unicode encoding.
7428 76. Added code to make require_files with a specific uid setting more usable in
7429 the case where statting the file as root fails - usually a non-root-mounted
7430 NFS file system. When this happens and the failure is EACCES, Exim now
7431 forks a subprocess and does the per-uid checking as the relevant uid.
7433 77. Added process_log_path.
7435 78. If log_file_path was not explicitly set, a setting of check_log_space or
7436 check_log_inodes was ignored.
7438 79. If a space check for the spool or log partitions fails, the incident is now
7439 logged. Of course, in the latter case the data may get lost...
7441 80. Added the %p formatting code to string_format() so that it can be used to
7442 print addresses in debug_print(). Adjusted all the address printing in the
7443 debugging in store.c to use %p rather than %d.
7445 81. There was a concern that a line of code in smtp_in.c could overflow a
7446 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7447 initially expressed, the concern was not well-founded, because trailing
7448 spaces are removed early. However, if the trailing spaces were followed by
7449 a NULL, they did not get removed, so the overflow was possible. Two fixes
7452 (a) I re-wrote the offending code in a cleaner fashion.
7453 (b) If an incoming SMTP command contains a NULL character, it is rejected
7456 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7457 runs initgroups(), so that if the Exim user is in any additional groups,
7458 they will be used during message reception.
7464 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob