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/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
74 shared (NFS) environment.
76 HS/03 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
105 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
106 A single TCP connection by a client will now hold a TLS connection open
107 for multiple message deliveries, by default. Previoud the default was to
110 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
111 default. If built with the facility, DANE will be used. The facility
112 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
114 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
115 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
116 must be defined and you must still, unless you define DISABLE_TLS, manage
117 the the include-dir and library-file requirements that go with that
118 choice. Non-TLS builds are still supported.
120 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
123 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
124 default. If the platform supports and has the facility enabled, it will
125 be requested on all coneections.
127 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
128 controlled by the build-time option SUPPORT_PIPE_CONNECT.
130 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
132 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
133 success-DSN messages. Previously the From: header was always the default
134 one for these; the option was ignored.
136 JH/28 Fix the timeout on smtp response to apply to the whole response.
137 Previously it was reset for every read, so a teergrubing peer sending
138 single bytes within the time limit could extend the connection for a
139 long time. Credit to Qualsys Security Advisory Team for the discovery.
141 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
142 delivery address, which leaked information of the results of local
143 forwarding. Change to the original envelope recipient address, per
146 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
147 requested. Previously not bounce was generated and a log entry of
148 error ignored was made.
150 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
152 JH/32 Introduce a general tainting mechanism for values read from the input
153 channel, and values derived from them. Refuse to expand any tainted
154 values, to catch one form of exploit.
156 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
157 was unused and the unexpanded text used for the test. Found and
158 fixed by Ruben Jenster.
160 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
161 an attempt to use a TLS library read routine dereffed a nul pointer,
164 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
165 them in our smtp response.
167 JH/36 Have the general router option retry_use_local_part default to true when
168 any of the restrictive preconditions are set (to anything). Previously it
169 was only for check_local user. The change removes one item of manual
170 configuration which is required for proper retries when a remote router
171 handles a subset of addresses for a domain.
173 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
174 link count into consideration.
176 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
177 caused the extension of big_buffer, the following lines were ignored.
179 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
180 accordance with RFC 2308. Previously there was no expiry, so a longlived
181 receive process (eg. due to ACL delays) versus a short SOA value could
184 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
186 JH/39 Promote DMARC support to mainline.
192 JH/01 Remove code calling the customisable local_scan function, unless a new
193 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
195 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
196 non-signal-safe functions being used.
198 JH/03 Bug 2269: When presented with a received message having a stupidly large
199 number of DKIM-Signature headers, disable DKIM verification to avoid
200 a resource-consumption attack. The limit is set at twenty.
202 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
203 report of oldest_pass in ${authres } in consequence, and separate out
204 some descriptions of reasons for verification fail.
206 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
207 files in the spool were present and unlocked. A queue-runner could spot
208 them, resulting in a duplicate delivery. Fix that by doing the unlock
209 after the unlink. Investigation by Tim Stewart. Take the opportunity to
210 add more error-checking on spoolfile handling while that code is being
213 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
214 No known attacks, no CVE, this is defensive hardening.
216 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
217 a queue-runner could start a delivery while other operations were ongoing.
218 Cutthrough delivery was a common victim, resulting in duplicate delivery.
219 Found and investigated by Tim Stewart. Fix by using the open message data
220 file handle rather than opening another, and not locally closing it (which
221 releases a lock) for that case, while creating the temporary .eml format
222 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
224 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
225 $sender_verify_failure/$recipient_verify_failure to "random".
227 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
230 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
231 Previously this would segfault.
233 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
236 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
237 like zero, since the resolver should be doing this for us, But we need one
238 as a CNAME but no MX presence gets the CNAME returned; we need to check
239 that doesn't point to an MX to declare it "no MX returned" rather than
240 "error, loop". A new main option is added so the older capability of
241 following some limited number of chain links is maintained.
243 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
245 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
246 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
247 now-deprecated earlier definitions used only specified the range up to TLS
248 1.2 (in the older-version library docs).
250 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
252 JH/15 Rework TLS client-side context management. Stop using a global, and
253 explicitly pass a context around. This enables future use of TLS for
254 connections to service-daemons (eg. malware scanning) while a client smtp
255 connection is using TLS; with cutthrough connections this is quite likely.
257 JH/16 Fix ARC verification to do AS checks in reverse order.
259 JH/17 Support a "tls" option on the ${readsocket } expansion item.
261 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
262 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
263 Previously the "utf8" would be re-prepended for every additional message.
265 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
266 Previously thery were accepted, resulting in issues when attempting to
267 forward messages to a non-supporting MTA.
269 PP/02 Let -n work with printing macros too, not just options.
271 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
272 one parent address was copied, and bogus data was used at delivery-logging
273 time. Either a crash (after delivery) or bogus log data could result.
274 Discovery and analysis by Tim Stewart.
276 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
277 Previously if the string ended mid-character, we did not insert the
278 promised '?' replacement.
280 PP/04 Documentation: current string operators work on bytes, not codepoints.
282 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
283 should pack well giving a smaller memory footprint so better caching and
284 therefore performance. Group the declarations where this can't be done so
285 that the byte-sized flag variables are not interspersed among pointer
286 variables, giving a better chance of good packing by the compiler.
288 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
289 non-null, to avoid issues with sites running BATV. Previously reports were
290 sent with an empty envelope sender so looked like bounces.
292 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
293 The ignore_error flag wasn't being returned from the filter subprocess so
294 was not set for later routers. Investigation and fix by Matthias Kurz.
296 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
297 and a msg:complete for the whole, when a message is manually removed using
298 -Mrm. Developement by Matthias Kurz, hacked on by JH.
300 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
301 a "Gnu special" function, asprintf() in the DB utility binary builds; I
302 hope that is portable enough.
304 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
305 requiring a known-CA anchor certificate; make it now rely entirely on the
306 TLSA as an anchor. Checking the name on the leaf cert against the name
307 on the A-record for the host is still done for TA (but not for EE mode).
309 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
310 included in delivery lines for non-proxied connections, when compiled with
311 SUPPORT_SOCKS and running with proxy logging enabled.
313 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
314 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
315 move the existing event to fire before the normal logging of message
316 failure so that custom logging is bracketed by normal logging.
318 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
319 msg:fail:internal event. Developement by Matthias Kurz.
321 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
322 far too small for todays use of crypto signatures stored there. Go all
323 the way to the max DNS message size of 64kB, even though this might be
324 overmuch for IOT constrained device use.
326 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
327 copy a string over itself. The library routine is documented as not
328 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
330 JH/32 For main options check_spool_space and check_inode_space, where the
331 platform supports 64b integers, support more than the previous 2^31 kB
332 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
333 the previous G, M, k.
335 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
336 $authenticated_fail_id variable on authentication failure. Previously
339 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
340 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
341 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
342 GNUTLS_SEC_PARAM_MEDIUM.
344 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
345 side. Previously we would continue as if no SNI had been received.
347 JH/36 Harden the handling of string-lists. When a list consisted of a sole
348 "<" character, which should be a list-separator specification, we walked
349 off past the nul-terimation.
351 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
352 causes) even when the retry time is not yet met. Previously they were
353 not, meaning that when (say) an account was over-quota and temp-rejecting,
354 and multiple senders' messages were queued, only one sender would get
355 notified on each configured delay_warning cycle.
357 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
359 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
360 cipher-suites, an error can be left on the stack even for a succeeding
361 accept; this results in impossible error messages when a later operation
364 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
365 return error codes indicating retry. Under TLS1.3 this becomes required.
367 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
368 it only wrote the new authenticators, resulting in a lack of tracking of
369 peer changes of ESMTP extensions until the next cache flush.
371 JH/41 Fix the loop reading a message header line to check for integer overflow,
372 and more-often against header_maxsize. Previously a crafted message could
373 induce a crash of the recive process; now the message is cleanly rejected.
375 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
376 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
382 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
383 When redis_servers is set to a list of > 1 element, and the Redis servers
384 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
385 case of MOVED into a DEFER case instead, thus moving the query onto the
386 next server in the list. For a cluster of N elements, all N servers must
387 be defined in redis_servers.
389 GF/02 Catch and remove uninitialized value warning in exiqsumm
390 Check for existence of @ARGV before looking at $ARGV[0]
392 JH/01 Replace the store_release() internal interface with store_newblock(),
393 which internalises the check required to safely use the old one, plus
394 the allocate and data copy operations duplicated in both (!) of the
395 extant use locations.
397 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
398 modifier. This matches the restriction on the commandline.
400 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
401 Previously only the last row was returned.
403 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
404 we assumed that tags in the header were well-formed, and parsed the
405 element content after inspecting only the first char of the tag.
406 Assumptions at that stage could crash the receive process on malformed
409 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
410 While running the DKIM ACL we operate on the Permanent memory pool so that
411 variables created with "set" persist to the DATA ACL. Also (at any time)
412 DNS lookups that fail create cache records using the Permanent pool. But
413 expansions release any allocations made on the current pool - so a dnsdb
414 lookup expansion done in the DKIM ACL releases the memory used for the
415 DNS negative-cache, and bad things result. Solution is to switch to the
416 Main pool for expansions.
417 While we're in that code, add checks on the DNS cache during store_reset,
418 active in the testsuite.
419 Problem spotted, and debugging aided, by Wolfgang Breyha.
421 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
422 When none of the hosts presented to a transport match an already-open
423 connection, close it and proceed with the list. Previously we would
424 queue the message. Spotted by Lena with Yahoo, probably involving
427 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
428 Previously a spurious "250 OK id=" response was appended to the proper
431 JH/08 The "support for" informational output now, which built with Content
432 Scanning support, has a line for the malware scanner interfaces compiled
433 in. Interface can be individually included or not at build time.
435 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
436 by the template makefile "src/EDITME". The "STREAM" support for an older
437 ClamAV interface method is removed.
439 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
440 rows affected is given instead).
442 JH/11 The runtime Berkeley DB library version is now additionally output by
443 "exim -d -bV". Previously only the compile-time version was shown.
445 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
446 SMTP connection. Previously, when one had more receipients than the
447 first, an abortive onward connection was made. Move to full support for
448 multiple onward connections in sequence, handling cutthrough connection
449 for all multi-message initiating connections.
451 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
452 routers. Previously, a multi-recipient message would fail to match the
453 onward-connection opened for the first recipient, and cause its closure.
455 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
456 a timeout on read on a GnuTLS initiating connection, resulting in the
457 initiating connection being dropped. This mattered most when the callout
458 was marked defer_ok. Fix to keep the two timeout-detection methods
461 JH/15 Relax results from ACL control request to enable cutthrough, in
462 unsupported situations, from error to silently (except under debug)
463 ignoring. This covers use with PRDR, frozen messages, queue-only and
466 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
468 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
469 metadata, resulting in a crash in free().
471 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
472 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
473 Broken also in d185889f4, with init system revamp.
475 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
476 Previously we abruptly closed the connection after reading a malware-
477 found indication; now we go on to read the "scan ok" response line,
480 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
481 ACL. Previously, a crash would result.
483 JH/19 Speed up macro lookups during configuration file read, by skipping non-
484 macro text after a replacement (previously it was only once per line) and
485 by skipping builtin macros when searching for an uppercase lead character.
487 JH/20 DANE support moved from Experimental to mainline. The Makefile control
488 for the build is renamed.
490 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
491 was allocated for every new TLS startup, meaning one per message. Fix
492 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
494 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
495 reported the original. Fix to report (as far as possible) the ACL
496 result replacing the original.
498 JH/23 Fix memory leak during multi-message connections using STARTTLS under
499 OpenSSL. Certificate information is loaded for every new TLS startup,
500 and the resources needed to be freed.
502 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
504 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
507 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
508 DATA response info to the (existing) per-recipient response info for
509 the "C=" log element. It can have useful tracking info from the
510 destination system. Patch from Simon Arlott.
512 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
513 length value. Previously this would segfault.
515 HS/02 Support Avast multiline protoocol, this allows passing flags to
516 newer versions of the scanner.
518 JH/28 Ensure that variables possibly set during message acceptance are marked
519 dead before release of memory in the daemon loop. This stops complaints
520 about them when the debug_store option is enabled. Discovered specifically
521 for sender_rate_period, but applies to a whole set of variables.
522 Do the same for the queue-runner and queue-list loops, for variables set
523 from spool message files. Do the same for the SMTP per-message loop, for
524 certain variables indirectly set in ACL operations.
526 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
527 as a multi-recipient message from a mailinglist manager). The coding had
528 an arbitrary cutoff number of characters while checking for more input;
529 enforced by writing a NUL into the buffer. This corrupted long / fast
530 input. The problem was exposed more widely when more pipelineing of SMTP
531 responses was introduced, and one Exim system was feeding another.
532 The symptom is log complaints of SMTP syntax error (NUL chars) on the
533 receiving system, and refused recipients seen by the sending system
534 (propating to people being dropped from mailing lists).
535 Discovered and pinpointed by David Carter.
537 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
538 replaced by the ${authresults } expansion.
540 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
542 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
543 allows proper process termination in container environments.
545 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
546 Previously the "final dot" had a newline after it; ensure it is CR,LF.
548 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
549 and "err_perm", deprecated since 4.83 when the RFC-defined words
550 "temperror" and "permerror" were introduced.
552 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
553 transport-filters or DKIM-signing. The restriction was lost in the
554 consolidation of verify-callout and delivery SMTP handling.
555 Extend the restriction to also cover ARC-signing.
557 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
558 in defer=pass mode supply a 450 to the initiator. Previously the message
561 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
562 tls_require_ciphers is used as before.
564 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
565 "pass_unscanned". Only tmpfails from the scanner are written to
566 the paniclog, as they may require admin intervention (permission
567 denied, license issues). Other scanner errors (like decompression
568 bombs) do not cause a paniclog entry.
570 JH/36 Fix reinitialisation of DKIM logging variable between messages.
571 Previously it was possible to log spurious information in receive log
574 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
575 triggered odd behaviour from Outlook Express clients.
577 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
580 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
581 since the IETF WG has not yet settled on that versus the original
582 "bare" representation.
584 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
585 Previously the millisecond value corrupted the output.
586 Fix also for syslog_pid=no and log_selector +pid, for which the pid
587 corrupted the output.
593 JH/01 Rework error string handling in TLS interface so that the caller in
594 more cases is responsible for logging. This permits library-sourced
595 string to be attached to addresses during delivery, and collapses
596 pairs of long lines into single ones.
598 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
599 during configuration. Wildcards are allowed and expanded.
601 JH/02 Rework error string handling in DKIM to pass more info back to callers.
602 This permits better logging.
604 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
605 do not close it down and have the child transport start it up again on
606 the passed-on TCP connection. Instead, proxy the child (and any
607 subsequent ones) for TLS via a unix-domain socket channel. Logging is
608 affected: the continued delivery log lines do not have any DNSSEC, TLS
609 Certificate or OCSP information. TLS cipher information is still logged.
611 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
612 identical IP addresses on different listening ports. Will also affect
615 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
616 add noisy ifdef guards to special-case this sillyness.
617 Patch from Bernd Kuhls.
619 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
620 than 255 are no longer allowed.
622 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
623 Disable the session-cache too, which might reduce our load. Since we
624 currrectly use a new context for every connection, both as server and
625 client, there is no benefit for these.
626 GnuTLS appears to not support tickets server-side by default (we don't
627 call gnutls_session_ticket_enable_server()) but client side is enabled
628 by default on recent versions (3.1.3 +) unless the PFS priority string
631 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
632 <https://reproducible-builds.org/specs/source-date-epoch/>.
634 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
635 the check for any unsuccessful recipients did not notice the limit, and
636 erroneously found still-pending ones.
638 JH/08 Pipeline CHUNKING command and data together, on kernels that support
639 MSG_MORE. Only in-clear (not on TLS connections).
641 JH/09 Avoid using a temporary file during transport using dkim. Unless a
642 transport-filter is involved we can buffer the headers in memory for
643 creating the signature, and read the spool data file once for the
644 signature and again for transmission.
646 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
647 4.77 as the kernel support then wasn't solid, having issues in 64bit
648 mode. Now, it's been long enough. Add support for FreeBSD also.
650 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
651 case where the routing stage had gathered several addresses to send to
652 a host before calling the transport for the first, we previously failed
653 to close down TLS in the old transport process before passing the TCP
654 connection to the new process. The new one sent a STARTTLS command
655 which naturally failed, giving a failed delivery and bloating the retry
656 database. Investigation and fix prototype from Wolfgang Breyha.
658 JH/12 Fix check on SMTP command input synchronisation. Previously there were
659 false-negatives in the check that the sender had not preempted a response
660 or prompt from Exim (running as a server), due to that code's lack of
661 awareness of the SMTP input buffering.
663 PP/04 Add commandline_checks_require_admin option.
664 Exim drops privileges sanely, various checks such as -be aren't a
665 security problem, as long as you trust local users with access to their
666 own account. When invoked by services which pass untrusted data to
667 Exim, this might be an issue. Set this option in main configuration
668 AND make fixes to the calling application, such as using `--` to stop
671 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
672 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
673 responses to those, into a single TLS record each way (this usually means
674 a single packet). As a side issue, smtp_enforce_sync now works on TLS
677 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
678 affects you only if you're dancing at the edge of the param size limits.
679 If you are, and this message makes sense to you, then: raise the
680 configured limit or use OpenSSL 1.1. Nothing we can do for older
683 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
684 cmdline element to get the documented default one. Previously it was
687 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
688 are unsuable for non-message-specific reasons - eg. connection timeout,
689 banner-time rejection.
691 JH/16 Fix logging of delivery remote port, when specified by router, under
694 PP/06 Repair manualroute's ability to take options in any order, even if one
695 is the name of a transport.
698 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
700 JH/17 Change the list-building routines interface to use the expanding-string
701 triplet model, for better allocation and copying behaviour.
703 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
704 Previously it was constructed the first time a possibly-matching string
705 was met in the configuration file input during startup; now it is done
708 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
709 compatible one, to avoid the (poorly documented) possibility of a config
710 file in the working directory redirecting the DB files, possibly correpting
711 some existing file. CVE-2017-10140 assigned for BDB.
713 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
714 cache-hot. Previously, although the result was properly cached, the
715 initial verify call returned a defer.
717 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
718 the main verify for receipient in uncached-mode.
720 JH/22 Retire historical build files to an "unsupported" subdir. These are
721 defined as "ones for which we have no current evidence of testing".
723 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
724 if present. Previously it was ignored.
726 JH/24 Start using specified-initialisers in C structure init coding. This is
727 a C99 feature (it's 2017, so now considered safe).
729 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
730 if was a fixed-sized field and bitmask ops via macros; it is now more
733 PP/07 GitHub PR 56: Apply MariaDB build fix.
734 Patch provided by Jaroslav Škarvada.
736 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
737 during Coverity cleanups [4.87 JH/47]
738 Diagnosis and fix provided by Michael Fischer v. Mollard.
740 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
741 the right size to place the terminating semicolon on its own folded
742 line, the header hash was calculated to an incorrect value thanks to
743 the (relaxed) space the fold became.
745 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
746 and confused the parent.
748 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
749 which could crash as a result. This could lead to undeliverable messages.
751 JH/28 Logging: "next input sent too soon" now shows where input was truncated
754 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
755 matters on fast-turnover and PID-randomising systems, which were getting
756 out-of-order delivery.
758 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
759 a possibly-overlapping copy. The symptom was that "Remote host closed
760 connection in response to HELO" was logged instead of the actual 4xx
763 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
764 Previously only that bufferd was discarded, resulting in SYMTP command
767 JH/32 DKIM: when a message has multiple signatures matching an identity given
768 in dkim_verify_signers, run the dkim acl once for each. Previously only
769 one run was done. Bug 2189.
771 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
772 "panic the current process" to "deliberately defer". The panic log is
773 still written with the problem list name; the mail and reject logs now
774 get a temp-reject line for the message that was being handled, saying
775 something like "domains check lookup or other defer". The SMTP 451
776 message is still "Temporary local problem".
778 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
779 A crafted sequence of BDAT commands could result in in-use memory beeing
780 freed. CVE-2017-16943.
782 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
783 from SMTP input. Previously it was always done; now only done for DATA
784 and not BDAT commands. CVE-2017-16944.
786 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
787 to the message (such as an overlong header line). Previously this was
788 not done and we did not exit BDAT mode. Followon from the previous item
789 though a different problem.
795 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
796 than -2003 did; needs libidn2 in addition to libidn.
798 JH/02 The path option on a pipe transport is now expanded before use.
800 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
801 Patch provided by "Björn", documentation fix added too.
803 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
804 missing a wire-to-host endian conversion.
806 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
807 close after a BDAT command line could be taken as a following command,
808 giving a synch failure. Fix by only checking for synch immediately
809 before acknowledging the chunk.
811 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
812 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
813 macro. Patches provided by Josh Soref.
815 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
816 Previously we did not; the RFC seems ambiguous and VRFY is not listed
817 by IANA as a service extension. However, John Klensin suggests that we
820 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
821 the dkim code may be unix-mode line endings rather than smtp wire-format
822 CRLF, so prepend a CR to any bare LF.
824 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
825 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
827 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
828 the first were themselves being wrongly included in the feed into dkim
829 processing; with most chunk sizes in use this resulted in an incorrect
830 body hash calculated value.
832 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
833 DKIM signature block, for verification. Although advised against by
834 standards it is specifically not ruled illegal.
836 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
838 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
839 missing a body hash (the bh= tag).
841 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
842 It seems that HAProxy sends the Proxy Protocol information in clear and
843 only then does a TLS startup, so do the same.
845 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
846 TCP connections (such as for Spamd) unless the daemon successfully set
847 Fast Open mode on its listening sockets. This fixes breakage seen on
848 too-old kernels or those not configured for Fast Open, at the cost of
849 requiring both directions being enabled for TFO, and TFO never being used
850 by non-daemon-related Exim processes.
852 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
853 endings, at least on the first header line. Try to canonify any that get
854 past that check, despite the cost.
856 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
857 now limited to an arbitrary five deep, while parsing addresses with the
858 strip_excess_angle_brackets option enabled.
860 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
861 instead leave the unprompted TLS handshake in socket buffer for the
862 TLS library to consume.
864 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
866 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
868 JH/16 Drop variables when they go out of scope. Memory management drops a whole
869 region in one operation, for speed, and this leaves assigned pointers
870 dangling. Add checks run only under the testsuite which checks all
871 variables at a store-reset and panics on a dangling pointer; add code
872 explicitly nulling out all the variables discovered. Fixes one known
873 bug: a transport crash, where a dangling pointer for $sending_ip_address
874 originally assigned in a verify callout, is re-used.
876 PP/06 Drop '.' from @INC in various Perl scripts.
878 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
880 PP/08 Reduce a number of compilation warnings under clang; building with
881 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
882 should be warning-free.
884 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
886 HS/01 Fix portability problems introduced by PP/08 for platforms where
887 realloc(NULL) is not equivalent to malloc() [SunOS et al].
889 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
890 chunk. This allows us to accept broken chunked messages. We need a more
891 general solution here.
893 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
894 already-broken messages in the queue.
896 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
898 JH/19 Fix reference counting bug in routing-generated-address tracking.
904 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
905 supports it and a size is available (ie. the sending peer gave us one).
907 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
908 years of being deprecated). The replacements are the ACLs
909 acl_smtp_mime and acl_not_smtp_mime.
911 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
912 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
913 or even an in-clear connection were permitted. Now, if the host lookup
914 was dnssec and dane was requested then the host is only used if the
915 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
916 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
917 if one fails this test.
918 This means that a poorly-configured remote DNS will make it incommunicado;
919 but it protects against a DNS-interception attack on it.
921 JH/04 Bug 1810: make continued-use of an open smtp transport connection
922 non-noisy when a race steals the message being considered.
924 JH/05 If main configuration option tls_certificate is unset, generate a
925 self-signed certificate for inbound TLS connections.
927 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
928 in rewrites and routers.
930 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
931 and logged a warning sing 4.83; now they are a configuration file error.
933 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
934 (lacking @domain). Apply the same qualification processing as RCPT.
936 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
938 JH/10 Support ${sha256:} applied to a string (as well as the previous
941 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
942 a cutthrough deliver is pending, as we always want to make a connection.
943 This also avoids re-routing the message when later placing the cutthrough
944 connection after a verify cache hit.
945 Do not update it with the verify result either.
947 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
948 when routing results in more than one destination address.
950 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
951 signing (which inhibits the cutthrough capability). Previously only
952 the presence of an option was tested; now an expansion evaluating as
953 empty is permissible (obviously it should depend only on data available
954 when the cutthrough connection is made).
956 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
957 the relevant preceding SMTP command did not note the pipelining mode.
959 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
960 Previously they were not counted.
962 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
963 as one having no matching records. Previously we deferred the message
964 that needed the lookup.
966 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
967 distinguished as "(=".
969 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
970 for missing MX records. Previously it only worked for missing A records.
972 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
974 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
975 after the data-go-ahead and data-ack. Patch from Jason Betts.
977 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
978 even for a "none" policy. Patch from Tony Meyer.
980 JH/22 Fix continued use of a connection for further deliveries. If a port was
981 specified by a router, it must also match for the delivery to be
984 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
985 When one of the recipients of a message was unsuitable for the connection
986 (has no matching addresses), we lost track of needing to mark it
987 deferred. As a result mail would be lost.
989 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
991 JH/25 Decoding ACL controls is now done using a binary search; the source code
992 takes up less space and should be simpler to maintain. Merge the ACL
993 condition decode tables also, with similar effect.
995 JH/26 Fix problem with one_time used on a redirect router which returned the
996 parent address unchanged. A retry would see the parent address marked as
997 delivered, so not attempt the (identical) child. As a result mail would
1000 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1001 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1002 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1005 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1006 Main config options check_{log,spool}_{inodes,space} are now
1007 100 inodes, 10MB unless set otherwise in the configuration.
1009 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1010 Previously it only applied to the main-section connection policy
1013 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1015 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1016 by me. Added RFC7919 DH primes as an alternative.
1018 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1019 are not in the system include path.
1021 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1022 GnuTLS, when a session startup failed (eg because the client disconnected)
1023 Exim did stdio operations after fclose. This was exposed by a recent
1024 change which nulled out the file handle after the fclose.
1026 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1027 signed directly by the cert-signing cert, rather than an intermediate
1028 OCSP-signing cert. This is the model used by LetsEncrypt.
1030 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1032 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1033 an incoming connection.
1035 HS/02 Bug 1802: Do not half-close the connection after sending a request
1038 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1039 fallback to "prime256v1".
1041 JH/34 SECURITY: Use proper copy of DATA command in error message.
1042 Could leak key material. Remotely exploitable. CVE-2016-9963.
1048 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1049 and 3.4.4 - once the server is enabled to respond to an OCSP request
1050 it does even when not requested, resulting in a stapling non-aware
1051 client dropping the TLS connection.
1053 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1054 support variable-length bit vectors. No functional change.
1056 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1057 The I= interface field on outgoing lines is now after the H= remote
1058 host field, same as incoming lines. There is a separate
1059 outgoing_interface log selector which allows you to disable the
1062 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1063 If not running log_selector +smtp_connection the mainlog would be held
1064 open indefinitely after a "too many connections" event, including to a
1065 deleted file after a log rotate. Leave the per net connection logging
1066 leaving it open for efficiency as that will be quickly detected by the
1067 check on the next write.
1069 HS/01 Bug 1671: Fix post transport crash.
1070 Processing the wait-<transport> messages could crash the delivery
1071 process if the message IDs didn't exist for some reason. When
1072 using 'split_spool_directory=yes' the construction of the spool
1073 file name failed already, exposing the same netto behaviour.
1075 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1076 mime_regex ACL conditions.
1078 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1079 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1080 to HELO, local diagnostic string.
1082 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1083 log line to debug. Even when configured with a tls authenticator many
1084 client connections are expected to not authenticate in this way, so
1085 an authenticate fail is not an error.
1087 HS/02 Add the Exim version string to the process info. This way exiwhat
1088 gives some more detail about the running daemon.
1090 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1091 matter for fast-change records such as DNSBLs.
1093 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1094 retry record, even if constant. There may be multiple transports with
1095 different interface settings and the retry behaviour needs to be kept
1098 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1100 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1102 JH/10 Bug 840: fix log_defer_output option of pipe transport
1104 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1105 hosts_randomize. This matters a lot when combined with mua_wrapper.
1107 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1108 ${quote_pgsql:<string>} operator.
1110 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1112 JH/14 Change header file name for freeradius-client. Relevant if compiling
1113 with Radius support; from the Gentoo tree and checked under Fedora.
1115 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1117 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1118 option transport_filter, meaning no filtering.
1120 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1122 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1123 defaults to "*" (all hosts). The variable is now available when not built
1124 with TLS, default unset, mainly to enable keeping the testsuite sane.
1125 If a server certificate is not supplied (via tls_certificate) an error is
1126 logged, and clients will find TLS connections fail on startup. Presumably
1127 they will retry in-clear.
1128 Packagers of Exim are strongly encouraged to create a server certificate
1129 at installation time.
1131 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1132 with the $config_file variable.
1134 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1135 in transport context, after the attempt, and per-recipient. The latter type
1136 is per host attempted. The event data is the error message, and the errno
1137 information encodes the lookup type (A vs. MX) used for the (first) host,
1138 and the trailing two digits of the smtp 4xx response.
1140 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1141 to write to mainlog (or rejectlog, paniclog) in the window between file
1142 creation and permissions/ownership being changed. Particularly affects
1143 installations where exicyclog is run as root, rather than exim user;
1144 result is that the running daemon panics and dies.
1146 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1148 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1149 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1150 "pri" and "weight". Note that the previous implicit priority given by the
1151 list order is no longer honoured.
1153 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1154 for DKIM processing.
1156 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1157 by defining SUPPORT_SOCKS.
1159 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1160 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1161 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1162 variables are renamed to proxy_{local,external}_{address,port}.
1164 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1165 for a build by defining SUPPORT_I18N
1167 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1168 of the query string, and make ${quote_redis:} do that quoting.
1170 JH/29 Move Events support from Experimental to mainline, enabled by default
1171 and removable for a build by defining DISABLE_EVENT.
1173 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1175 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1176 cached by the daemon.
1178 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1179 by defining LOOKUP_REDIS. The libhiredis library is required.
1181 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1182 keys are given for lookup.
1184 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1185 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1186 only supported when built with TLS support. The PolarSSL SHA routines
1187 are still used when the TLS library is too old for convenient support.
1189 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1190 openssl_options), for security. OpenSSL forces this from version 1.1.0
1191 server-side so match that on older versions.
1193 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1194 allocation for $value could be released as the expansion processing
1195 concluded, but leaving the global pointer active for it.
1197 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1198 and to use the domains and local_parts ACL conditions.
1200 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1201 incorrectly not doubled on cutthrough transmission, hence seen as a
1202 body-termination at the receiving system - resulting in truncated mails.
1203 Commonly the sender saw a TCP-level error, and retransmitted the message
1204 via the normal store-and-forward channel. This could result in duplicates
1205 received - but deduplicating mailstores were liable to retain only the
1206 initial truncated version.
1208 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1210 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1212 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1213 we're in there, support oversigning also; bug 1309.
1215 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1217 HS/04 Add support for keep_environment and add_environment options.
1219 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1220 either intentional arithmetic overflow during PRNG, or testing config-
1223 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1224 delivery resulted in actual delivery. Cancel cutthrough before DATA
1227 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1228 rejects a recipient: pass the reject to the originator.
1230 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1231 Many were false-positives and ignorable, but it's worth fixing the
1234 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1235 for the new environment-manipulation done at startup. Move the routines
1236 from being local to tls.c to being global via the os.c file.
1238 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1239 an extract embedded as result-arg for a map, the first arg for extract
1240 is unavailable so we cannot tell if this is a numbered or keyed
1241 extraction. Accept either.
1247 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1250 JH/02 The smtp transport option "multi_domain" is now expanded.
1252 JH/03 The smtp transport now requests PRDR by default, if the server offers
1255 JH/04 Certificate name checking on server certificates, when exim is a client,
1256 is now done by default. The transport option tls_verify_cert_hostnames
1257 can be used to disable this per-host. The build option
1258 EXPERIMENTAL_CERTNAMES is withdrawn.
1260 JH/05 The value of the tls_verify_certificates smtp transport and main options
1261 default to the word "system" to access the system default CA bundle.
1262 For GnuTLS, only version 3.0.20 or later.
1264 JH/06 Verification of the server certificate for a TLS connection is now tried
1265 (but not required) by default. The verification status is now logged by
1266 default, for both outbound TLS and client-certificate supplying inbound
1269 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1272 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1273 Status Notification (bounce) messages are now MIME format per RFC 3464.
1274 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1275 under the control of the dsn_advertise_hosts option, and routers may
1276 have a dsn_lasthop option.
1278 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1279 default, modifiable by a malware= option. The list separator for
1280 the options can now be changed in the usual way. Bug 68.
1282 JH/10 The smtp_receive_timeout main option is now expanded before use.
1284 JH/11 The incoming_interface log option now also enables logging of the
1285 local interface on delivery outgoing connections.
1287 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1288 if the interface and destination host and port all match.
1290 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1293 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1294 Patch from Andrew Lewis.
1296 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1297 now supports optional time-restrictions, weighting, and priority
1298 modifiers per server. Patch originally by <rommer@active.by>.
1300 JH/16 The spamd_address main option now supports a mixed list of local
1301 and remote servers. Remote servers can be IPv6 addresses, and
1302 specify a port-range.
1304 JH/17 Bug 68: The spamd_address main option now supports an optional
1305 timeout value per server.
1307 JH/18 Bug 1581: Router and transport options headers_add/remove can
1308 now have the list separator specified.
1310 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1313 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1316 JH/21 Support for the A6 type of dns record is withdrawn.
1318 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1319 rather than the verbs used.
1321 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1322 from 255 to 1024 chars.
1324 JH/24 Verification callouts now attempt to use TLS by default.
1326 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1327 are generic router options now. The defaults didn't change.
1329 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1330 Original patch from Alexander Shikoff, worked over by JH.
1332 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1333 files if ZCAT_COMMAND is not executable.
1335 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1337 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1339 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1340 Normally benign, it bites when the pair was led to by a CNAME;
1341 modern usage is to not canonicalize the domain to a CNAME target
1342 (and we were inconsistent anyway for A-only vs AAAA+A).
1344 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1346 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1347 when evaluating $sender_host_dnssec.
1349 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1350 $sender_helo_dnssec variable.
1352 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1354 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1356 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1358 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1359 documented as working, but never had. Support all but $spam_report.
1361 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1362 added for tls authenticator.
1364 HS/03 Add perl_taintmode main config option
1370 TL/01 When running the test suite, the README says that variables such as
1371 no_msglog_check are global and can be placed anywhere in a specific
1372 test's script, however it was observed that placement needed to be near
1373 the beginning for it to behave that way. Changed the runtest perl
1374 script to read through the entire script once to detect and set these
1375 variables, reset to the beginning of the script, and then run through
1376 the script parsing/test process like normal.
1378 TL/02 The BSD's have an arc4random API. One of the functions to induce
1379 adding randomness was arc4random_stir(), but it has been removed in
1380 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1381 function when detected.
1383 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1384 cause callback expansion.
1386 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1387 syntax errors in an expansion can be treated as a string instead of
1388 logging or causing an error, due to the internal use of bool_lax
1389 instead of bool when processing it.
1391 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1392 server certificates when making smtp deliveries.
1394 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1396 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1398 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1400 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1401 Merged patch from Sebastian Wiedenroth.
1403 JH/05 Fix results-pipe from transport process. Several recipients, combined
1404 with certificate use, exposed issues where response data items split
1405 over buffer boundaries were not parsed properly. This eventually
1406 resulted in duplicates being sent. This issue only became common enough
1407 to notice due to the introduction of connection certificate information,
1408 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1410 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1411 size buffer was used, resulting in syntax errors when an expansion
1414 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1415 version 3.3.6 or later.
1417 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1418 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1419 both become "event_action", the variables become $event_name, $event_data
1420 and $event_defer_errno. There is a new variable $verify_mode, usable in
1421 routers, transports and related events. The tls:cert event is now also
1422 raised for inbound connections, if the main configuration event_action
1425 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1426 early OCSP support, but no stapling (appears to be less than 1.0.0).
1428 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1429 server certificate names available under the smtp transport option
1430 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1433 JH/10 Time-related extraction expansions from certificates now use the main
1434 option "timezone" setting for output formatting, and are consistent
1435 between OpenSSL and GnuTLS compilations. Bug 1541.
1437 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1438 encoded parameter in the incoming message. Bug 1558.
1440 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1441 include certificate info, eximon was claiming there were spoolfile
1444 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1446 JH/14 Log delivery-related information more consistently, using the sequence
1447 "H=<name> [<ip>]" wherever possible.
1449 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1450 are problematic for Debian distribution, omit them from the release
1453 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1455 JH/16 Fix string representation of time values on 64bit time_t architectures.
1458 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1459 output list separator was used.
1464 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1465 checkers that were complaining about end of non-void function with no
1468 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1469 This was a regression introduced in 4.83 by another bugfix.
1471 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1473 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1474 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1480 TF/01 Correctly close the server side of TLS when forking for delivery.
1482 When a message was received over SMTP with TLS, Exim failed to clear up
1483 the incoming connection properly after forking off the child process to
1484 deliver the message. In some situations the subsequent outgoing
1485 delivery connection happened to have the same fd number as the incoming
1486 connection previously had. Exim would try to use TLS and fail, logging
1487 a "Bad file descriptor" error.
1489 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1490 utilities have not been installed.
1492 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1493 temporary space as the ACL may create new global variables.
1495 TL/01 LDAP support uses per connection or global context settings, depending
1496 upon the detected version of the libraries at build time.
1498 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1499 to extract and use the src ip:port in logging and expansions as if it
1500 were a direct connection from the outside internet. PPv2 support was
1501 updated based on HAProxy spec change in May 2014.
1503 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1505 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1506 Properly escape header and check for NULL return.
1508 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1511 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1513 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1514 characters in header names, implemented as a verify condition.
1515 Contributed by Michael Fischer v. Mollard.
1517 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1518 results permerror and temperror. Previous values are deprecated but
1519 still accepted. In a future release, err_perm and err_temp will be
1520 completely removed, which will be a backward incompatibility if the
1521 ACL tests for either of these two old results. Patch contributed by
1522 user bes-internal on the mailing list.
1524 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1526 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1527 selectors, in both main and reject logs.
1529 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1532 JH/07 Add malware type "sock" for talking to simple daemon.
1534 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1536 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1537 routers/transports under cutthrough routing.
1539 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1540 numbers. Touch up "bool" conditional to keep the same definition.
1542 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1544 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1546 JH/12 Expand items in router/transport headers_add or headers_remove lists
1547 individually rather than the list as a whole. Bug 1452.
1549 Required for reasonable handling of multiple headers_ options when
1550 they may be empty; requires that headers_remove items with embedded
1551 colons must have them doubled (or the list-separator changed).
1553 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1554 view the policy declared in the DMARC record. Currently, $dmarc_status
1555 is a combined value of both the record presence and the result of the
1558 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1560 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1561 dnslookup router and the smtp transport (applying to the forward
1564 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1565 of ldap servers used for a specific lookup. Patch provided by Heiko
1568 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1569 New variable $lookup_dnssec_authenticated for observability.
1571 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1572 Patch submitted by Lars Timman.
1574 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1576 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1577 Requires trusted mode and valid format message id, aborts otherwise.
1578 Patch contributed by Heiko Schlichting.
1580 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1581 certextract with support for various fields. Bug 1358.
1583 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1584 is requested by default, modifiable by smtp transport option
1587 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1588 operate on certificate variables to give certificate fingerprints
1589 Also new ${sha256:cert_variable}.
1591 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1593 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1594 Christian Aistleitner.
1596 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1598 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1599 file. Patch from Wolfgang Breyha.
1601 JH/25 Expand the coverage of the delivery $host and $host_address to
1602 client authenticators run in verify callout. Bug 1476.
1604 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1605 align with daemon_smtp_ports. Bug 72.
1607 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1608 support and error reporting did not work properly.
1610 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1611 and is readable. Patch from Andrew Colin Kissa.
1613 TL/14 Enhance documentation of ${run expansion and how it parses the
1614 commandline after expansion, particularly in the case when an
1615 unquoted variable expansion results in an empty value.
1617 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1619 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1620 by renaming away from it. Addresses GH issue 15, reported by
1623 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1625 TL/15 SECURITY: prevent double expansion in math comparison functions
1626 (can expand unsanitized data). Not remotely exploitable.
1633 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1635 PP/02 Make -n do something, by making it not do something.
1636 When combined with -bP, the name of an option is not output.
1638 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1641 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1642 $sender_host_name and config options to manage this, and basic check
1645 PP/05 DSCP support for outbound connections and control modifier for inbound.
1647 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1648 (Only plugin which currently uses this is kerberos4, which nobody should
1649 be using, but we should make it available and other future plugins might
1650 conceivably use it, even though it would break NAT; stuff *should* be
1651 using channel bindings instead).
1653 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1654 name; added for Sendmail compatibility; requires admin caller.
1655 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1656 just ignore it); requires trusted caller.
1657 Also parse but ignore: -Ac -Am -X<logfile>
1660 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1662 TL/02 Add +smtp_confirmation as a default logging option.
1664 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1665 Patch by Magnus Holmgren from 2007-02-20.
1667 TL/04 Bugzilla 1281 - Spec typo.
1668 Bugzilla 1283 - Spec typo.
1669 Bugzilla 1290 - Spec grammar fixes.
1671 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1673 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1675 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1676 mailing list by by Dmitry Isaikin.
1678 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1680 JH/02 Support "G" suffix to numbers in ${if comparisons.
1682 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1684 NM/01 Bugzilla 1197 - Spec typo
1685 Bugzilla 1196 - Spec examples corrections
1687 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1689 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1690 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1693 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1694 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1696 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1697 "acl {{name}{arg}...}", and optional args on acl condition
1700 JH/05 Permit multiple router/transport headers_add/remove lines.
1702 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1704 JH/07 Avoid using a waiting database for a single-message-only transport.
1705 Performance patch from Paul Fisher. Bugzilla 1262.
1707 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1710 JH/09 Add $headers_added variable, with content from use of ACL modifier
1711 add_header (but not yet added to the message). Bugzilla 199.
1713 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1714 Pulled from Bugzilla 817 by Wolfgang Breyha.
1716 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1718 (nb: this is the same fix as in Exim 4.80.1)
1720 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1723 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1726 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1728 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1729 advertise SMTP AUTH mechanism to us, instead of a generic
1730 protocol violation error. Also, make Exim more robust to bad
1731 data from the Dovecot auth socket.
1733 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1735 When a queue runner is handling a message, Exim first routes the
1736 recipient addresses, during which it prunes them based on the retry
1737 hints database. After that it attempts to deliver the message to
1738 any remaining recipients. It then updates the hints database using
1741 So if a recipient address works intermittently, it can get repeatedly
1742 deferred at routing time. The retry hints record remains fresh so the
1743 address never reaches the final cutoff time.
1745 This is a fairly common occurrence when a user is bumping up against
1746 their storage quota. Exim had some logic in its local delivery code
1747 to deal with this. However it did not apply to per-recipient defers
1748 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1750 This change adds a proper retry rule check during routing so that the
1751 final cutoff time is checked against the message's age. We only do
1752 this check if there is an address retry record and there is not a
1753 domain retry record; this implies that previous attempts to handle
1754 the address had the retry_use_local_parts option turned on. We use
1755 this as an approximation for the destination being like a local
1756 delivery, as in LMTP.
1758 I suspect this new check makes the old local delivery cutoff check
1759 redundant, but I have not verified this so I left the code in place.
1761 TF/02 Correct gecos expansion when From: is a prefix of the username.
1763 Test 0254 submits a message to Exim with the header
1767 When I ran the test suite under the user fanf2, Exim expanded
1768 the header to contain my full name, whereas it should have added
1769 a Resent-Sender: header. It erroneously treats any prefix of the
1770 username as equal to the username.
1772 This change corrects that bug.
1774 GF/01 DCC debug and logging tidyup
1775 Error conditions log to paniclog rather than rejectlog.
1776 Debug lines prefixed by "DCC: " to remove any ambiguity.
1778 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1780 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1781 Bug spotted by Jeremy Harris; was flawed since initial commit.
1782 Would have resulted in OCSP responses post-SNI triggering an Exim
1783 NULL dereference and crash.
1785 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1787 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1788 Bug detection, analysis and fix by Samuel Thibault.
1789 Bugzilla 1331, Debian bug #698092.
1791 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1793 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1794 Server implementation by Todd Lyons, client by JH.
1795 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1796 config variable "prdr_enable" controls whether the server
1797 advertises the facility. If the client requests PRDR a new
1798 acl_data_smtp_prdr ACL is called once for each recipient, after
1799 the body content is received and before the acl_smtp_data ACL.
1800 The client is controlled by both of: a hosts_try_prdr option
1801 on the smtp transport, and the server advertisement.
1802 Default client logging of deliveries and rejections involving
1803 PRDR are flagged with the string "PRDR".
1805 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1806 fclose(). Diagnosis by Todd Lyons.
1808 PP/17 Update configure.default to handle IPv6 localhost better.
1809 Patch by Alain Williams (plus minor tweaks).
1812 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1813 This is now consistent with GnuTLS, and is now documented: the
1814 previous undocumented portable approach to treating the option as
1815 unset was to force an expansion failure. That still works, and
1816 an empty string is now equivalent.
1818 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1819 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1820 not performing validation itself.
1822 PP/20 Added force_command boolean option to pipe transport.
1823 Patch from Nick Koston, of cPanel Inc.
1825 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1828 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1830 PP/21 Fix eximon continuous updating with timestamped log-files.
1831 Broken in a format-string cleanup in 4.80, missed when I repaired the
1832 other false fix of the same issue.
1833 Report and fix from Heiko Schlichting.
1836 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1837 Report from Prashanth Katuri.
1839 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1840 It's SecureTransport, so affects any MacOS clients which use the
1841 system-integrated TLS libraries, including email clients.
1843 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1844 using a MIME ACL for non-SMTP local injection.
1845 Report and assistance in diagnosis by Warren Baker.
1847 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1849 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1851 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1852 last id that failed so it may be referenced in subsequent ACL's.
1854 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1857 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1858 ldap library initialization, allowing self-signed CA's to be
1859 used. Also properly sets require_cert option later in code by
1860 using NULL (global ldap config) instead of ldap handle (per
1861 session). Bug diagnosis and testing by alxgomz.
1863 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1864 the src/util/ subdirectory.
1866 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1867 renamed to Transport Post Delivery Action by Jeremy Harris, as
1870 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1871 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1872 redis_servers = needs to be configured which will be used by the redis
1873 lookup. Patch from Warren Baker, of The Packet Hub.
1875 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1877 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1878 hostname or reverse DNS when processing a host list. Used suggestions
1879 from multiple comments on this bug.
1881 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1883 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1884 Missed a few lines, added it to make the runtest require no keyboard
1887 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1888 contains upper case chars. Make router use caseful_local_part.
1890 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1891 support when GnuTLS has been built with p11-kit.
1897 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1899 This, or similar/improved, will also be change PP/11 of 4.82.
1905 PP/01 Handle short writes when writing local log-files.
1906 In practice, only affects FreeBSD (8 onwards).
1907 Bugzilla 1053, with thanks to Dmitry Isaikin.
1909 NM/01 Bugzilla 949 - Documentation tweak
1911 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1914 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1916 PP/02 Implemented gsasl authenticator.
1918 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1920 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1921 `pkg-config foo` for cflags/libs.
1923 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1924 with rest of GSASL and with heimdal_gssapi.
1926 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1927 `pkg-config foo` for cflags/libs for the TLS implementation.
1929 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1930 properties get this fed in as external SSF. A number of robustness
1931 and debugging improvements to the cyrus_sasl authenticator.
1933 PP/08 cyrus_sasl server now expands the server_realm option.
1935 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1936 Patch by Jeremy Harris.
1938 PP/10 Added dbmjz lookup type.
1940 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1942 PP/12 MAIL args handles TAB as well as SP, for better interop with
1943 non-compliant senders.
1944 Analysis and variant patch by Todd Lyons.
1946 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1947 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1948 Patch from Dirk Mueller <dmueller@suse.com>
1950 PP/13 tls_peerdn now print-escaped for spool files.
1951 Observed some $tls_peerdn in wild which contained \n, which resulted
1952 in spool file corruption.
1954 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1955 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1956 or write after TLS renegotiation, which otherwise led to messages
1959 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1960 as a tracking header (ie: a signed header comes before the signature).
1961 Patch from Wolfgang Breyha.
1963 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1964 comma-sep list; embedded commas doubled.
1966 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1968 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1970 Report and patch from Dmitry Banschikov.
1972 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1973 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1974 failures. We appear to now support TLS1.1+ with Exim.
1976 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1977 lets Exim select keys and certificates based upon TLS SNI from client.
1978 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1979 before an outbound SMTP session. New log_selector, +tls_sni.
1981 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1982 NULL dereference. Report and patch from Alun Jones.
1984 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1985 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1986 Not seeing resolver debug output on NetBSD, but suspect this is a
1987 resolver implementation change.
1989 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1990 Left warnings. Added "eximon gdb" invocation mode.
1992 PP/21 Defaulting "accept_8bitmime" to true, not false.
1994 PP/22 Added -bw for inetd wait mode support.
1996 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1997 locate the relevant includes and libraries. Made this the default.
1999 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2000 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2002 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2003 This may cause build issues on older platforms.
2005 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2006 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2007 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2008 Added SNI support via GnuTLS too.
2009 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2011 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2013 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2014 Applied second patch from Janne, implementing suggestion to default
2015 multiple-strings-in-record handling to match SPF spec.
2017 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2019 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2020 read-only, out of scope).
2021 Patch from Wolfgang Breyha, report from Stuart Northfield.
2023 PP/29 Fix three issues highlighted by clang analyser static analysis.
2024 Only crash-plausible issue would require the Cambridge-specific
2025 iplookup router and a misconfiguration.
2026 Report from Marcin Mirosław.
2028 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2030 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2031 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2032 As part of this, removing so much warning spew let me fix some minor
2033 real issues in debug logging.
2035 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2036 assignment on my part. Fixed.
2038 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2039 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2040 Janne Snabb (who went above and beyond: thank you).
2042 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2043 string otherwise requires a connection and a bunch more work and it's
2044 relatively easy to get wrong. Should also expose TLS library linkage
2047 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2048 64-bit ${eval} (JH/03).
2050 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2051 GNU libc to support some of the 64-bit stuff, should not lead to
2052 conflicts. Defined before os.h is pulled in, so if a given platform
2053 needs to override this, it can.
2055 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2056 protection layer was required, which is not implemented.
2057 Bugzilla 1254, patch from Wolfgang Breyha.
2059 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2060 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2061 tls_dhparam take prime identifiers. Also unbreak combination of
2062 OpenSSL+DH_params+TLSSNI.
2064 PP/39 Disable SSLv2 by default in OpenSSL support.
2070 PP/01 Solaris build fix for Oracle's LDAP libraries.
2071 Bugzilla 1109, patch from Stephen Usher.
2073 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2075 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2078 TF/02 Fix a couple more cases where we did not log the error message
2079 when unlink() failed. See also change 4.74-TF/03.
2081 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2082 lock up or crash if it happened to be inside a call to libc when it
2083 got a SIGUSR1 from exiwhat.
2085 The SIGUSR1 handler appends the current process status to the process
2086 log which is later printed by exiwhat. It used to use the general
2087 purpose logging code to do this, but several functions it calls are
2088 not safe for signals.
2090 The new output code in the SIGUSR1 handler is specific to the process
2091 log, and simple enough that it's easy to inspect for signal safety.
2092 Removing some special cases also simplifies the general logging code.
2093 Removing the spurious timestamps from the process log simplifies
2096 TF/04 Improved ratelimit ACL condition.
2098 The /noupdate option has been deprecated in favour of /readonly which
2099 has clearer semantics. The /leaky, /strict, and /readonly update modes
2100 are mutually exclusive. The update mode is no longer included in the
2101 database key; it just determines when the database is updated. (This
2102 means that when you upgrade Exim will forget old rate measurements.)
2104 Exim now checks that the per_* options are used with an update mode that
2105 makes sense for the current ACL. For example, when Exim is processing a
2106 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2107 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2108 must specify per_mail/readonly. If you omit the update mode it defaults to
2109 /leaky where that makes sense (as before) or /readonly where required.
2111 The /noupdate option is now undocumented but still supported for
2112 backwards compatibility. It is equivalent to /readonly except that in
2113 ACLs where /readonly is required you may specify /leaky/noupdate or
2114 /strict/noupdate which are treated the same as /readonly.
2116 A useful new feature is the /count= option. This is a generalization
2117 of the per_byte option, so that you can measure the throughput of other
2118 aggregate values. For example, the per_byte option is now equivalent
2119 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2121 The per_rcpt option has been generalized using the /count= mechanism
2122 (though it's more complicated than the per_byte equivalence). When it is
2123 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2124 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2125 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2126 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2127 Note that using per_rcpt with a non-readonly update mode in more than
2128 one ACL will cause the recipients to be double-counted. (The per_mail
2129 and per_byte options don't have this problem.)
2131 The handling of very low rates has changed slightly. If the computed rate
2132 is less than the event's count (usually one) then this event is the first
2133 after a long gap. In this case the rate is set to the same as this event's
2134 count, so that the first message of a spam run is counted properly.
2136 The major new feature is a mechanism for counting the rate of unique
2137 events. The new per_addr option counts the number of different
2138 recipients that someone has sent messages to in the last time period. It
2139 behaves like per_rcpt if all the recipient addresses are different, but
2140 duplicate recipient addresses do not increase the measured rate. Like
2141 the /count= option this is a general mechanism, so the per_addr option
2142 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2143 example, measure the rate that a client uses different sender addresses
2144 with the options per_mail/unique=$sender_address. There are further
2145 details in the main documentation.
2147 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2149 TF/06 Removed a few PCRE remnants.
2151 TF/07 Automatically extract Exim's version number from tags in the git
2152 repository when doing development or release builds.
2154 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2155 Bugzilla 879. Patch from Paul Fisher.
2157 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2158 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2161 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2163 PP/05 Variable $av_failed, true if the AV scanner deferred.
2164 Bugzilla 1078. Patch from John Horne.
2166 PP/06 Stop make process more reliably on build failure.
2167 Bugzilla 1087. Patch from Heiko Schlittermann.
2169 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2170 Bugzilla 1089. Patch from Heiko Schlittermann.
2172 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2173 Bugzilla 1131. Patch from Holger Weiß.
2175 PP/09 Handle IPv6 addresses with SPF.
2176 Bugzilla 860. Patch from Wolfgang Breyha.
2178 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2180 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2183 PP/11 match_* no longer expand right-hand-side by default.
2184 New compile-time build option, EXPAND_LISTMATCH_RHS.
2185 New expansion conditions, "inlist", "inlisti".
2187 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2189 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2191 PP/14 fix log_write() format string regression from TF/03.
2192 Bugzilla 1152. Patch from Dmitry Isaikin.
2198 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2200 PP/02 Harmonised TLS library version reporting; only show if debugging.
2201 Layout now matches that introduced for other libraries in 4.74 PP/03.
2203 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2205 PP/04 New "dns_use_edns0" global option.
2207 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2210 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2211 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2213 TK/01 Updated PolarSSL code to 0.14.2.
2214 Bugzilla 1097. Patch from Andreas Metzler.
2216 PP/07 Catch divide-by-zero in ${eval:...}.
2217 Fixes bugzilla 1102.
2219 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2222 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2223 format-string attack -- SECURITY: remote arbitrary code execution.
2225 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2226 time unintentionally subject to list matching rules, letting the header
2227 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2228 arbitrary string expansion). This allowed for information disclosure.
2230 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2231 INT_MIN/-1 -- value coerced to INT_MAX.
2237 NM/01 Workaround for PCRE version dependency in version reporting
2240 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2241 This fixes portability to compilers other than gcc, notably
2242 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2244 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2245 makefiles for portability to HP-UX and POSIX correctness.
2247 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2248 Also via indented variable definition in the Makefile.
2249 (Debugging by Oliver Heesakkers).
2251 PP/02 Restore caching of spamd results with expanded spamd_address.
2252 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2254 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2255 Improves build reliability. Fix from: Frank Elsner
2257 NM/02 Fix wide character breakage in the rfc2047 coding
2258 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2260 NM/03 Allow underscore in dnslist lookups
2261 Fixes bug 1026. Patch from Graeme Fowler
2263 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2264 Code patches from Adam Ciarcinski of NetBSD.
2266 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2269 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2270 is logged, to avoid truncation. Patch from John Horne.
2272 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2273 Patch from Jakob Hirsch.
2275 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2276 SQL string expansion failure details.
2277 Patch from Andrey Oktyabrski.
2279 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2280 Patch from Simon Arlott.
2282 PP/09 New lookups functionality failed to compile on old gcc which rejects
2283 extern declarations in function scope.
2284 Patch from Oliver Fleischmann
2286 PP/10 Use sig_atomic_t for flags set from signal handlers.
2287 Check getgroups() return and improve debugging.
2288 Fixed developed for diagnosis in bug 927 (which turned out to be
2291 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2292 Patch from Mark Zealey.
2294 PP/12 Bugzilla 1056: Improved spamd server selection.
2295 Patch from Mark Zealey.
2297 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2298 Based on patch from Heiko Schlittermann.
2300 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2301 Patch from Uwe Doering, sign-off by Michael Haardt.
2303 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2304 variable declaration deep within a block. Bug and patch from
2307 PP/15 lookups-Makefile IRIX compatibility coercion.
2309 PP/16 Make DISABLE_DKIM build knob functional.
2311 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2312 Patch by Simon Arlott
2314 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2315 variable argument macros. Our copy now differs from upstream.
2321 TF/01 Failure to get a lock on a hints database can have serious
2322 consequences so log it to the panic log.
2324 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2325 controlled using the smtp_confirmation log selector.
2327 TF/03 Include the error message when we fail to unlink a spool file.
2329 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2330 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2331 for maintaining out-of-tree patches for some time.
2333 PP/01 Bugzilla 139: Documentation and portability issues.
2334 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2335 Handle per-OS dynamic-module compilation flags.
2337 PP/02 Let /dev/null have normal permissions.
2338 The 4.73 fixes were a little too stringent and complained about the
2339 permissions on /dev/null. Exempt it from some checks.
2340 Reported by Andreas M. Kirchwitz.
2342 PP/03 Report version information for many libraries, including
2343 Exim version information for dynamically loaded libraries. Created
2344 version.h, now support a version extension string for distributors
2345 who patch heavily. Dynamic module ABI change.
2347 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2348 privilege escalation vulnerability whereby the Exim run-time user
2349 can cause root to append content of the attacker's choosing to
2352 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2355 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2356 If dropping privileges for untrusted macros, we disabled normal logging
2357 on the basis that it would fail; for the Exim run-time user, this is not
2358 the case, and it resulted in successful deliveries going unlogged.
2359 Fixed. Reported by Andreas Metzler.
2365 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2366 only prepend for the Resent-* case. Fixes regression introduced in
2367 Exim 4.70 by NM/22 for Bugzilla 607.
2369 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2370 increasing numbers of administrators be bitten by this.
2372 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2374 PP/03 Bugzilla 994: added openssl_options main configuration option.
2376 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2378 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2380 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2382 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2383 without a peer certificate, leading to a segfault because of an
2384 assumption that peers always have certificates. Be a little more
2385 paranoid. Problem reported by Martin Tscholak.
2387 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2388 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2389 NB: ClamAV planning to remove STREAM in "middle of 2010".
2390 CL also introduces -bmalware, various -d+acl logging additions and
2391 more caution in buffer sizes.
2393 PP/09 Implemented reverse_ip expansion operator.
2395 PP/10 Bugzilla 937: provide a "debug" ACL control.
2397 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2399 PP/12 Bugzilla 973: Implement --version.
2401 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2403 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2405 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2407 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2408 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2409 ignore trailing whitespace.
2411 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2413 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2416 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2417 Notification from Dr Andrew Aitchison.
2419 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2420 ExtendedDetectionInfo response format.
2421 Notification from John Horne.
2423 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2426 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2427 XSL and documented dependency on system catalogs, with examples of how
2430 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2433 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2434 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2435 configuration file which is writeable by the Exim user or group.
2437 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2438 of configuration files to cover files specified with the -C option if
2439 they are going to be used with root privileges, not just the default
2442 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2443 option (effectively making it always true).
2445 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2446 files to be used while preserving root privileges.
2448 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2449 that rogue child processes cannot use them.
2451 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2452 run-time user, instead of root.
2454 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2455 Exim run-time user without dropping privileges.
2457 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2458 result string, instead of calling string_vformat() twice with the same
2461 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2462 for other users. Others should always drop root privileges if they use
2463 -C on the command line, even for a whitelisted configure file.
2465 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2467 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2473 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2474 $data_path, and $header_path variables; fixed documentation bugs and
2477 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2478 exipick to access non-standard spools, including the "frozen" queue
2481 NM/01 Bugzilla 965: Support mysql stored procedures.
2482 Patch from Alain Williams
2484 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2486 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2487 Patch from Andreas Metzler
2489 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2490 Patch from Kirill Miazine
2492 NM/05 Bugzilla 671: Added umask to procmail example.
2494 JJ/03 installed exipick 20100323.0, fixing doc bug
2496 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2497 directory. Notification and patch from Dan Rosenberg.
2499 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2501 TK/02 Improve log output when DKIM signing operation fails.
2503 MH/01 Treat the transport option dkim_domain as a colon separated
2504 list, not as a single string, and sign the message with each element,
2505 omitting multiple occurences of the same signer.
2507 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2508 Bugzilla 985, 986. Patch by Simon Arlott
2510 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2511 Patch by Simon Arlott
2513 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2514 MBX locking. Notification from Dan Rosenberg.
2520 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2522 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2524 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2526 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2528 NM/04 Bugzilla 915: Flag broken perl installation during build.
2534 TK/01 Added patch by Johannes Berg that expands the main option
2535 "spamd_address" if it starts with a dollar sign.
2537 TK/02 Write list of recipients to X-Envelope-Sender header when building
2538 the mbox-format spool file for content scanning (suggested by Jakob
2541 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2542 (http://www.dcc-servers.net/) support via dccifd. Activated by
2543 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2545 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2546 by Mark Daniel Reidel <mr@df.eu>.
2548 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2549 When building exim an external PCRE library is now needed -
2550 PCRE is a system library on the majority of modern systems.
2551 See entry on PCRE_LIBS in EDITME file.
2553 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2554 conversation. Added nologin parameter to request.
2555 Patch contributed by Kirill Miazine.
2557 TF/01 Do not log submission mode rewrites if they do not change the address.
2559 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2561 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2562 log files in place. Contributed by Roberto Lima.
2564 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2566 TF/03 Bugzilla 615: When checking the local_parts router precondition
2567 after a local_part_suffix or local_part_prefix option, Exim now
2568 does not use the address's named list lookup cache, since this
2569 contains cached lookups for the whole local part.
2571 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2572 Robert Millan. Documentation is in experimental-spec.txt.
2574 TF/04 Bugzilla 668: Fix parallel build (make -j).
2576 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2578 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2579 Patch provided by Jan Srzednicki.
2581 TF/05 Leading white space used to be stripped from $spam_report which
2582 wrecked the formatting. Now it is preserved.
2584 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2585 that they are available at delivery time.
2587 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2589 TF/08 TLS error reporting now respects the incoming_interface and
2590 incoming_port log selectors.
2592 TF/09 Produce a more useful error message if an SMTP transport's hosts
2593 setting expands to an empty string.
2595 NM/06 Bugzilla 744: EXPN did not work under TLS.
2596 Patch provided by Phil Pennock.
2598 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2599 Patch provided by Richard Godbee.
2601 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2602 acl_smtp_notquit, added index entry.
2604 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2605 Patch provided by Eugene Bujak.
2607 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2608 accept(). Patch provided by Maxim Dounin.
2610 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2611 Patch provided by Phil Pennock.
2613 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2615 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2616 Patch provided by Brad "anomie" Jorsch.
2618 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2619 Patch provided by Dean Brooks.
2621 TK/05 Add native DKIM support (does not depend on external libraries).
2623 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2624 Patch provided by Graeme Fowler.
2626 NM/16 Bugzilla 851: Documentation example syntax fix.
2628 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2630 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2633 NM/19 Bugzilla 745: TLS version reporting.
2634 Patch provided by Phil Pennock.
2636 NM/20 Bugzilla 167: bool: condition support.
2637 Patch provided by Phil Pennock.
2639 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2640 clients. Patch provided by Phil Pennock.
2642 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2643 Patch provided by Brad "anomie" Jorsch.
2645 NM/23 Bugzilla 687: Fix misparses in eximstats.
2646 Patch provided by Heiko Schlittermann.
2648 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2649 Patch provided by Heiko Schlittermann.
2651 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2652 plus update to original patch.
2654 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2656 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2657 Patch provided by David Brownlee.
2659 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2661 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2663 NM/30 Bugzilla 888: TLS documentation bugfixes.
2665 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2667 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2668 Unlike the original bugzilla I have changed all shell scripts in src tree.
2670 NM/33 Bugzilla 898: Transport filter timeout fix.
2671 Patch by Todd Rinaldo.
2673 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2674 Patch by Serge Demonchaux.
2676 NM/35 Bugzilla 39: Base64 decode bug fixes.
2677 Patch by Jakob Hirsch.
2679 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2681 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2683 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2685 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2691 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2692 ALT-N's libdkim that I have put here:
2693 http://duncanthrax.net/exim-experimental/
2695 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2696 were called 'true' and it seems that C99 defines that as a reserved
2697 keyword to be used with 'bool' variable types. That means you could
2698 not include C99-style headers which use bools without triggering
2699 build errors in sieve.c.
2701 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2702 as mailq or other aliases. Changed the --help handling significantly
2703 to do whats expected. exim_usage() emits usage/help information.
2705 SC/01 Added the -bylocaldomain option to eximstats.
2707 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2709 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2711 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2717 PH/01 Another patch from the Sieve maintainer.
2719 PH/02 When an IPv6 address is converted to a string for single-key lookup
2720 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2721 dots are used instead of colons so that keys in lsearch files need not
2722 contain colons. This was done some time before quoting was made available
2723 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2724 (notated using the quote facility) so as to distinguish them from IPv4
2725 keys. This meant that lookups for IP addresses in host lists did not work
2726 for iplsearch lookups.
2728 This has been fixed by arranging for IPv6 addresses to be expressed with
2729 colons if the lookup type is iplsearch. This is not incompatible, because
2730 previously such lookups could never work.
2732 The situation is now rather anomalous, since one *can* have colons in
2733 ordinary lsearch keys. However, making the change in all cases is
2734 incompatible and would probably break a number of configurations.
2736 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2739 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2740 conversion specification without a maximum field width, thereby enabling
2741 a rogue spamd server to cause a buffer overflow. While nobody in their
2742 right mind would setup Exim to query an untrusted spamd server, an
2743 attacker that gains access to a server running spamd could potentially
2744 exploit this vulnerability to run arbitrary code as the Exim user.
2746 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2747 $primary_hostname instead of what libspf2 thinks the hosts name is.
2749 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2750 a directory entry by the name of the lookup key. Previously, if a
2751 symlink pointed to a non-existing file or a file in a directory that
2752 Exim lacked permissions to read, a lookup for a key matching that
2753 symlink would fail. Now it is enough that a matching directory entry
2754 exists, symlink or not. (Bugzilla 503.)
2756 PH/03 The body_linecount and body_zerocount variables are now exported in the
2759 PH/04 Added the $dnslist_matched variable.
2761 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2762 This means they are set thereafter only if the connection becomes
2765 PH/06 Added the client_condition to authenticators so that some can be skipped
2766 by clients under certain conditions.
2768 PH/07 The error message for a badly-placed control=no_multiline_responses left
2769 "_responses" off the end of the name.
2771 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2773 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2774 (without spaces) instead of just copying the configuration text.
2776 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2778 PH/11 Added $max_received_linelength.
2780 PH/12 Added +ignore_defer and +include_defer to host lists.
2782 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2783 way in which PCRE > 7.0 is built.
2785 PH/14 Implemented queue_only_load_latch.
2787 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2788 MAIL command. The effect was to mangle the value on 64-bit systems.
2790 PH/16 Another patch from the Sieve maintainer.
2792 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2794 PH/18 If a system quota error occurred while trying to create the file for
2795 a maildir delivery, the message "Mailbox is full" was not appended to the
2796 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2797 only to a quota excession during the actual writing of the file.
2799 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2800 characters?) which causes problems in log lines. The DN values are now
2801 passed through string_printing() before being added to log lines.
2803 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2804 and InterBase are left for another time.)
2806 PH/21 Added message_body_newlines option.
2808 PH/22 Guard against possible overflow in moan_check_errorcopy().
2810 PH/23 POSIX allows open() to be a macro; guard against that.
2812 PH/24 If the recipient of an error message contained an @ in the local part
2813 (suitably quoted, of course), incorrect values were put in $domain and
2814 $local_part during the evaluation of errors_copy.
2820 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2821 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2824 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2825 issue a MAIL command.
2827 PH/02 In an ACL statement such as
2829 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2831 if a client was not listed at all, or was listed with a value other than
2832 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2833 the condition was not true (as it should be), so access was not denied.
2834 The bug was that the ! inversion was incorrectly passed on to the second
2835 item. This has been fixed.
2837 PH/03 Added additional dnslists conditions == and =& which are different from
2838 = and & when the dns lookup returns more than one IP address.
2840 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2841 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2843 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2844 FSYNC, which compiles an option called disable_fsync that allows for
2845 bypassing fsync(). The documentation is heavily laced with warnings.
2847 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2849 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2850 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2851 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2852 including adding "make clean"; (3) Added -fPIC when compiling the test
2853 dynamically loaded module, to get rid of a warning.
2855 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2856 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2857 = 0s. The bug is otherwise harmless.
2859 PH/07 There was a bug in the dovecot authenticator such that the value of
2860 $auth1 could be overwritten, and so not correctly preserved, after a
2861 successful authentication. This usually meant that the value preserved by
2862 the server_setid option was incorrect.
2864 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2866 PH/09 Installed PCRE release 7.0.
2868 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2869 run for batched SMTP input. It is now run at the start of every message
2870 in the batch. While fixing this I discovered that the process information
2871 (output by running exiwhat) was not always getting set for -bs and -bS
2872 input. This is fixed, and it now also says "batched" for BSMTP.
2874 PH/11 Added control=no_pipelining.
2876 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2877 patch, slightly modified), and move the expansion of helo_data till after
2878 the connection is made in the smtp transport (so it can use these
2881 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2883 PH/14 Added log_selector = +pid.
2885 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2887 PH/16 Add ${if forany and ${if forall.
2889 PH/17 Added dsn_from option to vary the From: line in DSNs.
2891 PH/18 Flush SMTP output before performing a callout, unless control =
2892 no_callout_flush is set.
2894 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2895 was true (the default) a successful delivery failed to delete the retry
2896 item, thus causing premature timeout of the address. The bug is now
2899 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2901 PH/21 Long custom messages for fakedefer and fakereject are now split up
2902 into multiline responses in the same way that messages for "deny" and
2903 other ACL rejections are.
2905 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2906 with slight modification.
2908 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2909 draft, changing the syntax and factoring some duplicate code".
2911 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2912 for deliveries of the second and subsequent messages over the same SMTP
2915 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2916 ${reduce, with only minor "tidies".
2918 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2920 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2921 expansion side effects.
2923 PH/27 When a message times out after an over-quota error from an Exim-imposed
2924 quota, the bounce message says "mailbox is full". This message was not
2925 being given when it was a system quota that was exceeded. It now should
2928 MH/03 Made $recipients available in local_scan(). local_scan() already has
2929 better access to the recipient list through recipients_list[], but
2930 $recipients can be useful in postmaster-provided expansion strings.
2932 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2933 in the case of a MAIL command with additional options following the
2934 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2935 were accidentally chopped off.
2937 PH/29 SMTP synchronization checks are implemented when a command is read -
2938 there is a check that no more input is waiting when there shouldn't be
2939 any. However, for some commands, a delay in an ACL can mean that it is
2940 some time before the response is written. In this time, more input might
2941 arrive, invalidly. So now there are extra checks after an ACL has run for
2942 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2943 pipelining has not been advertised.
2945 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2947 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2948 returned in the SMTP response when smtp_return_error_details was false.
2949 This has been fixed.
2951 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2952 instead of the C I/O that was originally supplied, because problems were
2953 reported on Solaris.
2955 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2956 Exim which did not show up earlier: it was assuming that a call to
2957 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2958 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2959 was a macro that became an assignment, so it seemed to work. This has
2960 changed to a proper function call with a void return, hence the compile
2961 error. Exim's code has been fixed.
2963 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2966 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2968 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2970 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2971 because it turns out to be too big or there is a timeout) there is no
2972 "Completed" line in the log. When some messages of this type were
2973 selected by exigrep, they were listed as "not completed". Others were
2974 picked up by some special patterns. I have improved the selection
2975 criteria to be more general.
2977 PH/38 The host_find_failed option in the manualroute router can now be set
2978 to "ignore", to completely ignore a host whose IP address cannot be
2979 found. If all hosts are ignored, the behaviour is controlled by the new
2980 host_all_ignored option.
2982 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2983 homing or because of multiple MX records with /mx) generated more than
2984 one IP address, and the following item turned out to be the local host,
2985 all the secondary addresses of the first item were incorrectly removed
2986 from the list, along with the local host and any following hosts (which
2987 is what is supposed to happen).
2989 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2990 whoever called Exim into the -H file. In the case of the daemon it was
2991 behaving confusingly. When first started, it used values for whoever
2992 started the daemon, but after a SIGHUP it used the Exim user (because it
2993 calls itself on a restart). I have changed the code so that it now always
2996 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2997 message are rejected with the same error (e.g. no authentication or bad
2998 sender address), and a DATA command is nevertheless sent (as can happen
2999 with PIPELINING or a stupid MUA), the error message that was given to the
3000 RCPT commands is included in the rejection of the DATA command. This is
3001 intended to be helpful for MUAs that show only the final error to their
3004 PH/42 Another patch from the Sieve maintainer.
3006 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3007 Eximstats - Fixed some broken HTML links and added missing column headers
3009 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3010 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3012 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3014 PH/43 Yet another patch from the Sieve maintainer.
3016 PH/44 I found a way to check for a TCP/IP connection going away before sending
3017 the response to the final '.' that terminates a message, but only in the
3018 case where the client has not sent further data following the '.'
3019 (unfortunately, this is allowed). However, in many cases there won't be
3020 any further data because there won't be any more messages to send. A call
3021 to select() can be used: if it shows that the input is "ready", there is
3022 either input waiting, or the socket has been closed. An attempt to read
3023 the next input character can distinguish the two cases. Previously, Exim
3024 would have sent an OK response which the client would never have see.
3025 This could lead to message repetition. This fix should cure that, at
3026 least in a lot of common cases.
3028 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3029 advertised in response to EHLO.
3035 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3036 fixed by 4.65/MH/01 (is this a record?) are fixed:
3038 (i) An empty string was always treated as zero by the numeric comparison
3039 operators. This behaviour has been restored.
3041 (ii) It is documented that the numeric comparison operators always treat
3042 their arguments as decimal numbers. This was broken in that numbers
3043 starting with 0 were being interpreted as octal.
3045 While fixing these problems I realized that there was another issue that
3046 hadn't been noticed. Values of message_size_limit (both the global option
3047 and the transport option) were treated as octal if they started with 0.
3048 The documentation was vague. These values are now always treated as
3049 decimal, and I will make that clear in the documentation.
3055 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3056 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3059 MH/01 Don't check that the operands of numeric comparison operators are
3060 integers when their expansion is in "skipping" mode (fixes bug
3061 introduced by 4.64-PH/07).
3063 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3064 child addresses, Exim now panics and dies. Previously, because the count
3065 is held in a short int, deliveries were likely to be lost. As such a
3066 large number of recipients for a single message is ridiculous
3067 (performance will be very, very poor), I have chosen to impose a limit
3068 rather than extend the field.
3074 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3075 leftover -K file (the existence of which was triggered by #402).
3076 While we were at it, introduced process PID as part of the -K
3077 filename. This should rule out race conditions when creating
3080 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3081 processing considerably. Previous code took too long for large mails,
3082 triggering a timeout which in turn triggers #401.
3084 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3085 in the DK code in transports.c. sendfile() is not really portable,
3086 hence the _LINUX specificness.
3088 TF/01 In the add_headers option to the mail command in an Exim filter,
3089 there was a bug that Exim would claim a syntax error in any
3090 header after the first one which had an odd number of characters
3093 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3094 callout verification, Exim cached a "reject" for the entire domain. This
3095 is correct for most verifications, but it is not correct for a recipient
3096 verification with use_sender or use_postmaster set, because in that case
3097 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3098 case of MAIL FROM:<> rejection from other early rejections (e.g.
3099 rejection of HELO). When verifying a recipient using a non-null MAIL
3100 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3101 Whatever the result of the callout, the value of the domain cache is
3102 left unchanged (for any other kind of callout, getting as far as trying
3103 RCPT means that the domain itself is ok).
3105 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3108 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3109 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3110 session, the daemon ignores ECONNECT errors and logs others; it now
3111 ignores EPIPE as well.
3113 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3114 (quoted-printable decoding).
3116 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3117 later the small subsequent patch to fix an introduced bug.
3119 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3121 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3123 PH/08 An error is now given if message_size_limit is specified negative.
3125 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3126 to be given (somewhat) arbitrary names.
3128 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3131 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3132 miscellaneous code fixes
3134 PH/10 Added the log_reject_target ACL modifier to specify where to log
3137 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3138 hostname. This is wrong, because it relates to the incoming message (and
3139 probably the interface on which it is arriving) and not to the outgoing
3140 callout (which could be using a different interface). This has been
3141 changed to use the value of the helo_data option from the smtp transport
3142 instead - this is what is used when a message is actually being sent. If
3143 there is no remote transport (possible with a router that sets up host
3144 addresses), $smtp_active_hostname is used.
3146 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3147 tweaks were necessary in order to get it to work (see also 21 below):
3148 (a) The code assumed that strncpy() returns a negative number on buffer
3149 overflow, which isn't the case. Replaced with Exim's string_format()
3151 (b) There were several signed/unsigned issues. I just did the minimum
3152 hacking in of casts. There is scope for a larger refactoring.
3153 (c) The code used strcasecmp() which is not a standard C function.
3154 Replaced with Exim's strcmpic() function.
3155 (d) The code set only $1; it now sets $auth1 as well.
3156 (e) A simple test gave the error "authentication client didn't specify
3157 service in request". It would seem that Dovecot has changed its
3158 interface. Fortunately there's a specification; I followed it and
3159 changed what the client sends and it appears to be working now.
3161 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3164 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3165 address A is aliased to B and C, where B exists and C does not. Without
3166 -v the output is "A verified" because verification stops after a
3167 successful redirection if more than one address is generated. However,
3168 with -v the child addresses are also verified. Exim was outputting "A
3169 failed to verify" and then showing the successful verification for C,
3170 with its parentage. It now outputs "B failed to verify", showing B's
3171 parentage before showing the successful verification of C.
3173 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3174 look up a TXT record in a specific list after matching in a combined
3177 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3178 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3179 they consult the DNS. I had assumed they would set it the way they
3180 wanted; and indeed my experiments on Linux seem to show that in some
3181 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3182 To be on the safe side, however, I have now made the interface to
3183 host_find_byname() similar to host_find_bydns(), with an argument
3184 containing the DNS resolver options. The host_find_byname() function now
3185 sets these options at its start, just as host_find_bydns() does. The smtp
3186 transport options dns_qualify_single and dns_search_parents are passed to
3187 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3188 of host_find_byname() use the default settings of RES_DEFNAMES
3189 (qualify_single) but not RES_DNSRCH (search_parents).
3191 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3192 spool_read_header() do less string testing, by means of a preliminary
3193 switch on the second character of optional "-foo" lines. (This is
3194 overdue, caused by the large number of possibilities that now exist.
3195 Originally there were few.) While I was there, I also converted the
3196 str(n)cmp tests so they don't re-test the leading "-" and the first
3197 character, in the hope this might squeeze out yet more improvement.
3199 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3200 flag allowing group syntax was set by the header_syntax check but not
3201 turned off, possible causing trouble later; (2) The flag was not being
3202 set at all for the header_verify test, causing "group"-style headers to
3203 be rejected. I have now set it in this case, and also caused header_
3204 verify to ignore an empty address taken from a group. While doing this, I
3205 came across some other cases where the code for allowing group syntax
3206 while scanning a header line wasn't quite right (mostly, not resetting
3207 the flag correctly in the right place). These bugs could have caused
3208 trouble for malformed header lines. I hope it is now all correct.
3210 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3211 with the "reply" argument non-NULL. The code, however (which originally
3212 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3213 but it didn't always do it. This confused somebody who was copying the
3214 code for some other use. I have removed all the tests.
3216 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3217 feature that was used to support insecure browsers during the U.S. crypto
3218 embargo. It requires special client support, and Exim is probably the
3219 only MTA that supported it -- and would never use it because real RSA is
3220 always available. This code has been removed, because it had the bad
3221 effect of slowing Exim down by computing (never used) parameters for the
3222 RSA_EXPORT functionality.
3224 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3225 authenticator to fail if there's a tab character in the incoming data
3226 (there should never be unless someone is messing about, as it's supposed
3227 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3228 if the connection is using TLS or if the remote IP is the same as the
3229 local IP, and the "valid-client-cert option" if a client certificate has
3232 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3233 authenticators. This can be used for authorization after authentication
3234 succeeds. (In the case of plaintext, it servers for both authentication
3237 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3238 if any retry times were supplied.
3240 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3241 connection, where there is no client IP address to check. In this
3242 situation, the verify now always succeeds.
3244 PH/25 Applied John Jetmore's -Mset patch.
3246 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3248 PH/27 In a string expansion for a processed (not raw) header when multiple
3249 headers of the same name were present, leading whitespace was being
3250 removed from all of them, but trailing whitespace was being removed only
3251 from the last one. Now trailing whitespace is removed from each header
3252 before concatenation. Completely empty headers in a concatenation (as
3253 before) are ignored.
3255 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3256 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3258 PH/29 [Removed. This was a change that I later backed out, and forgot to
3259 correct the ChangeLog entry (that I had efficiently created) before
3260 committing the later change.]
3262 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3263 address errors (4xx response to RCPT) over the same connection as other
3264 messages routed to the same hosts. Such deliveries are always "forced",
3265 so retry times are not inspected. This resulted in far too many retries
3266 for the affected addresses. The effect occurred only when there were more
3267 hosts than the hosts_max_try setting in the smtp transport when it had
3268 the 4xx errors. Those hosts that it had tried were not added to the list
3269 of hosts for which the message was waiting, so if all were tried, there
3270 was no problem. Two fixes have been applied:
3272 (i) If there are any address or message errors in an SMTP delivery, none
3273 of the hosts (tried or untried) are now added to the list of hosts
3274 for which the message is waiting, so the message should not be a
3275 candidate for sending over the same connection that was used for a
3276 successful delivery of some other message. This seems entirely
3277 reasonable: after all the message is NOT "waiting for some host".
3278 This is so "obvious" that I'm not sure why it wasn't done
3279 previously. Hope I haven't missed anything, but it can't do any
3280 harm, as the worst effect is to miss an optimization.
3282 (ii) If, despite (i), such a delivery is accidentally attempted, the
3283 routing retry time is respected, so at least it doesn't keep
3284 hammering the server.
3286 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3287 in ${readsocket because some servers need this prod.
3289 PH/32 Added some extra debug output when updating a wait-xxx database.
3291 PH/33 The hint "could be header name not terminated by colon", which has been
3292 given for certain expansion errors for a long time, was not being given
3293 for the ${if def:h_colon_omitted{... case.
3295 PH/34 The spec says: "With one important exception, whenever a domain list is
3296 being scanned, $domain contains the subject domain." There was at least
3297 one case where this was not true.
3299 PH/35 The error "getsockname() failed: connection reset by peer" was being
3300 written to the panic log as well as the main log, but it isn't really
3301 panic-worthy as it just means the connection died rather early on. I have
3302 removed the panic log writing for the ECONNRESET error when getsockname()
3305 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3306 runs only) independently of the message's sender address. This meant
3307 that, if the 4xx error was in fact related to the sender, a different
3308 message to the same recipient with a different sender could confuse
3309 things. In particular, this can happen when sending to a greylisting
3310 server, but other circumstances could also provoke similar problems.
3311 I have changed the default so that the retry time for these errors is now
3312 based a combination of the sender and recipient addresses. This change
3313 can be overridden by setting address_retry_include_sender=false in the
3316 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3317 remote server are returned as part of bounce messages. This was not
3318 happening for LMTP over a pipe (the lmtp transport), but now it is the
3319 same for both kinds of LMTP.
3321 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3322 in header lines that were in fact CNAMEs. This is no longer the case.
3324 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3325 and queue runs started by the daemon processed all messages. This has
3326 been fixed so that -R and -S can now usefully be given with -q<time>.
3328 PH/40 Import PCRE release 6.7 (fixes some bugs).
3330 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3332 PH/42 Give an error if -q is specified more than once.
3334 PH/43 Renamed the variables $interface_address and $interface_port as
3335 $received_ip_address and $received_port, to make it clear that these
3336 values apply to message reception, and not to the outgoing interface when
3337 a message is delivered. (The old names remain recognized, of course.)
3339 PH/44 There was no timeout on the connect() call when using a Unix domain
3340 socket in the ${readsocket expansion. There now is.
3342 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3343 be meaningful with "accept".
3345 SC/01 Eximstats V1.43
3346 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3348 SC/02 Eximstats V1.44
3349 Use a glob alias rather than an array ref in the generated
3350 parser. This improves both readability and performance.
3352 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3353 Collect SpamAssassin and rejection statistics.
3354 Don't display local sender or destination tables unless
3355 there is data to show.
3356 Added average volumes into the top table text output.
3358 SC/04 Eximstats V1.46
3359 Collect data on the number of addresses (recipients)
3360 as well as the number of messages.
3362 SC/05 Eximstats V1.47
3363 Added 'Message too big' to the list of mail rejection
3364 reasons (thanks to Marco Gaiarin).
3366 SC/06 Eximstats V1.48
3367 Mainlog lines which have GMT offsets and are too short to
3368 have a flag are now skipped.
3370 SC/07 Eximstats V1.49 (Alain Williams)
3371 Added the -emptyok flag.
3373 SC/08 Eximstats V1.50
3374 Fixes for obtaining the IP address from reject messages.
3376 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3377 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3378 whitespace changes from 4.64-PH/27
3380 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3383 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3384 are found, allow negative numbers in numeric criteria)
3386 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3388 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3389 to match changes made in 4.64-PH/43
3391 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3393 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3394 "domain" is actually an IP address, and force a failure. This locks out
3395 those revolvers/nameservers that support "A-for-A" lookups, in
3396 contravention of the specifications.
3398 PH/48 When a host name was looked up from an IP address, and the subsequent
3399 forward lookup of the name timed out, the host name was left in
3400 $sender_host_name, contrary to the specification.
3402 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3403 restricted to single-key lookups, Exim was not diagnosing an error if
3404 * or *@ was used with a query-style lookup.
3406 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3408 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3409 long ago, but noone interested enough thought of it. Let's just say that
3410 the "1.1" means that there are some new functions that weren't there at
3411 some point in the past.
3413 PH/51 Error processing for expansion failure of helo_data from an smtp
3414 transport during callout processing was broken.
3416 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3417 tested/used via the -bh/-bhc/-bs options.
3419 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3420 bug, fixed in subsequent PCRE releases).
3422 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3423 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3425 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3431 SC/01 Use a glob alias rather than an array ref in eximstats generated
3432 parser. This improves both readability and performance.
3434 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3435 Don't display local sender or destination tables in eximstats unless
3436 there is data to show.
3437 Added average volumes into the eximstats top table text output.
3439 SC/03 Collect data on the number of addresses (recipients) as well
3440 as the number of messages in eximstats.
3442 TF/01 Correct an error in the documentation for the redirect router. Exim
3443 does (usually) call initgroups() when daemonizing.
3445 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3446 with consistent privilege compared to when running as a daemon.
3448 TF/03 Note in the spec that $authenticated_id is not set for local
3449 submissions from trusted users.
3451 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3452 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3454 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3455 by adding some example configuration directives to the default
3456 configuration file. A little bit of work is required to uncomment the
3457 directives and define how usernames and passwords are checked, but
3458 there is now a framework to start from.
3460 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3461 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3462 without this. I don't know how relevant this is to other LDAP libraries.
3464 PH/02 Add the verb name to the "unknown ACL verb" error.
3466 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3468 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3470 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3471 directory not expanded when it should be if an expanded home directory
3472 was set for the address (which is overridden by the transport).
3474 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3477 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3478 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3479 because it is too late at that time, and has no effect.
3481 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3482 security issue with \' (bugzilla #107). I could not use the
3483 PQescapeStringConn() function, because it needs a PGconn value as one of
3486 PH/08 When testing addresses using -bt, indicate those final addresses that
3487 are duplicates that would not cause an additional delivery. At least one
3488 person was confused, thinking that -bt output corresponded to deliveries.
3489 (Suppressing duplicates isn't a good idea as you lose the information
3490 about possibly different redirections that led to the duplicates.)
3492 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3493 systems where poll() doesn't work, in particular OS X.
3495 PH/10 Added more information to debugging output for retry time not reached.
3497 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3498 operations in malware.c.
3500 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3503 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3504 syslog_duplication set false, log lines that would normally be written
3505 both the the main log and to the reject log were not written to syslog at
3508 PH/14 In the default configuration, change the use of "message" in ACL warn
3509 statements to "add_header".
3511 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3512 not followed by a command (e.g. "seen endif").
3514 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3515 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3518 PH/17 Added extra conditions to the default value of delay_warning_condition
3522 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3523 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3524 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3527 The Auto-Submitted: and various List- headers are standardised, whereas I
3528 don't think Precedence: ever was.
3530 PH/18 Refactored debugging code in route_finduser() to show more information,
3531 in particular, the error code if getpwnam() issues one.
3533 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3534 This is apparently needed in addition to the PH/07 change above to avoid
3535 any possible encoding problems.
3537 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3538 but not after initializing Perl.
3540 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3541 output them only if debugging. By default they are written stderr,
3542 apparently, which is not desirable.
3544 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3547 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3550 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3552 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3553 authenticated or an ident call has been made. Suppress the default
3554 values for $authenticated_id and $authenticated_sender (but permit -oMai
3555 and -oMas) when testing with -bh.
3557 PH/24 Re-jigged the order of the tests in the default configuration so that the
3558 tests for valid domains and recipients precede the DNS black list and CSA
3559 tests, on the grounds that those ones are more expensive.
3561 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3562 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3563 This bug exists in every version of Exim that I still have, right back to
3566 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3567 However, an attempt to turn on case-sensitivity in a regex key by
3568 including (?-i) didn't work because the subject string was already
3569 lowercased, and the effects were non-intuitive. It turns out that a
3570 one-line patch can be used to allow (?-i) to work as expected.
3576 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3577 other effects) broke the use of negated acl sub-conditions.
3579 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3582 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3583 "Deny" causes Exim to reject the incoming connection with a 554 error.
3584 Unfortunately, if there is a major crisis, such as a disk failure,
3585 tcp-wrappers gives "deny", whereas what one would like would be some
3586 kind of temporary error. A kludge has been added to help with this.
3587 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3588 554 error is used if errno is still zero or contains ENOENT (which occurs
3589 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3592 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3594 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3595 errors. Otherwise a message that provokes a temporary error (when other
3596 messages do not) can cause a whole host to time out.
3598 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3599 addresses were routed directly to files or pipes from a redirect router.
3600 File deliveries just didn't batch; pipe deliveries might have suffered
3603 PH/06 A failure to get a lock for a hints database would erroneously always say
3604 "Failed to get write lock", even when it was really a read lock.
3606 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3607 of 0600. This has been changed to use the value of the lockfile_mode
3608 option (which defaults to 0600).
3610 PH/08 Applied small patch from the Sieve maintainer.
3612 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3613 folder from quota calculations, a direct delivery into this folder messed
3614 up the contents of the maildirsize file. This was because the regex was
3615 used only to exclude .Trash (or whatever) when the size of the mailbox
3616 was calculated. There was no check that a delivery was happening into an
3617 excluded directory. This bug has been fixed by ignoring all quota
3618 processing for deliveries into excluded directories.
3620 PH/10 Added the maildirfolder_create_regex option to appendfile.
3626 PH/01 The code for finding all the local interface addresses on a FreeBSD
3627 system running IPv6 was broken. This may well have applied to all BSD
3628 systems, as well as to others that have similar system calls. The broken
3629 code found IPv4 interfaces correctly, but gave incorrect values for the
3630 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3631 that it would not match correctly against @[] and not recognize the IPv6
3634 PH/02 The ipliteral router was not recognizing addresses of the form user@
3635 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3637 PH/03 Added disable_ipv6.
3639 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3640 decoded form, because it is most often used to construct To: headers
3641 lines in autoreplies, and the decoded form may well be syntactically
3642 invalid. However, $reply_address has leading white space removed, and all
3643 newlines turned into spaces so that the autoreply transport does not
3646 PH/05 If group was specified without a user on a router, and no group or user
3647 was specified on a transport, the group from the router was ignored.
3649 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3650 for visible compile-time settings that can be used to change these
3651 numbers, for those that want even more. Backwards compatibility with old
3652 spool files has been maintained. However, going back to a previous Exim
3653 release will lost any variables that are in spool files.
3655 PH/07 Two small changes when running in the test harness: increase delay when
3656 passing a TCP/IP connection to a new process, in case the original
3657 process has to generate a bounce, and remove special handling of
3658 127.0.0.2 (sic), which is no longer necessary.
3660 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3661 be the same on different OS.
3663 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3666 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3667 whether --show-vars was specified or not
3669 JJ/02 exipick: Added support for new ACL variable spool format introduced
3672 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3673 syntactically invalid From: or Reply-to: line, and a filter used this to
3674 generate an autoreply, and therefore failed to obtain an address for the
3675 autoreply, Exim could try to deliver to a non-existent relative file
3676 name, causing unrelated and misleading errors. What now happens is that
3677 it logs this as a hard delivery error, but does not attempt to create a
3680 PH/11 The exinext utility has a -C option for testing purposes, but although
3681 the given file was scanned by exinext itself; it wasn't being passed on
3682 when Exim was called.
3684 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3685 an end-of-file indication when reading a command response.
3687 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3688 compiled. In many other places in Exim, IPv6 addresses are always
3689 recognized, so I have changed this. It also means that IPv4 domain
3690 literals of the form [IPV4:n.n.n.n] are now always recognized.
3692 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3693 used if the router is not running as root, for example, when verifying at
3694 ACL time, or when using -bh. The debugging output from this situation was
3695 non-existent - all you got was a failure to exec. I have made two
3698 (a) Failures to set uid/gid, the current directory, or a process leader
3699 in a subprocess such as that created by queryprogram now generate
3700 suitable debugging output when -d is set.
3702 (b) The queryprogram router detects when it is not running as root,
3703 outputs suitable debugging information if -d is set, and then runs
3704 the subprocess without attempting to change uid/gid.
3706 PH/15 Minor change to Makefile for building test_host (undocumented testing
3709 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3710 additional section of a DNS packet that returns MX or SRV records.
3711 Instead, it always explicitly searches for A/AAAA records. This avoids
3712 major problems that occur when a DNS server includes only records of one
3713 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3714 fixed another bug: if SRV records were looked up and the corresponding
3715 address records were *not* found in the additional section, the port
3716 values from the SRV records were lost.
3718 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3719 using the correct key (the original address) when searching the retry
3720 rules in order to find which one to use for generating the retry hint.
3722 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3723 adding the default one. Similarly, if it contains a Reply-To: header, the
3724 errors_reply_to option, if set, is not used.
3726 PH/19 When calculating a retry time, Exim used to measure the "time since
3727 failure" by looking at the "first failed" field in the retry record. Now
3728 it does not use this if it is later than than the arrival time of the
3729 message. Instead it uses the arrival time. This makes for better
3730 behaviour in cases where some deliveries succeed, thus re-setting the
3731 "first failed" field. An example is a quota failure for a huge message
3732 when small messages continue to be delivered. Without this change, the
3733 "time since failure" will always be short, possible causing more frequent
3734 delivery attempts for the huge message than are intended.
3735 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3737 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3738 $1, $2, $3) because the numerical variables can be reset during some
3739 expansion items (e.g. "match"), thereby losing the authentication data.
3741 PH/21 Make -bV show the size of off_t variables so that the test suite can
3742 decide whether to run tests for quotas > 2G.
3744 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3745 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3746 filecount value is greater than 2G or if a quota value is greater than 2G
3747 on a system where the size of off_t is not greater than 4, a panic error
3750 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3751 never match. The debug and -bh output now contains an explicit error
3752 message indicating a malformed IPv4 address or mask.
3754 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3755 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3756 PH/23 above applies.
3758 PH/25 Do not write to syslog when running in the test harness. The only
3759 occasion when this arises is a failure to open the main or panic logs
3760 (for which there is an explicit test).
3762 PH/26 Added the /no_tell option to "control=freeze".
3764 PH/27 If a host name lookup failed very early in a connection, for example, if
3765 the IP address matched host_lookup and the reverse lookup yielded a name
3766 that did not have a forward lookup, an error message of the form "no IP
3767 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3768 could be logged. Now it outputs the IP address instead of "NULL".
3770 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3771 allows the sender and the authenticated sender to be set when
3772 submitting a message from within Exim. Since child_open_exim() is
3773 documented for local_scan(), the new function should be too.
3775 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3776 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3777 results in an empty string is now treated as unset.
3779 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3781 PH/31 Added sender_verify_fail logging option.
3783 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3784 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3785 tidied the source and removed it altogether.
3787 PH/33 When a queue run was abandoned because the load average was too high, a
3788 log line was always written; now it is written only if the queue_run log
3789 selector is set. In addition, the log line for abandonment now contains
3790 information about the queue run such as the pid. This is always present
3791 in "start" and "stop" lines but was omitted from the "abandon" line.
3793 PH/34 Omit spaces between a header name and the colon in the error message that
3794 is given when verify = headers_syntax fails (if there are lots of them,
3795 the message gets confusing).
3797 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3798 names, as there are now some PTR records that contain slashes. This check
3799 is only to protect against broken name servers that fall over on strange
3800 characters, so the fact that it applies to all lookups doesn't matter.
3802 PH/36 Now that the new test suite is complete, we can remove some of the
3803 special code in Exim that was needed for the old test suite. For example,
3804 sorting DNS records because real resolvers return them in an arbitrary
3805 order. The new test suite's fake resolver always returns records in the
3808 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3809 bounces) except when queue_only is set, to avoid logging races between
3810 the different processes.
3812 PH/38 Panic-die if .include specifies a non-absolute path.
3814 PH/39 A tweak to the "H" retry rule from its user.
3816 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3817 a label. They prevented compilation on older perls.
3819 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3820 a warning to be raised on newish perls.
3822 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3823 on queue. Changes to match documented behaviour of showing count of
3824 messages matching specified criteria.
3826 PH/40 Changed the default ident timeout from 30s to 5s.
3828 PH/41 Added support for the use of login_cap features, on those BSD systems
3829 that have them, for controlling the resources used by pipe deliveries.
3831 PH/42 The content-scanning code uses fopen() to create files in which to put
3832 message data. Previously it was not paying any attention to the mode of
3833 the files. Exim runs with umask(0) because the rest of the code creates
3834 files with open(), and sets the required mode explicitly. Thus, these
3835 files were ending up world-writeable. This was not a big issue, because,
3836 being within the spool directory, they were not world-accessible. I have
3837 created a function called modefopen, which takes an additional mode
3838 argument. It sets umask(777), creates the file, chmods it to the required
3839 mode, then resets the umask. All the relevant calls to fopen() in the
3840 content scanning code have been changed to use this function.
3842 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3843 to 24 hours. This avoids potential overflow problems when processing G
3844 and H retry rules. I suspect nobody ever tinkers with this value.
3846 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3848 PH/45 When the plaintext authenticator is running as a client, the server's
3849 challenges are checked to ensure they are valid base64 strings. By
3850 default, the authentication attempt is cancelled if an invalid string is
3851 received. Setting client_ignore_invalid_base64 true ignores these errors.
3852 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3853 they are received. Thus, the responses can be made to depend on the
3854 challenges. If an invalid string is ignored, an empty string is placed in
3857 PH/46 Messages that are created by the autoreply transport now contains a
3858 References: header, in accordance with RFCs 2822 and 3834.
3860 PH/47 Added authenticated_sender_force to the smtp transport.
3862 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3864 PH/49 Installed latest patch from the Sieve maintainer.
3866 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3867 was also set, the appendfile transport was unnecessarily scanning a
3868 directory of message files (e.g. for maildir delivery) to find the count
3869 of files (along with the size), even though it did not need this
3870 information. It now does the scan only if it needs to find either the
3871 size of the count of files.
3873 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3875 PH/52 Two bugs concerned with error handling when the smtp transport is
3878 (i) Exim was not creating retry information for temporary errors given
3879 for individual recipients after the DATA command when the smtp transport
3880 was used in LMTP mode. This meant that they could be retried too
3881 frequently, and not timed out correctly.
3883 (ii) Exim was setting the flag that allows error details to be returned
3884 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3885 recipients that were returned after the DATA command.
3887 PH/53 This is related to PH/52, but is more general: for any failing address,
3888 when detailed error information was permitted to be returned to the
3889 sender, but the error was temporary, then after the final timeout, only
3890 "retry timeout exceeded" was returned. Now it returns the full error as
3891 well as "retry timeout exceeded".
3893 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3894 do this, and (what is worse) MTAs that accept it.
3896 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3897 will now be deprecated.
3899 PH/56 New os.c-cygwin from the Cygwin maintainer.
3901 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3902 formats (previously only available in exim formats via -bpr, -bpru,
3903 and -bpra. Now also available in native and exiqgrep formats)
3905 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3906 with very large, slow to parse queues
3908 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3910 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3912 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3913 responses to authentication challenges, though it was showing the
3914 challenges; (ii) I've removed the CR characters from the debug output for
3917 PH/58 Allow for the insertion of a newline as well as a space when a string
3918 is turned into more than one encoded-word during RFC 2047 encoding. The
3919 Sieve code now uses this.
3921 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3922 data_4xx, lost_connection, tls_required.
3924 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3925 message was being sent as an SMTP response.
3927 PH/61 Add -l and -k options to exicyclog.
3929 PH/62 When verifying, if an address was redirected to one new address, so that
3930 verification continued, and the new address failed or deferred after
3931 having set something in $address_data, the value of $address_data was not
3932 passed back to the ACL. This was different to the case when no
3933 redirection occurred. The value is now passed back in both cases.
3935 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3936 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3937 use login_cap.h, so on its own it isn't the distinguishing feature. The
3938 new name refers directly to the setclassresources() function.
3940 PH/65 Added configuration files for NetBSD3.
3942 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3944 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3945 is preferred over IPv4.
3947 PH/68 The bounce_return_message and bounce_return_body options were not being
3948 honoured for bounces generated during the reception of non-SMTP messages.
3949 In particular, this applied to messages rejected by the ACL. This bug has
3950 been fixed. However, if bounce_return_message is true and bounce_return_
3951 body is false, the headers that are returned for a non-SMTP message
3952 include only those that have been read before the error was detected.
3953 (In the case of an ACL rejection, they have all been read.)
3955 PH/69 The HTML version of the specification is now built in a directory called
3956 spec_html instead of spec.html, because the latter looks like a path with
3957 a MIME-type, and this confuses some software.
3959 PH/70 Catch two compiler warnings in sieve.c.
3961 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3962 function verify_get_ident() calls ip_connect() to connect a socket, but
3963 if the "connect()" function timed out, ip_connect() used to close the
3964 socket. However, verify_get_ident() also closes the socket later, and in
3965 between Exim writes to the log, which may get opened at this point. When
3966 the socket was closed in ip_connect(), the log could get the same file
3967 descriptor number as the socket. This naturally causes chaos. The fix is
3968 not to close the socket in ip_connect(); the socket should be closed by
3969 the function that creates it. There was only one place in the code where
3970 this was missing, in the iplookup router, which I don't think anybody now
3971 uses, but I've fixed it anyway.
3973 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3974 well as to direct DNS lookups. Otherwise the handling of names in host
3975 lists is inconsistent and therefore confusing.
3981 PH/01 Two changes to the default runtime configuration:
3983 (1) Move the checks for relay_from_hosts and authenticated clients from
3984 after to before the (commented out) DNS black list checks.
3986 (2) Add control=submission to the relay_from_hosts and authenticated
3987 clients checks, on the grounds that messages accepted by these
3988 statements are most likely to be submissions.
3990 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3992 (1) Generate an error if the third argument for the ${prvs expansion is
3995 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3998 (3) Reset the variables that are obtained from the first argument of
3999 ${prvscheck and used in the second argument before leaving the code,
4000 because their memory is reclaimed, so using them afterwards may do
4003 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4004 one (it's much easier than Tom thought :-).
4006 (5) Because of (4), we can now allow for the use of $prvscheck_result
4007 inside the third argument.
4009 PH/03 For some reason, the default setting of PATH when running a command from
4010 a pipe transport was just "/usr/bin". I have changed it to
4013 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4014 anything to be listed in the output from -bV.
4016 PH/05 When a filter generated an autoreply, the entire To: header line was
4017 quoted in the delivery log line, like this:
4019 => >A.N.Other <ano@some.domain> <original@ddress> ...
4021 This has been changed so that it extracts the operative address. There
4022 may be more than one such address. If so, they are comma-separated, like
4025 => >ano@some.domain,ona@other.domain <original@ddress> ...
4027 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4028 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4029 not being looked up in the rDNS to get a host name, Exim was showing the
4030 IP address twice in Received: lines, even though the IP addresses were
4031 identical. For example:
4033 Received: from [1.2.3.4] (helo=[1.2.3.4])
4035 However, if the real host name was known, it was omitting the HELO data
4036 if it matched the actual IP address. This has been tidied up so that it
4037 doesn't show the same IP address twice.
4039 PH/07 When both +timestamp and +memory debugging was on, the value given by
4040 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4041 called by the debug printing, thereby overwriting the timestamp buffer.
4042 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4044 PH/08 When the original message was included in an autoreply transport, it
4045 always said "this is a copy of the message, including all the headers",
4046 even if body_only or headers_only was set. It now gives an appropriate
4049 PH/09 Applied a patch from the Sieve maintainer which:
4051 o fixes some comments
4052 o adds the (disabled) notify extension core
4053 o adds some debug output for the result of if/elsif tests
4054 o points to the current vacation draft in the documentation
4055 and documents the missing references header update
4059 o fixes a bug in processing the envelope test (when testing
4060 multiple envelope elements, the last element determined the
4063 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4064 Electronic Mail") by including:
4066 Auto-submitted: auto-generated
4068 in the messages that it generates (bounce messages and others, such as
4069 warnings). In the case of bounce messages for non-SMTP messages, there was
4070 also a typo: it was using "Auto_submitted" (underscore instead of
4071 hyphen). Since every message generated by Exim is necessarily in response
4072 to another message, thes have all been changed to:
4074 Auto-Submitted: auto-replied
4076 in accordance with these statements in the RFC:
4078 The auto-replied keyword:
4080 - SHOULD be used on messages sent in direct response to another
4081 message by an automatic process,
4083 - MUST NOT be used on manually-generated messages,
4085 - MAY be used on Delivery Status Notifications (DSNs) and Message
4086 Disposition Notifications (MDNs),
4088 - MUST NOT be used on messages generated by automatic or periodic
4089 processes, except for messages which are automatic responses to
4092 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4093 to the default Received: header definition.
4095 PH/12 Added log selector acl_warn_skipped (default on).
4097 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4098 variables because (a) they are in the wrong storage pool and (b) even if
4099 they were copied, it wouldn't work properly because of the caching.
4101 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4102 checking when decoding. Apparently there are clients that generate
4103 overlong encoded strings. Why am I not surprised?
4105 PH/15 If the first argument of "${if match_address" was not empty, but did not
4106 contain an "@" character, Exim crashed. Now it writes a panic log message
4107 and treats the condition as false.
4109 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4111 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4112 extension "envelope-auth". The code is finished and in agreement with
4113 other implementations, but there is no documentation so far and in fact,
4114 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4115 not changing the active code.
4117 Print executed "if" and "elsif" statements when debugging is used. This
4118 helps a great deal to understand what a filter does.
4120 Document more things not specified clearly in RFC3028. I had all this
4121 sorted out, when out of a sudden new issues came to my mind. Oops."
4123 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4126 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4127 canonical form (as documented). However, after a host name lookup from
4128 the IP address, check_host() was doing a simple string comparison with
4129 addresses acquired from the DNS when checking that the found name did
4130 have the original IP as one of its addresses. Since any found IPv6
4131 addresses are likely to be in abbreviated form, the comparison could
4132 fail. Luckily, there already exists a function for doing the comparison
4133 by converting both addresses to binary, so now that is used instead of
4134 the text comparison.
4136 PH/20 There was another similar case to PH/19, when a complete host name was
4137 given in a host list; looking up its IP address could give an abbreviated
4138 form, whereas the current host's name might or might not be abbreviated.
4139 The same fix has been applied.
4145 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4146 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4149 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4150 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4152 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4154 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4155 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4156 Also, for "Auto-Submitted", treat anything other than "no" as
4157 non-personal, in accordance with RFC 3834. (Previously it treated
4158 anything starting "auto-" as non-personal.)
4160 TF/01 The control=submission/name=... option had a problem with syntax
4161 errors if the name included a slash character. The /name= option
4162 now slurps the rest of the string, so it can include any characters
4163 but it must come last in the list of options (after /sender_retain
4166 PH/05 Some modifications to the interface to the fake nameserver for the new
4174 TK/01 Added the "success_on_redirect" address verification option. See
4175 NewStuff for rationale and an example.
4177 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4179 PH/02 Patch to exigrep to allow it to work on syslog lines.
4181 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4182 fread() to skip over the body file's header line, because in Cygwin the
4183 header line is locked and is inaccessible.
4185 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4186 co-exist for some time) to make it clear that it is the Exim ID that is
4187 referenced, not the Message-ID: header line.
4189 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4190 string_format() function, because snprintf() does not exist on all
4193 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4194 ${stat: expansion item.
4196 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4197 protocol synchronization error", to keep the pedants happy.
4199 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4200 well as for IRIX systems, when gcc is being used. See the host.c source
4203 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4205 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4208 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4209 transport and to the smtp transport in LMTP mode.
4211 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4213 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4214 run a filter in a subprocess. This could lead to confusion in subsequent
4215 lookups in the parent process. There should also be a search_tidyup() at
4216 the end of the subprocess.
4218 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4219 only if the host matched helo_try_verify_hosts, which caused the
4220 verification to occur when the EHLO/HELO command was issued. The ACL just
4221 tested the remembered result. Now, if a previous verification attempt has
4222 not happened, "verify = helo" does it there and then.
4224 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4226 TK/03 Fix log output including CR from clamd.
4228 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4229 exist provoked a memory error which could cause a segfault.
4231 PH/15 Installed PCRE 6.2
4233 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4235 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4236 of the problem. Specifically, suggested +O2 rather than +O1 for the
4239 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4241 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4242 record, the port was not correctly logged when the outgoing_port log
4243 selector was set (it logged the transort's default port).
4245 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4246 fallback_hosts, and "hosts" in the smtp transport.
4248 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4249 host errors such as "Connection refused".
4251 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4252 authentication with radiusclient 0.4.9:
4254 - Error returned from rc_read_config was caught wrongly
4255 - Username/password not passed on to radius server due to wrong length.
4257 The presumption is that some radiusclient API changes for 4.51/PH/17
4258 were not taken care of correctly. The code is still untested by me (my
4259 Linux distribution still has 0.3.2 of radiusclient), but it was
4260 contributed by a Radius user.
4262 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4263 expanding the "port" option of the smtp transport.
4265 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4266 while reading a MIME header. Thanks to Tom Hughes for a patch.
4268 PH/24 Include config.h inside local_scan.h so that configuration settings are
4271 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4272 that in an ACL for RCPT (for example), you can examine exactly what was
4275 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4276 commands, but it was not correctly comparing the address with the actual
4277 client host address. Thus, it would show the EHLO address in Received:
4278 header lines when this was not necessary.
4280 PH/27 Added the % operator to ${eval:}.
4282 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4283 it should be ignoring failures (because with -C, for example, it has lost
4284 privilege). It wasn't ignoring creation failures other than "already
4287 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4290 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4291 because an input line was too long, either on its own, or by virtue of
4292 too many continuations, the temporary file was not being removed, and the
4293 return code was incorrect.
4295 PH/31 Missing "BOOL" in function definition in filtertest.c.
4297 PH/32 Applied Sieve patches from the maintainer.
4299 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4301 PH/33 Added "verify = not_blind".
4303 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4304 Local/Makefile (with some defaults set). These are used in built scripts
4305 such as exicyclog, but they have never been used in the exim_install
4306 script (though there are many overriding facilities there). I have
4307 arranged that the exim_install script now takes note of these two
4310 PH/35 Installed configuration files for Dragonfly.
4312 PH/36 When a locally submitted message by a trusted user did not contain a
4313 From: header, and the sender address was obtained from -f or from an SMTP
4314 MAIL command, and the trusted user did not use -F to supply a sender
4315 name, $originator_name was incorrectly used when constructing a From:
4316 header. Furthermore, $originator_name was used for submission mode
4317 messages from external hosts without From: headers in a similar way,
4318 which is clearly wrong.
4320 PH/37 Added control=suppress_local_fixups.
4322 PH/38 When log_selector = +received_sender was set, and the addition of the
4323 sender made the log line's construction buffer exactly full, or one byte
4324 less than full, an overflow happened when the terminating "\n" was
4327 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4328 when the result of a list match is failure because a DNS lookup failed.
4330 PH/40 RM_COMMAND is now used in the building process.
4332 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4333 the "build-* directories that it finds.
4335 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4336 address in a domain literal was a prefix of an interface address.
4338 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4339 when verifying a sender address, unless rewrite_headers is false.
4341 PH/44 Wrote a long comment about why errors_to addresses are verified as
4342 recipients, not senders.
4344 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4345 the ratelimit ACL was added.
4347 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4349 PH/46 Added extra information about PostgreSQL errors to the error string.
4351 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4352 avoiding the need to install special zones in a real server. This is
4353 backwards compatible; if it can't find the fake resolver, it drops back.
4354 Thus, both old and new test suites can be run.
4356 TF/02 Added util/ratelimit.pl
4358 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4359 clock is set back in time.
4361 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4362 Candler <B.Candler@pobox.com>.
4364 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4365 OK for addresses that are the result of redirections.
4367 PH/48 A number of further additions for the benefit of the new test suite,
4368 including a fake gethostbyname() that interfaces to the fake DNS resolver
4371 TF/06 The fix for widen_domains has also been applied to qualify_single and
4372 search_parents which are the other dnslookup options that can cause
4375 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4378 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4380 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4381 service" if it encountered an MX record with an empty target hostname.
4382 The message is now "an MX or SRV record indicated no SMTP service".
4384 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4385 if the DNS of the sending site is misconfigured. This is quite a
4386 common situation. This change restores the behaviour of treating a
4387 helo verification defer as a failure.
4389 PH/51 If self=fail was set on a router, the bounce message did not include the
4390 actual error message.
4396 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4398 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4399 command itself ended in error, the underlying message about the transport
4400 filter timeout was being overwritten with the pipe command error. Now the
4401 underlying error message should be appended to the second error message.
4403 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4405 PH/02 Reduce the amount of output that "make" produces by default. Full output
4406 can still be requested.
4408 PH/03 The warning log line about a condition test deferring for a "warn" verb
4409 was being output only once per connection, rather than after each
4410 occurrence (because it was using the same function as for successful
4411 "warn" verbs). This seems wrong, so I have changed it.
4413 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4414 it should not have, which might have caused a crash in the right
4415 circumstances, but probably never did.
4417 PH/04 Installed a modified version of Tony Finch's patch to make submission
4418 mode fix the return path as well as the Sender: header line, and to
4419 add a /name= option so that you can make the user's friendly name appear
4422 TF/03 Added the control = fakedefer ACL modifier.
4424 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4425 Mark Lowes for thorough testing.
4427 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4429 TK/03 Merged latest SRS patch from Miles Wilton.
4431 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4432 with the definition in sysexits.h (which is #included earlier).
4433 Fortunately, Exim does not actually use EX_OK. The code used to try to
4434 preserve the sysexits.h value, by assuming that macro definitions were
4435 scanned for macro replacements. I have been disabused of this notion,
4436 so now the code just undefines EX_OK before #including unistd.h.
4438 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4439 in the smtp transport. When a block could not be written in a single
4440 write() function, the timeout was being re-applied to each part-write.
4441 This seems wrong - if the receiver was accepting one byte at a time it
4442 would take for ever. The timeout is now adjusted when this happens. It
4443 doesn't have to be particularly precise.
4445 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4446 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4448 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4449 without a domain if the check to <postmaster@domain> fails.
4451 SC/01 Eximstats: added -xls and the ability to specify output files
4452 (patch written by Frank Heydlauf).
4454 SC/02 Eximstats: use FileHandles for outputting results.
4456 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4458 SC/04 Eximstats: fixed display of large numbers with -nvr option
4460 SC/05 Eximstats: fixed merging of reports with empty tables.
4462 SC/06 Eximstats: added the -include_original_destination flag
4464 SC/07 Eximstats: removed tabs and trailing whitespace.
4466 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4468 TK/06 MBOX spool code: Add real "From " MBOX separator line
4469 so the .eml file is really in mbox format (even though
4470 most programs do not really care). Patch from Alex Miller.
4472 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4473 The latter is generated from $received_to and is only set if the
4474 message has one envelope recipient. SA can use these headers,
4475 obviously out-of-the-box. Patch from Alex Miller.
4477 PH/08 The ${def test on a variable was returning false if the variable's
4478 value was "0", contrary to what the specification has always said!
4479 The result should be true unless the variable is empty.
4481 PH/09 The syntax error of a character other than { following "${if
4482 def:variable_name" (after optional whitespace) was not being diagnosed.
4483 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4484 accidental colon was present, for example, could give incorrect results.
4486 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4487 result is used (not including appendfile, where other changes are about
4490 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4491 This involved changing a lot of size variables from int to off_t. It
4492 should work with maildirs and everything.
4494 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4495 spamd dying while we are connected to it.
4497 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4500 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4501 function for BDB 4.3.
4503 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4505 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4506 the sources for PCRE have been re-arranged and more files are now
4509 PH/15 The code I had for printing potentially long long variables in PH/11
4510 above was not the best (it lost precision). The length of off_t variables
4511 is now inspected at build time, and an appropriate printing format (%ld
4512 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4513 to be "long long int" or "long int". This is needed for the internal
4514 formatting function string_vformat().
4516 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4517 the configuration file to be ":syslog", then the script "guesses" where
4518 the logs files are, rather than using the compiled in default. In our
4519 case the guess is not the same as the compiled default, so the script
4520 suddenly stopped working when I started to use syslog. The patch checks
4521 to see if log_file_path is "". If so, it attempts to read it from exim
4522 with no configuration file to get the compiled in version, before it
4523 falls back to the previous guessing code."
4525 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4526 implementing BATV in an Exim configuration. See NewStuff for the gory
4529 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4530 Makefile that are specific to HP-UX.
4532 PH/18 If the "use_postmaster" option was set for a recipient callout together
4533 with the "random" option, the postmaster address was used as the MAIL
4534 FROM address for the random test, but not for the subsequent recipient
4535 test. It is now used for both.
4537 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4538 patch removes a few documentation additions to RFC 3028, because the
4539 latest draft now contains them. It adds the new en;ascii-case comparator
4540 and a new error check for 8bit text in MIME parts. Comparator and
4541 require names are now matched exactly. I enabled the subaddress
4542 extension, but it is not well tested yet (read: it works for me)."
4544 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4545 rework some of the code of TK/09 above to avoid the hardwired use of
4546 "%lld" and "long long". Replaced the call to snprintf() with a call to
4549 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4550 records point to non-existent hosts", "retry timeout exceeded", and
4551 "retry time not reached for any host after a long failure period".
4553 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4554 experimental DomainKeys support:
4556 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4557 (2) On an error such as an illegally used "control", the wrong name for
4558 the control was given.
4560 These problems did NOT occur unless DomainKeys support was compiled.
4562 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4564 PH/24 Added ${if match_ip condition.
4566 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4567 databases so that it will be absolutely obvious if a crash occurs in the
4568 DB library. This is a regular occurrence (often caused by mis-matched
4571 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4572 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4573 picked up on a user's system that detects such things. There doesn't seem
4574 to be a gcc warning option for this - only an attribute that has to be
4575 put on the function's prototype. It seems that in Fedora Core 4 they have
4576 set this on a number of new functions. No doubt there will be more in due
4579 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4580 specify a transport. However, if an address that was verified by such a
4581 router was the subject of a callout, Exim crashed because it tried to
4582 read the rcpt_include_affixes from the non-existent transport. Now it
4583 just assumes that the setting of that option is false. This bug was
4584 introduced by 4.51/PH/31.
4586 PH/28 Changed -d+all to exclude +memory, because that information is very
4587 rarely of interest, but it makes the output a lot bigger. People tend to
4588 do -d+all out of habit.
4590 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4591 code in os-type was giving problems when libc.so lives in lib64, like on
4594 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4595 aren't the modern standard, and it seems that some systems' include files
4596 don't always have them. Exim was already checking for some of the newer
4597 ones like T_AAAA, and defining it itself. I've added checks for all the
4598 record types that Exim uses.
4600 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4601 not automatically generating a new one, as it is supposed to. This
4602 prevented TLS from working. If the file did exist, but contained invalid
4603 data, a new version was generated, as expected. It was only the case of a
4604 non-existent file that was broken.
4606 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4607 with a change in libdomainkeys > 0.64.
4609 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4610 from DNS. If the selector record carries the flag, it now has
4611 precedence over the domain-wide flag.
4613 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4615 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4616 the use of a port name that isn't defined in /etc/services) occurred, the
4617 message was deferred as in a normal delivery, and thus remained on the
4618 spool, instead of being failed because of the mua_wrapper setting. This
4619 is now fixed, and I tidied up some of the mua_wrapper messages at the
4622 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4623 the messages in a hash of arrays rather than using individual hashes.
4624 This is a bit cleaner and results in dramatic memory savings, albeit
4625 at a slight CPU cost.
4627 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4628 as requested by Marc Sherman.
4630 SC/10 Eximstats: added histograms for user specified patterns as requested
4633 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4635 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4636 fopen() in the content-scanning modules that did not already have it.
4642 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4643 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4645 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4647 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4649 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4650 to test Sieve filters that use "vacation".
4652 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4653 that changes the way the GnuTLS parameters are stored in the cache file.
4654 The new format can be generated externally. For backward compatibility,
4655 if the data in the cache doesn't make sense, Exim assumes it has read an
4656 old-format file, and it generates new data and writes a new file. This
4657 means that you can't go back to an older release without removing the
4660 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4661 work if there are any delivery delays because "one_time" forces the
4662 parent to be marked "delivered", so its unseen clone is never tried
4663 again. For this reason, Exim now forbids the simultaneous setting of
4666 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4667 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4668 entry if you want to know the details. The fix, however, neglected to
4669 consider the case when local delivery batching is involved. The test for
4670 "previously delivered" was not happening when checking to see if an
4671 address could be batched with a previous (undelivered) one; under
4672 certain circumstances this could lead to multiple deliveries to the same
4675 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4676 in its include files, and this causes problems building Exim.
4678 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4679 header_syntax) but Exim was just ignoring anything given after a slash.
4680 In particular, this caused confusion with an attempt to use "verify =
4681 reverse_host_lookup/defer_ok". An error is now given when options are
4682 supplied for verify items that do not have them. (Maybe reverse_host_
4683 lookup should have a defer_ok option, but that's a different point.)
4685 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4686 defined by RFC 821) to 2048, because there were problems with some AUTH
4687 commands, and RFC 1869 says the size should be increased for extended
4688 SMTP commands that take arguments.
4690 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4693 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4694 "unknown" error; now it says that the functionality isn't in the binary.
4696 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4697 an address' error message when a string expansion fails (syntax or
4698 whatever). Otherwise the password may appear in the log. Following change
4699 PH/42 below, there is no longer a chance of it appearing in a bounce
4702 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4704 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4705 of its addresses was ever tried. (Bugzilla bug #2.)
4707 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4708 the result incorrectly in the debug output. (It correctly added a newline
4709 to what was transported.)
4711 TF/01 Added $received_time.
4713 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4714 commented out examples of how to interface to a virus scanner and to
4715 SpamAssassin. Also added commented examples of av_scanner and
4716 spamd_address settings.
4718 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4719 and controls are allowed in which ACLs. There were a couple of minor
4720 errors. Some of the entries in the conditions table (which is a table of
4721 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4722 negation of where the condition IS allowed.
4724 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4726 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4727 header file does not have a version number, so I've had to invent a new
4728 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4729 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4730 radiusclient), but it was contributed by a Radius user.
4732 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4733 files or directories, for OpenSSL.
4735 PH/19 When an Exim process that is running as root has to create an Exim log
4736 file, it does so in a subprocess that runs as exim:exim so as to get the
4737 ownership right at creation (otherwise, other Exim processes might see
4738 the file with the wrong ownership). There was no test for failure of this
4739 fork() call, which would lead to the process getting stuck as it waited
4740 for a non-existent subprocess. Forks do occasionally fail when resources
4741 run out. I reviewed all the other calls to fork(); they all seem to check
4744 PH/20 When checking for unexpected SMTP input at connect time (before writing
4745 the banner), Exim was not dealing correctly with a non-positive return
4746 from the read() function. If the client had disconnected by this time,
4747 the result was a log entry for a synchronization error with an empty
4748 string after "input=" when read() returned zero. If read() returned -1
4749 (an event I could not check), uninitialized data bytes were printed.
4750 There were reports of junk text (parts of files, etc) appearing after
4753 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4755 PH/22 Added support for macro redefinition, and (re)definition in between
4756 driver and ACL definitions.
4758 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4759 forgetting to use the resulting value; it was using the unexpanded value.
4761 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4762 hadn't been configured. The fix is from Juergen Kreileder, who
4763 understands it better than I do:
4765 "Here's what I see happening with three configured cyrus_sasl
4766 authenticators configured (plain, login, cram-md5):
4768 On startup auth_cyrus_sasl_init() gets called for each of these.
4769 This means three calls to sasl_listmech() without a specified mech_list.
4770 => SASL tests which mechs of all available mechs actually work
4771 => three warnings about OTP not working
4772 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4774 With the patch, sasl_listmech() also gets called three times. But now
4775 SASL's mech_list option is set to the server_mech specified in the the
4776 authenticator. Or in other words, the answer from sasl_listmech()
4777 gets limited to just the mech you're testing for (which is different
4779 => the return list contains just 'plain' or 'login', 'cram-md5' or
4780 nothing depending on the value of ob->server_mech.
4782 I've just tested the patch: Authentication still works fine,
4783 unavailable mechs specified in the exim configuration are still
4784 caught, and the auth.log warnings about OTP are gone."
4786 PH/25 When debugging is enabled, the contents of the command line are added
4787 to the debugging output, even when log_selector=+arguments is not
4790 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4791 answer is "GNU", and only if the return is "GNU/something" is the answer
4794 PH/27 $acl_verify_message is now set immediately after the failure of a
4795 verification in an ACL, and so is available in subsequent modifiers. In
4796 particular, the message can be preserved by coding like this:
4798 warn !verify = sender
4799 set acl_m0 = $acl_verify_message
4801 Previously, $acl_verify_message was set only while expanding "message"
4802 and "log_message" when a very denied access.
4804 PH/28 Modified OS/os.c-Linux with
4806 -#ifndef OS_LOAD_AVERAGE
4807 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4809 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4810 nomenclature these days.)
4812 PH/29 Installed patch from the Sieve maintainer that adds the options
4813 sieve_useraddress and sieve_subaddress to the redirect router.
4815 PH/30 In these circumstances:
4816 . Two addresses routed to the same list of hosts;
4817 . First host does not offer TLS;
4818 . First host accepts first address;
4819 . First host gives temporary error to second address;
4820 . Second host offers TLS and a TLS session is established;
4821 . Second host accepts second address.
4822 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4823 and peerdn, if requested) that were in fact used only for the second
4826 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4827 attention to any local part prefix or suffix that was matched by the
4828 router that accepted the address. It now behaves in the same way as it
4829 does for delivery: the affixes are removed from the local part unless
4830 rcpt_include_affixes is set on the transport.
4832 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4833 timeout during the DATA phase of an incoming message.
4835 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4836 applied a patch sanctioned by the Sieve maintainer.
4838 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4839 the uid or gid is negative. A case of a negative gid caused this to be
4840 noticed. The fix allows for either to be negative.
4842 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4843 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4846 PH/36 The change PH/12 above was broken. Fixed it.
4848 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4849 the grounds that routing the same address twice would always produce the
4850 same answer. This might have been true once, but it is certainly no
4851 longer true now. Routing a child address may depend on the previous
4852 routing that produced that child. Some complicated redirection strategies
4853 went wrong when messages had multiple recipients, and made Exim's
4854 behaviour dependent on the order in which the addresses were given.
4856 I have moved the duplicate checking until after the routing is complete.
4857 Exim scans the addresses that are assigned to local and remote
4858 transports, and removes any duplicates. This means that more work will be
4859 done, as duplicates will always all be routed, but duplicates are
4860 presumably rare, so I don't expect this is of any significance.
4862 For deliveries to pipes, files, and autoreplies, the duplicate checking
4863 still happens during the routing process, since they are not going to be
4866 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4867 It corrects a timeout issue with spamd. This is Ian's comment: "The
4868 background is that sometimes spamd either never reads data from a
4869 connection it has accepted, or it never writes response data. The exiscan
4870 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4871 blindly assumes that writes won't block so it may never time out."
4873 PH/39 Allow G after quota size as well as K and M.
4875 PH/40 The value set for $authenticated_id in an authenticator may not contain
4876 binary zeroes or newlines because the value is written to log lines and
4877 to spool files. There was no check on this. Now the value is run through
4878 the string_printing() function so that such characters are converted to
4879 printable escape sequences.
4881 PH/41 $message_linecount is a new variable that contains the total number of
4882 lines in the message. Compare $body_linecount, which is the count for the
4885 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4886 bounce and delay warning messages, except in certain special cases, which
4889 (a) An SMTP error message from a remote host;
4890 (b) A message specified in a :fail: redirection;
4891 (c) A message specified in a "fail" command in a system filter;
4892 (d) A message specified in a FAIL return from the queryprogram router;
4893 (e) A message specified by the cannot_route_message router option.
4895 In these cases only, Exim does include the error details in bounce and
4896 warning messages. There are also a few cases where bland messages such
4897 as "unrouteable address" or "local delivery error" are given.
4899 PH/43 $value is now also set for the "else" part of a ${run expansion.
4901 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4902 being worked on, but at least Exim now implements the latest version to
4905 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4906 process to complete was treated as a delivery failure, a timeout while
4907 writing the message to the pipe was logged, but erroneously treated as a
4908 successful delivery. Such timeouts include transport filter timeouts. For
4909 consistency with the overall process timeout, these timeouts are now
4910 treated as errors, giving rise to delivery failures by default. However,
4911 there is now a new Boolean option for the pipe transport called
4912 timeout_defer, which, if set TRUE, converts the failures into defers for
4913 both kinds of timeout. A transport filter timeout is now identified in
4916 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4917 systems where "make" and "gmake" are different, calling "gmake" at top
4918 level broke things. I've arranged for the value of $(MAKE) to be passed
4919 from the Makefile to this script so that it can call the same version of
4923 A note about Exim versions 4.44 and 4.50
4924 ----------------------------------------
4926 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4927 changes of various kinds. As a consequence, a big documentation update was
4928 needed. This delayed the release for rather longer than seemed good, especially
4929 in the light of a couple of (minor) security issues. Therefore, the changes
4930 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4931 release. So 4.44 and 4.50 are in effect two different branches that both start
4934 I have left the 4.50 change log unchanged; it contains all the changes since
4935 4.43. The change log for 4.44 is below; many of its items are identical to
4936 those for 4.50. This seems to be the most sensible way to preserve the
4937 historical information.
4943 1. Minor wording change to the doc/README.SIEVE file.
4945 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4946 computation of the current number of files was incorrect.
4948 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4949 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4952 4. Give more explanation in the error message when the command for a transport
4953 filter fails to execute.
4955 5. There are several places where Exim runs a non-Exim command in a
4956 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4957 was being done only for the command run by the queryprogram router. It is
4958 now done for all such subprocesses. The other cases are: ${run, transport
4959 filters, and the commands run by the lmtp and pipe transports.
4961 6. Added CONFIGURE_GROUP build-time option.
4963 7. Some older OS have a limit of 256 on the maximum number of file
4964 descriptors. Exim was using setrlimit() to set 1000 as a large value
4965 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4966 systems. I've change it so that if it can't get 1000, it tries for 256.
4968 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4969 was an oversight, and furthermore, ever since the addition of extra
4970 controls (e.g. 4.43/32), the checks on when to allow different forms of
4971 "control" were broken. There should now be diagnostics for all cases when a
4972 control that does not make sense is encountered.
4974 9. Added the /retain_sender option to "control=submission".
4976 10. $recipients is now available in the predata ACL (oversight).
4978 11. Tidy the search cache before the fork to do a delivery from a message
4979 received from the command line. Otherwise the child will trigger a lookup
4980 failure and thereby defer the delivery if it tries to use (for example) a
4981 cached ldap connection that the parent has called unbind on.
4983 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4984 of $address_data from the recipient verification was clobbered by the
4985 sender verification.
4987 13. The value of address_data from a sender verification is now available in
4988 $sender_address_data in subsequent conditions in the ACL statement.
4990 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4992 15. Added a new option "connect=<time>" to callout options, to set a different
4995 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4996 was its contents. (It was OK if the option was not defined at all.)
4998 17. A "Completed" log line is now written for messages that are removed from
4999 the spool by the -Mrm option.
5001 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5002 information about exactly what failed.
5004 19. Added -dd to debug only the daemon process.
5006 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5007 handles timeouts, both on the server side and network timeouts. Renamed the
5008 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5010 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5011 It is now set to "smtps".
5013 22. $host_address is now set to the target address during the checking of
5014 ignore_target_hosts.
5016 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5017 being passed; this would have caused $sender_host_name to have been used if
5018 matching the list had actually called for a host name (not very likely,
5019 since this list is usually IP addresses). A host name is now passed as
5022 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5023 code that specifies a non-restarting handler (typically sigaction() in
5024 modern systems) in an attempt to fix a rare and obscure crash bug.
5026 25. Narrowed the window for a race in the daemon that could cause it to ignore
5027 SIGCHLD signals. This is not a major problem, because they are used only to
5028 wake it up if nothing else does.
5030 26. A malformed maildirsize file could cause Exim to calculate negative values
5031 for the mailbox size or file count. Odd effects could occur as a result.
5032 The maildirsize information is now recalculated if the size or filecount
5035 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5036 support for a long time. Removed HAVE_SYS_VFS_H.
5038 28. Installed the latest version of exipick from John Jetmore.
5040 29. In an address list, if the pattern was not a regular expression, an empty
5041 subject address (from a bounce message) matched only if the pattern was an
5042 empty string. Non-empty patterns were not even tested. This was the wrong
5043 because it is perfectly reasonable to use an empty address as part of a
5044 database query. An empty address is now tested by patterns that are
5045 lookups. However, all the other forms of pattern expect the subject to
5046 contain a local part and a domain, and therefore, for them, an empty
5047 address still always fails if the pattern is not itself empty.
5049 30. Exim went into a mad DNS loop when attempting to do a callout where the
5050 host was specified on an smtp transport, and looking it up yielded more
5051 than one IP address.
5053 31. Re-factored the code for checking spool and log partition space into a
5054 function that finds that data and another that does the check. The former
5055 is then used to implement four new variables: $spool_space, $log_space,
5056 $spool_inodes, and $log_inodes.
5058 32. The RFC2047 encoding function was originally intended for short strings
5059 such as real names; it was not keeping to the 75-character limit for
5060 encoded words that the RFC imposes. It now respects the limit, and
5061 generates multiple encoded words if necessary. To be on the safe side, I
5062 have increased the buffer size for the ${rfc2047: expansion operator from
5065 33. It is now permitted to omit both strings after an "if" condition; if the
5066 condition is true, the result is "true". As before, when the second string
5067 is omitted, a false condition yields an empty string. This makes it less
5068 cumbersome to write custom ACL and router conditions.
5070 34. Failure to deliver a bounce message always caused it to be frozen, even if
5071 there was an errors_to setting on the router. The errors_to setting is now
5074 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5075 canonical form (fully expanded) before being placed in
5076 $sender_host_address.
5078 36. The table in the code that translates DNS record types into text (T_A to
5079 "A" for instance) was missing entries for NS and CNAME. It is just possible
5080 that this could have caused confusion if both these types were looked up
5081 for the same domain, because the text type is used as part of Exim's
5082 per-process caching. But the chance of anyone hitting this buglet seems
5085 37. The dnsdb lookup has been extended in a number of ways.
5087 (1) There is a new type, "zns", which walks up the domain tree until it
5088 finds some nameserver records. It should be used with care.
5090 (2) There is a new type, "mxh", which is like "mx" except that it returns
5091 just the host names, not the priorities.
5093 (3) It is now possible to give a list of domains (or IP addresses) to be
5094 looked up. The behaviour when one of the lookups defers can be
5095 controlled by a keyword.
5097 (4) It is now possible to specify the separator character for use when
5098 multiple records are returned.
5100 38. The dnslists ACL condition has been extended: it is now possible to supply
5101 a list of IP addresses and/or domains to be looked up in a particular DNS
5104 39. Added log_selector=+queue_time_overall.
5106 40. When running the queue in the test harness, wait just a tad after forking a
5107 delivery process, to get repeatability of debugging output.
5109 41. Include certificate and key file names in error message when GnuTLS fails
5110 to set them up, because the GnuTLS error message doesn't include the name
5111 of the failing file when there is a problem reading it.
5113 42. Allow both -bf and -bF in the same test run.
5115 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5117 44. The "Exiscan patch" is now merged into the mainline Exim source.
5119 45. Sometimes the final signoff response after QUIT could fail to get
5120 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5121 < 0 before doing a fflush(). This bug looks as though it goes back to the
5122 introduction of TLS in release 3.20, but "sometimes" must have been rare
5123 because the tests only now provoked it.
5125 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5126 (this can affect the format of dates).
5128 47. exim_tidydb, when checking for the continued existence of a message for
5129 which it has found a message-specific retry record, was not finding
5130 messages that were in split spool directories. Consequently, it was
5131 deleting retry records that should have stayed in existence.
5133 48. Steve fixed some bugs in eximstats.
5135 49. The SPA authentication driver was not abandoning authentication and moving
5136 on to the next authenticator when an expansion was forced to fail,
5137 contradicting the general specification for all authenticators. Instead it
5138 was generating a temporary error. It now behaves as specified.
5140 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5141 (the order specifies the preference for clients). The order is now AES256,
5142 AES128, 3DES, ARCFOUR128.
5144 51. Small patch to Sieve code - explicitly set From: when generating an
5147 52. Exim crashed if a remote delivery caused a very long error message to be
5148 recorded - for instance if somebody sent an entire SpamAssassin report back
5149 as a large number of 550 error lines. This bug was coincidentally fixed by
5150 increasing the size of one of Exim's internal buffers (big_buffer) that
5151 happened as part of the Exiscan merge. However, to be on the safe side, I
5152 have made the code more robust (and fixed the comments that describe what
5155 53. Now that there can be additional text after "Completed" in log lines (if
5156 the queue_time_overall log selector is set), a one-byte patch to exigrep
5157 was needed to allow it to recognize "Completed" as not the last thing in
5160 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5161 patch that reportedly fixes this has been added. I am not expert enough to
5162 create a test for it. This is what the patch creator wrote:
5164 "I found a little strange behaviour of ldap code when working with
5165 Windows 2003 AD Domain, where users was placed in more than one
5166 Organization Units. When I tried to give exim partial DN, the exit code
5167 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5168 But simultaneously result of request was absolutely normal ldap result,
5169 so I produce this patch..."
5171 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5172 REFERENCE, so I have modified the code to exclude the patch when that macro
5175 55. Some experimental protocols are using DNS PTR records for new purposes. The
5176 keys for these records are domain names, not reversed IP addresses. The
5177 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5178 leaves it alone. Component reversal etc. now happens only for IP addresses.
5181 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5183 57. Double the size of the debug message buffer (to 2048) so that more of very
5184 long debug lines gets shown.
5186 58. The exicyclog utility now does better if the number of log files to keep
5187 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5189 59. Two changes related to the smtp_active_hostname option:
5191 (1) $smtp_active_hostname is now available as a variable.
5192 (2) The default for smtp_banner uses $smtp_active_hostname instead
5193 of $primary_hostname.
5195 60. The host_aton() function is supposed to be passed a string that is known
5196 to be a valid IP address. However, in the case of IPv6 addresses, it was
5197 not checking this. This is a hostage to fortune. Exim now panics and dies
5198 if the condition is not met. A case was found where this could be provoked
5199 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5200 components; fortuitously, this particular loophole had already been fixed
5201 by change 4.50/55 above.
5203 If there are any other similar loopholes, the new check in host_aton()
5204 itself should stop them being exploited. The report I received stated that
5205 data on the command line could provoke the exploit when Exim was running as
5206 exim, but did not say which command line option was involved. All I could
5207 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5208 running as the user.
5211 61. There was a buffer overflow vulnerability in the SPA authentication code
5212 (which came originally from the Samba project). I have added a test to the
5213 spa_base64_to_bits() function which I hope fixes it.
5216 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5217 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5219 63. The daemon start-up calls getloadavg() while still root for those OS that
5220 need the first call to be done as root, but it missed one case: when
5221 deliver_queue_load_max is set with deliver_drop_privilege. This is
5222 necessary for the benefit of the queue runner, because there is no re-exec
5223 when deliver_drop_privilege is set.
5225 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5226 This has been fixed.
5228 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5229 was in use, was not putting the data itself into the right store pool;
5230 consequently, it could be overwritten for a subsequent message in the same
5231 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5234 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5236 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5237 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5238 the function were treating the return as a boolean value, which happened to
5239 work because 0=false and not-0=true, but is not correct code.
5241 68. The host_aton() function was not handling scoped IPv6 addresses (those
5242 with, for example, "%eth0" on the end) correctly.
5244 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5245 negated items (that is, ~something) in unsigned ints. Some compilers
5246 apparently mutter when there is no cast.
5248 70. If an address verification called from an ACL failed, and did not produce a
5249 user-specific message (i.e. there was only a "system" message), nothing was
5250 put in $acl_verify_message. In this situation, it now puts the system
5253 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5254 session; change 4.31/43 added the unwanted input to the log line - except
5255 that it did not do this in the start of session case. It now does.
5257 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5258 This is wrong and can cause the other end to generate a synchronization
5259 error if it is another Exim or anything else that does the synchronization
5260 check. A QUIT command is no longer sent after a timeout.
5262 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5263 during host lookups.
5265 74. The defer_ok option of callout verification was not working if it was used
5266 when verifying addresses in header lines, that is, for this case:
5268 verify = header_sender/callout=defer_ok
5270 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5271 those file descriptors could be used for SMTP connections. If anything
5272 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5273 could be sent to the SMTP client, causing chaos. The daemon now opens
5274 stdin, stdout, and stderr to /dev/null when it puts itself into the
5277 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5278 log by default. The user can override this with suitable Perl magic.
5280 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5281 the log message when discard triggers, was not working for the DATA ACL or
5282 for the non-SMTP ACL.
5284 78. Error message wording change in sieve.c.
5286 79. If smtp_accept_max_per_host was set, the number of connections could be
5287 restricted to fewer than expected, because the daemon was trying to set up
5288 a new connection before checking whether the processes handling previous
5289 connections had finished. The check for completed processes is now done
5290 earlier. On busy systems, this bug wouldn't be noticed because something
5291 else would have woken the daemon, and it would have reaped the completed
5294 80. If a message was submitted locally by a user whose login name contained one
5295 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5296 It caused a spool format error. I have fixed the spool reading code. A
5297 related problem was that the "from" clause in the Received: line became
5298 illegal because of the space(s). It is now covered by ${quote_local_part.
5300 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5303 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5305 83. Patch from Sieve maintainer to fix unterminated string problem in
5306 "vacation" handling.
5308 84. Some minor changes to the Linux configuration files to help with other
5309 OS variants using glibc.
5311 85. One more patch for Sieve to update vacation handling to latest spec.
5314 ----------------------------------------------------
5315 See the note above about the 4.44 and 4.50 releases.
5316 ----------------------------------------------------
5322 1. Change 4.43/35 introduced a bug that caused file counts to be
5323 incorrectly computed when quota_filecount was set in an appendfile
5326 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5327 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5330 3. Give more explanation in the error message when the command for a transport
5331 filter fails to execute.
5333 4. There are several places where Exim runs a non-Exim command in a
5334 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5335 was being done only for the command run by the queryprogram router. It is
5336 now done for all such subprocesses. The other cases are: ${run, transport
5337 filters, and the commands run by the lmtp and pipe transports.
5339 5. Some older OS have a limit of 256 on the maximum number of file
5340 descriptors. Exim was using setrlimit() to set 1000 as a large value
5341 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5342 systems. I've change it so that if it can't get 1000, it tries for 256.
5344 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5345 was an oversight, and furthermore, ever since the addition of extra
5346 controls (e.g. 4.43/32), the checks on when to allow different forms of
5347 "control" were broken. There should now be diagnostics for all cases when a
5348 control that does not make sense is encountered.
5350 7. $recipients is now available in the predata ACL (oversight).
5352 8. Tidy the search cache before the fork to do a delivery from a message
5353 received from the command line. Otherwise the child will trigger a lookup
5354 failure and thereby defer the delivery if it tries to use (for example) a
5355 cached ldap connection that the parent has called unbind on.
5357 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5358 of $address_data from the recipient verification was clobbered by the
5359 sender verification.
5361 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5362 was its contents. (It was OK if the option was not defined at all.)
5364 11. A "Completed" log line is now written for messages that are removed from
5365 the spool by the -Mrm option.
5367 12. $host_address is now set to the target address during the checking of
5368 ignore_target_hosts.
5370 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5371 being passed; this would have caused $sender_host_name to have been used if
5372 matching the list had actually called for a host name (not very likely,
5373 since this list is usually IP addresses). A host name is now passed as
5376 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5377 code that specifies a non-restarting handler (typically sigaction() in
5378 modern systems) in an attempt to fix a rare and obscure crash bug.
5380 15. Narrowed the window for a race in the daemon that could cause it to ignore
5381 SIGCHLD signals. This is not a major problem, because they are used only to
5382 wake it up if nothing else does.
5384 16. A malformed maildirsize file could cause Exim to calculate negative values
5385 for the mailbox size or file count. Odd effects could occur as a result.
5386 The maildirsize information is now recalculated if the size or filecount
5389 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5390 support for a long time. Removed HAVE_SYS_VFS_H.
5392 18. Updated exipick to current release from John Jetmore.
5394 19. Allow an empty sender to be matched against a lookup in an address list.
5395 Previously the only cases considered were a regular expression, or an
5398 20. Exim went into a mad DNS lookup loop when doing a callout where the
5399 host was specified on the transport, if the DNS lookup yielded more than
5402 21. The RFC2047 encoding function was originally intended for short strings
5403 such as real names; it was not keeping to the 75-character limit for
5404 encoded words that the RFC imposes. It now respects the limit, and
5405 generates multiple encoded words if necessary. To be on the safe side, I
5406 have increased the buffer size for the ${rfc2047: expansion operator from
5409 22. Failure to deliver a bounce message always caused it to be frozen, even if
5410 there was an errors_to setting on the router. The errors_to setting is now
5413 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5414 canonical form (fully expanded) before being placed in
5415 $sender_host_address.
5417 24. Updated eximstats to version 1.33
5419 25. Include certificate and key file names in error message when GnuTLS fails
5420 to set them up, because the GnuTLS error message doesn't include the name
5421 of the failing file when there is a problem reading it.
5423 26. Expand error message when OpenSSL has problems setting up cert/key files.
5426 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5427 (this can affect the format of dates).
5429 28. exim_tidydb, when checking for the continued existence of a message for
5430 which it has found a message-specific retry record, was not finding
5431 messages that were in split spool directories. Consequently, it was
5432 deleting retry records that should have stayed in existence.
5434 29. eximstats updated to version 1.35
5435 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5436 1.35 - bugfix such that pie charts by volume are generated correctly
5438 30. The SPA authentication driver was not abandoning authentication and moving
5439 on to the next authenticator when an expansion was forced to fail,
5440 contradicting the general specification for all authenticators. Instead it
5441 was generating a temporary error. It now behaves as specified.
5443 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5444 (the order specifies the preference for clients). The order is now AES256,
5445 AES128, 3DES, ARCFOUR128.
5447 31. Small patch to Sieve code - explicitly set From: when generating an
5450 32. Exim crashed if a remote delivery caused a very long error message to be
5451 recorded - for instance if somebody sent an entire SpamAssassin report back
5452 as a large number of 550 error lines. This bug was coincidentally fixed by
5453 increasing the size of one of Exim's internal buffers (big_buffer) that
5454 happened as part of the Exiscan merge. However, to be on the safe side, I
5455 have made the code more robust (and fixed the comments that describe what
5458 33. Some experimental protocols are using DNS PTR records for new purposes. The
5459 keys for these records are domain names, not reversed IP addresses. The
5460 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5461 leaves it alone. Component reversal etc. now happens only for IP addresses.
5464 34. The host_aton() function is supposed to be passed a string that is known
5465 to be a valid IP address. However, in the case of IPv6 addresses, it was
5466 not checking this. This is a hostage to fortune. Exim now panics and dies
5467 if the condition is not met. A case was found where this could be provoked
5468 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5469 components; fortuitously, this particular loophole had already been fixed
5470 by change 4.50/55 or 4.44/33 above.
5472 If there are any other similar loopholes, the new check in host_aton()
5473 itself should stop them being exploited. The report I received stated that
5474 data on the command line could provoke the exploit when Exim was running as
5475 exim, but did not say which command line option was involved. All I could
5476 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5477 running as the user.
5480 35. There was a buffer overflow vulnerability in the SPA authentication code
5481 (which came originally from the Samba project). I have added a test to the
5482 spa_base64_to_bits() function which I hope fixes it.
5485 36. The daemon start-up calls getloadavg() while still root for those OS that
5486 need the first call to be done as root, but it missed one case: when
5487 deliver_queue_load_max is set with deliver_drop_privilege. This is
5488 necessary for the benefit of the queue runner, because there is no re-exec
5489 when deliver_drop_privilege is set.
5491 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5492 was in use, was not putting the data itself into the right store pool;
5493 consequently, it could be overwritten for a subsequent message in the same
5494 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5497 38. Sometimes the final signoff response after QUIT could fail to get
5498 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5499 < 0 before doing a fflush(). This bug looks as though it goes back to the
5500 introduction of TLS in release 3.20, but "sometimes" must have been rare
5501 because the tests only now provoked it.
5507 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5508 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5509 Now it returns an integer. A number of calls were still expecting a T/F
5510 return. Fortuitously, in all cases, the tests worked in OK situations,
5511 which is the norm. However, things would have gone wrong on any write
5512 failures on the smtp file descriptor. This function is used when sending
5513 messages over SMTP and also when doing verify callouts.
5515 2. When Exim is called to do synchronous delivery of a locally submitted
5516 message (the -odf or -odi options), it no longer closes stderr before doing
5519 3. Implemented the mua_wrapper option.
5521 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5523 5. Implemented the functions header_remove(), header_testname(),
5524 header_add_at_position(), and receive_remove_recipient(), and exported them
5527 6. If an ACL "warn" statement specified the addition of headers, Exim already
5528 inserted X-ACL-Warn: at the start if there was no header name. However, it
5529 was not making this test for the second and subsequent header lines if
5530 there were newlines in the string. This meant that an invalid header could
5531 be inserted if Exim was badly configured.
5533 7. Allow an ACL "warn" statement to add header lines at the start or after all
5534 the Received: headers, as well as at the end.
5536 8. Added the rcpt_4xx retry error code.
5538 9. Added postmaster_mailfrom=xxx to callout verification option.
5540 10. Added mailfrom=xxxx to the callout verification option, for verify=
5543 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5544 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5546 12. Inserted some casts to stop certain compilers warning when using pointer
5547 differences as field lengths or precisions in printf-type calls (mostly
5548 affecting debugging statements).
5550 13. Added optional readline() support for -be (dynamically loaded).
5552 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5553 same clock tick as a message's arrival, so that its received time was the
5554 same as the "first fail" time on the retry record, and that message
5555 remained on the queue past the ultimate address timeout, every queue runner
5556 would try a delivery (because it was past the ultimate address timeout) but
5557 after another failure, the ultimate address timeout, which should have then
5558 bounced the address, did not kick in. This was a "< instead of <=" error;
5559 in most cases the first failure would have been in the next clock tick
5560 after the received time, and all would be well.
5562 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5563 being recognized when the domain list was tested by the match_domain
5564 condition in an expansion string.
5566 16. Added the ${str2b64: operator.
5568 17. Exim was always calling setrlimit() to set a large limit for the number of
5569 processes, without checking whether the existing limit was already
5570 adequate. (It did check for the limit on file descriptors.) Furthermore,
5571 errors from getrlimit() and setrlimit() were being ignored. Now they are
5572 logged to the main and panic logs, but Exim does carry on, to try to do its
5573 job under whatever limits there are.
5575 18. Imported PCRE 5.0.
5577 19. Trivial typo in log message " temporarily refused connection" (the leading
5580 20. If the log selector return_path_on_delivery was set and an address was
5581 redirected to /dev/null, the delivery process crashed because it assumed
5582 that a return path would always be set for a "successful" delivery. In this
5583 case, the whole delivery is bypassed as an optimization, and therefore no
5586 21. Internal re-arrangement: the function for sending a challenge and reading
5587 a response while authentication was assuming a zero-terminated challenge
5588 string. It's now changed to take a pointer and a length, to allow for
5589 binary data in such strings.
5591 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5593 23. Exim was not respecting finduser_retries when seeking the login of the
5594 uid under which it was called; it was always trying 10 times. (The default
5595 setting of finduser_retries is zero.) Also, it was sleeping after the final
5596 failure, which is pointless.
5598 24. Implemented tls_on_connect_ports.
5600 25. Implemented acl_smtp_predata.
5602 26. If the domain in control=submission is set empty, Exim assumes that the
5603 authenticated id is a complete email address when it generates From: or
5604 Sender: header lines.
5606 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5607 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5608 chown and chgrp in /bin and hostname in /usr/bin.
5610 28. Exim was keeping the "process log" file open after each use, just as it
5611 does for the main log. This opens the possibility of it remaining open for
5612 long periods when the USR1 signal hits a daemon. Occasional processlog
5613 errors were reported, that could have been caused by this. Anyway, it seems
5614 much more sensible not to leave this file open at all, so that is what now
5617 29. The long-running daemon process does not normally write to the log once it
5618 has entered its main loop, and it closes the log before doing so. This is
5619 so that log files can straightforwardly be renamed and moved. However,
5620 there are a couple of unusual error situations where the daemon does write
5621 log entries, and I had neglected to close the log afterwards.
5623 30. The text of an SMTP error response that was received during a remote
5624 delivery was being truncated at 512 bytes. This is too short for some of
5625 the long messages that one sometimes sees. I've increased the limit to
5628 31. It is now possible to make retry rules that apply only when a message has a
5629 specific sender, in particular, an empty sender.
5631 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5632 it possible to be selective about when SMTP synchronization is enforced.
5634 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5636 32. Implemented hosts_connection_nolog.
5638 33. Added an ACL for QUIT.
5640 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5643 35. Added mailbox_size and mailbox_filecount to appendfile.
5645 36. Added control = no_multiline_responses to ACLs.
5647 37. There was a bug in the logic of the code that waits for the clock to tick
5648 in the case where the clock went backwards by a substantial amount such
5649 that the microsecond fraction of "now" was more than the microsecond
5650 fraction of "then" (but the whole seconds number was less).
5652 38. Added support for the libradius Radius client library this is found on
5653 FreeBSD (previously only the radiusclient library was supported).
5659 1. When certain lookups returned multiple values in the form name=value, the
5660 quoting of the values was not always being done properly. Specifically:
5661 (a) If the value started with a double quote, but contained no whitespace,
5663 (b) If the value contained whitespace other than a space character (i.e.
5664 tabs or newlines or carriage returns) it was not quoted.
5665 This fix has been applied to the mysql and pgsql lookups by writing a
5666 separate quoting function and calling it from the lookup code. The fix
5667 should probably also be applied to nisplus, ibase and oracle lookups, but
5668 since I cannot test any of those, I have not disturbed their existing code.
5670 2. A hit in the callout cache for a specific address caused a log line with no
5671 reason for rejecting RCPT. Now it says "Previous (cached) callout
5672 verification failure".
5674 3. There was an off-by-one bug in the queryprogram router. An over-long
5675 return line was truncated at 256 instead of 255 characters, thereby
5676 overflowing its buffer with the terminating zero. As well as fixing this, I
5677 have increased the buffer size to 1024 (and made a note to document this).
5679 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5680 when Exim is waiting for an SMTP response from a remote server, Exim
5681 restarts its select() call on the socket, thereby resetting its timeout.
5682 This is not a problem when such interrupts are rare. Somebody set up a cron
5683 job to run exiwhat every 2 minutes, which is less than the normal select()
5684 timeout (5 or 10 minutes). This meant that the select() timeout never
5685 kicked in because it was always reset. I have fixed this by comparing the
5686 time when an interrupt arrives with the time at the start of the first call
5687 to select(). If more time than the timeout has elapsed, the interrupt is
5688 treated as a timeout.
5690 5. Some internal re-factoring in preparation for the addition of Sieve
5691 extensions (by MH). In particular, the "personal" test is moved to a
5692 separate function, and given an option for scanning Cc: and Bcc: (which is
5693 not set for Exim filters).
5695 6. When Exim created an email address using the login of the caller as the
5696 local part (e.g. when creating a From: or Sender: header line), it was not
5697 quoting the local part when it contained special characters such as @.
5699 7. Installed new OpenBSD configuration files.
5701 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5702 try to make them clearer.
5704 9. Callout options, other than the timeout value, were being ignored when
5705 verifying sender addresses in header lines. For example, when using
5707 verify = header_sender/callout=no_cache
5709 the cache was (incorrectly) being used.
5711 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5712 only the Cygwin environment.
5714 11. When return_path_on_delivery was set as a log selector, if different remote
5715 addresses in the same message used different return paths and parallel
5716 remote delivery occurred, the wrong values would sometimes be logged.
5717 (Whenever a remote delivery process finished, the return path value from
5718 the most recently started remote delivery process was logged.)
5720 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5721 lines when AUTH and/or TLS are in use. This is the "received protocol"
5722 field. Exim used to use "asmtp" for authenticated SMTP, without any
5723 indication (in the protocol name) for TLS use. Now it follows the RFC and
5724 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5725 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5726 names appear in log lines as well as in Received: header lines.
5728 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5729 extensions, and comparison tests, and to fix some bugs.
5731 14. Changes to the "personal" filter test:
5733 (1) The test was buggy in that it was just doing the equivalent of
5734 "contains" tests on header lines. For example, if a user's address was
5735 anne@some.where, the "personal" test would incorrectly be true for
5737 To: susanne@some.where
5739 This test is now done by extracting each address from the header in turn,
5740 and checking the entire address. Other tests that are part of "personal"
5741 are now done using regular expressions (for example, to check local parts
5742 of addresses in From: header lines).
5744 (2) The list of non-personal local parts in From: addresses has been
5745 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5746 taken from the Sieve specification recommendations.
5748 (3) If the message contains any header line starting with "List-" it is
5749 treated as non-personal.
5751 (4) The test for "circular" in the Subject: header line has been removed
5752 because it now seems ill-conceived.
5754 15. Minor typos in src/EDITME comments corrected.
5756 16. Installed latest exipick from John Jetmore.
5758 17. If headers_add on a router specified a text string that was too long for
5759 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5760 of string_sprintf() is now avoided.
5762 18. $message_body_size was not set (it was always zero) when running the DATA
5763 ACL and the local_scan() function.
5765 19. For the "mail" command in an Exim filter, no default was being set for
5766 the once_repeat time, causing a random time value to be used if "once" was
5767 specified. (If the value happened to be <= 0, no repeat happened.) The
5768 default is now 0s, meaning "never repeat". The "vacation" command was OK
5769 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5770 (I found it when inspecting the code).
5772 20. There is now an overall timeout for performing a callout verification. It
5773 defaults to 4 times the callout timeout, which applies to individual SMTP
5774 commands during the callout. The overall timeout applies when there is more
5775 than one host that can be tried. The timeout is checked before trying the
5776 next host. This prevents very long delays if there are a large number of
5777 hosts and all are timing out (e.g. when the network connections are timing
5778 out). The value of the overall timeout can be changed by specifying an
5779 additional sub-option for "callout", called "maxwait". For example:
5781 verify = sender/callout=5s,maxwait=20s
5783 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5784 to the end before writing, but this should make it even safer).
5786 22. Exim was forgetting that it had advertised PIPELINING for the second and
5787 subsequent messages on an SMTP connection. It was also not resetting its
5788 memory on STARTTLS and an internal HELO.
5790 23. When Exim logs an SMTP synchronization error within a session, it now
5791 records whether PIPELINING has been advertised or not.
5793 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5794 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5795 rather than long int.
5797 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5799 26. Added the never_mail option to autoreply.
5805 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5806 crash if the getsockname() call failed; this can happen if a connection is
5807 closed very soon after it is established. The problem was simply in the
5808 order in which certain operations were done, causing Exim to try to write
5809 to the SMTP stream before it had set up the file descriptor. The bug has
5810 been fixed by making things happen in the correct order.
5816 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5817 before the connection was closed, thus losing the rejection response.
5819 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5820 some early Solaris releases, but causes trouble in current releases where
5821 socklen_t is defined.
5823 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5826 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5829 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5830 that releasing the top of it at the end releases what was used for sub-
5831 expansions (unless the block got too big). However, discard this block if
5832 the first thing is a variable or header, so that we can use its block when
5833 it is dynamic (useful for very large $message_headers, for example).
5835 6. Lookups now cache *every* query, not just the most recent. A new, separate
5836 store pool is used for this. It can be recovered when all lookup caches are
5837 flushed. Lookups now release memory at the end of their result strings.
5838 This has involved some general refactoring of the lookup sources.
5840 7. Some code has been added to the store_xxx() functions to reduce the amount
5841 of flapping under certain conditions.
5843 8. log_incoming_interface used to affect only the <= reception log lines. Now
5844 it causes the local interface and port to be added to several more SMTP log
5845 lines, for example "SMTP connection from", and rejection lines.
5847 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5849 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5851 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5852 could be overwritten at the end of the current message (or the start of a
5853 new message if it was set in a HELO ACL). The value is now preserved for
5854 the duration of the SMTP connection.
5856 12. If a transport had a headers_rewrite setting, and a matching header line
5857 contained an unqualified address, that address was qualified, even if it
5858 did not match any rewriting rules. The underlying bug was that the values
5859 of the flags that permit the existence of unqualified sender and recipient
5860 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5861 non-local messages, and by -bnq for local messages) were not being
5862 preserved with the message after it was received.
5864 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5865 "next input=" as part of the text comprising the host identity instead of
5866 the correct text. The code was using the same buffer for two different
5867 strings. However, depending on which order the printing function evaluated
5868 its arguments, the bug did not always show up. Under Linux, for example, my
5869 test suite worked just fine.
5871 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5872 doesn't work with some older versions of Perl. It has been changed to "my",
5873 which in any case is probably the better facility to use.
5875 15. A really picky compiler found some instances of statements for creating
5876 error messages that either had too many or two few arguments for the format
5879 16. The size of the buffer for calls to the DNS resolver has been increased
5880 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5881 for addresses that have a lot of PTR records. This alleviates a problem; it
5882 does not fully solve it.
5884 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5885 buffer now truncates the list and logs the incident, which is the same
5886 action as happens when Exim is looking up a host name and its aliases.
5887 Previously in this situation something unpredictable would happen;
5888 sometimes it was "internal error: store_reset failed".
5890 18. If a server dropped the connection unexpectedly when an Exim client was
5891 using GnuTLS and trying to read a response, the client delivery process
5892 crashed while trying to generate an error log message.
5894 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5895 string, for example:
5897 warn message = H1: something\nH2: something
5899 the text was added as a single header line from Exim's point of view
5900 though it ended up OK in the delivered message. However, searching for the
5901 second and subsequent header lines using $h_h2: did not work. This has been
5902 fixed. Similarly, if a system filter added multiple headers in this way,
5903 the routers could not see them.
5905 20. Expanded the error message when iplsearch is called with an invalid key to
5906 suggest using net-iplsearch in a host list.
5908 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5909 ACLs are no longer actually imposed (and a message to that effect is
5912 22. If a "gecos" field in a passwd entry contained escaped characters, in
5913 particular, if it contained a \" sequence, Exim got it wrong when building
5914 a From: or a Sender: header from that name. A second bug also caused
5915 incorrect handling when an unquoted " was present following a character
5916 that needed quoting.
5918 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5919 was not being matched caselessly.
5921 24. Arranged for all hyphens in the exim.8 source to be escaped with
5924 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5925 properly. Recipient callout cache records were still being keyed to include
5926 the sender, even when use_sender was set false. This led to far more
5927 callouts that were necessary. The sender is no longer included in the key
5928 when use_sender is false.
5930 26. Added "control = submission" modifier to ACLs.
5932 27. Added the ${base62d: operator to decode base 62 numbers.
5934 28. dnsdb lookups can now access SRV records.
5936 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5937 the configuration file.
5939 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5940 (-v) mode. This makes the output for a verbose queue run more intelligible.
5942 31. Added a use_postmaster feature to recipient callouts.
5944 32. Added the $body_zerocount variable, containing the number of binary zero
5945 bytes in the message body.
5947 33. The time of last modification of the "new" subdirectory is now used as the
5948 "mailbox time last read" when there is a quota error for a maildir
5951 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5953 35. Added +ignore_unknown as a special item in host lists.
5955 36. Code for decoding IPv6 addresses in host lists is now included, even if
5956 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5957 address was recognized as an IP address, but was then not correctly decoded
5958 into binary, causing unexpected and incorrect effects when compared with
5965 1. Very minor rewording of debugging text in manualroute to say "list of
5966 hosts" instead of "hostlist".
5968 2. If verify=header_syntax was set, and a header line with an unqualified
5969 address (no domain) and a large number of spaces between the end of the
5970 name and the colon was received, the reception process suffered a buffer
5971 overflow, and (when I tested it) crashed. This was caused by some obsolete
5972 code that should have been removed. The fix is to remove it!
5974 3. When running in the test harness, delay a bit after writing a bounce
5975 message to get a bit more predictability in the log output.
5977 4. Added a call to search_tidyup() just before forking a reception process. In
5978 theory, someone could use a lookup in the expansion of smtp_accept_max_
5979 per_host which, without the tidyup, could leave open a database connection.
5981 5. Added the variables $recipient_data and $sender_data which get set from a
5982 lookup success in an ACL "recipients" or "senders" condition, or a router
5983 "senders" option, similar to $domain_data and $local_part_data.
5985 6. Moved the writing of debug_print from before to after the "senders" test
5988 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5989 problems for message scanning, either using a data ACL, or using
5990 local_scan() because the Received: header was not generated till after they
5991 were called (in order to set the time as the time of reception completion).
5992 I have revised the way this works. The header is now generated after the
5993 body is received, but before the ACL or local_scan() are called. After they
5994 are run, the timestamp in the header is updated.
6000 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6001 before starting a queue runner without re-exec. This happened only when
6002 deliver_drop_privilege was set or when the Exim user was set to root. The
6003 effect of the bug was that timeouts during subsequent deliveries caused
6004 crashes instead of being properly handled. The handler is now left at its
6005 default (and expected) setting.
6007 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6008 message, again when deliver_drop_privilege is set or Exim is run as root.
6009 The bug described in (1) was not present in this case, but the tidying up
6010 of the other signals was missing. I have made the two cases consistent.
6012 3. The ignore_target_hosts setting on a manualroute router was being ignored
6013 for hosts that were looked up using the /MX notation.
6015 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6018 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6019 operated on the sender address. After changing the $sender_address to <>
6020 for the sender address verify, Exim was re-instated it as the original
6021 (before rewriting) address, but remembering that it had rewritten it, so it
6022 wasn't rewriting it again. This bug also had the effect of breaking the
6023 sender address verification caching when the sender address was rewritten.
6025 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6026 This has been changed so that if the ip literal address matches
6027 ignore_target_hosts, the router declines.
6029 7. Added expansion conditions match_domain, match_address, and match_local_
6030 part (NOT match_host).
6032 8. The placeholder for the Received: header didn't have a length field set.
6034 9. Added code to Exim itself and to exim_lock to test for a specific race
6035 condition that could lead to file corruption when using MBX delivery. The
6036 issue is with the lockfile that is created in /tmp. If this file is removed
6037 after a process has opened it but before that process has acquired a lock,
6038 there is the potential for a second process to recreate the file and also
6039 acquire a lock. This could lead to two Exim processes writing to the file
6040 at the same time. The added code performs the same test as UW imapd; it
6041 checks after acquiring the lock that its file descriptor still refers to
6042 the same named file.
6044 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6045 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6048 11. Added the smtp_active_hostname option. If used, this will typically be made
6049 to depend on the incoming interface address. Because $interface_address is
6050 not set up until the daemon has forked a reception process, error responses
6051 that can happen earlier (such as "too many connections") no longer contain
6054 12. If an expansion in a condition on a "warn" statement fails because a lookup
6055 defers, the "warn" statement is abandoned, and the next ACL statement is
6056 processed. Previously this caused the whole ACL to be aborted.
6058 13. Added the iplsearch lookup type.
6060 14. Added ident_timeout as a log selector.
6062 15. Added tls_certificate_verified as a log selector.
6064 16. Added a global option tls_require_ciphers (compare the smtp transport
6065 option of the same name). This controls incoming TLS connections.
6067 17. I finally figured out how to make tls_require_ciphers do a similar thing
6068 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6069 before starting the TLS session.
6071 18. Tabs are now shown as \t in -bP output.
6073 19. If the log selector return_path_on_delivery was set, Exim crashed when
6074 bouncing a message because it had too many Received: header lines.
6076 20. If two routers both had headers_remove settings, and the first one included
6077 a superfluous trailing colon, the final name in the first list and the
6078 first name in the second list were incorrectly joined into one item (with a
6079 colon in the middle).
6085 1. Added -C and -D options to the exinext utility, mainly to make it easier
6086 to include in the automated testing, but these could be helpful when
6087 multiple configurations are in use.
6089 2. The exinext utility was not formatting the output nicely when there was
6090 an alternate port involved in the retry record key, nor when there was a
6091 message id as well (for retries that were specific to a specific message
6092 and a specific host). It was also confused by IPv6 addresses, because of
6093 the additional colons they contain. I have fixed the IPv4 problem, and
6094 patched it up to do a reasonable job for IPv6.
6096 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6097 delivery, the log line now contains "pipelined" if PIPELINING was used.
6099 4. An SMTP transport process used to panic and die if the bind() call to set
6100 an explicit outgoing interface failed. This has been changed; it is now
6101 treated in the same way as a connect() failure.
6103 5. A reference to $sender_host_name in the part of a conditional expansion
6104 that was being skipped was still causing a DNS lookup. This no longer
6107 6. The def: expansion condition was not recognizing references to header lines
6108 that used bh_ and bheader_.
6110 7. Added the _cache feature to named lists.
6112 8. The code for checking quota_filecount in the appendfile transport was
6113 allowing one more file than it should have been.
6115 9. For compatibility with Sendmail, the command line option
6123 and sets the incoming protocol and host name (for trusted callers). The
6124 host name and its colon can be omitted when only the protocol is to be set.
6125 Note the Exim already has two private options, -pd and -ps, that refer to
6126 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6127 "s", but I don't think that's a major issue.
6129 10. A number of refactoring changes to the code, none of which should affect
6132 (a) The number of logging options was getting close to filling up the
6133 32-bit word that was used as a bit map. I have split them into two classes:
6134 those that are passed in the argument to log_write(), and those that are
6135 only ever tested independently outside of that function. These are now in
6136 separate 32-bit words, so there is plenty of room for expansion again.
6137 There is no change in the user interface or the logging behaviour.
6139 (b) When building, for example, log lines, the code previously used a
6140 macro that called string_cat() twice, in order to add two strings. This is
6141 not really sufficiently general. Furthermore, there was one instance where
6142 it was actually wrong because one of the argument was used twice, and in
6143 one call a function was used. (As it happened, calling the function twice
6144 did not affect the overall behaviour.) The macro has been replaced by a
6145 function that can join an arbitrary number of extra strings onto a growing
6148 (c) The code for expansion conditions now uses a table and a binary chop
6149 instead of a serial search (which was left over from when there were very
6150 few conditions). Also, it now recognizes conditions like "pam" even when
6151 the relevant support is not compiled in: a suitably worded error message is
6152 given if an attempt is made to use such a condition.
6154 11. Added ${time_interval:xxxxx}.
6156 12. A bug was causing one of the ddress fields not to be passed back correctly
6157 from remote delivery subprocesses. The field in question was not being
6158 subsequently used, so this caused to problems in practice.
6160 13. Added new log selectors queue_time and deliver_time.
6162 14. Might have fixed a bug in maildirsizefile handling that threw up
6163 "unexpected character" debug warnings, and recalculated the data
6164 unnecessarily. In any case, I expanded the warning message to give more
6167 15. Added the message "Restricted characters in address" to the statements in
6168 the default ACL that block characters like @ and % in local parts.
6170 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6171 Three changes have been made:
6173 (a) There was a serious bug; a negative response to MAIL caused the whole
6174 recipient domain to be cached as invalid, thereby blocking all messages
6175 to all local parts at the same domain, from all senders. This bug has
6176 been fixed. The domain is no longer cached after a negative response to
6177 MAIL if the sender used is not empty.
6179 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6182 (c) A new callout option, "use_sender" has been added for people who want
6183 the modified behaviour.
6189 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6192 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6193 indeed breaks things for older releases.
6195 3. Added additional logging to the case where there is a problem reading data
6196 from a filter that is running in a subprocess using a pipe, in order to
6197 try to track down a specific problem.
6199 4. Testing facility fudge: when running in the test harness and attempting
6200 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6201 getting "No route to host". Convert this to a timeout.
6203 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6206 6. Some OS don't have socklen_t but use size_t instead. This affects the
6207 fifth argument of getsockopt() amongst other things. This is now
6208 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6209 can be set for individual OS. I have set it for SunOS5, OSF1, and
6210 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6211 some earlier ones do not.
6213 7. Change 4.30/15 was not doing the test caselessly.
6215 8. The standard form for an IPv6 address literal was being rejected by address
6216 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6217 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6218 this, as well as the form without the "IPv6" on the front (but only when
6219 address literals are enabled, of course).
6221 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6223 10. Exim crashed if a message with an empty sender address specified by -f
6224 encountered a router with an errors_to setting. This could be provoked only
6225 by a command such as
6229 where an empty string was supplied; "<>" did not hit this bug.
6231 11. Installed PCRE release 4.5.
6233 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6234 remained set. It is now erased.
6236 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6237 times from message ids (which are base 36 rather than the normal 62).
6239 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6240 were being counted as actual protocol errors, and logged if the log
6241 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6242 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6243 and DATA following a set of rejected RCPTs do not count as protocol errors.
6244 In other words, Exim assumes they were pipelined, though this may not
6245 actually be the case. Of course, in all cases the client gets an
6246 appropriate error code.
6248 15. If a lookup fails in an ACL condition, a message about the failure may
6249 be available; it is used if testing the ACL cannot continue, because most
6250 such messages specify what the cause of the deferral is. However, some
6251 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6252 that caused an old message to be retained and used if a later statement
6253 caused a defer, replacing the real cause of the deferral.
6255 16. If an IP address had so many PTR records that the DNS lookup buffer
6256 was not large enough to hold them, Exim could crash while trying to process
6257 the truncated data. It now detects and logs this case.
6259 17. Further to 4.21/58, another change has been made: if (and only if) the
6260 first line of a message (the first header line) ends with CRLF, a bare LF
6261 in a subsequent header line has a space inserted after it, so as not to
6262 terminate the header.
6264 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6265 unnecessarily, used atoi() instead of strtol(), and didn't check the
6266 termination when getting file sizes from file names by regex.
6268 19. Completely re-implemented the support for maildirsize files, in the light
6269 of a number of problems with the previous contributed implementation
6270 (4.30/29). In particular:
6272 . If the quota is zero, the maildirsize file is maintained, but no quota is
6275 . If the maildir directory does not exist, it is created before any attempt
6276 to write a maildirsize file.
6278 . The quota value in the file is just a cache; if the quota is changed in
6279 the transport, the new value overrides.
6281 . A regular expression is available for excluding directories from the
6284 20. The autoreply transport checks the characters in options that define the
6285 message's headers; it allows continued headers, but it was checking with
6286 isspace() after an embedded newline instead of explicitly looking for a
6289 21. If all the "regular" hosts to which an address was routed had passed their
6290 expiry times, and had not reached their retry times, the address was
6291 bounced, even if fallback hosts were defined. Now Exim should go on to try
6294 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6295 equivalent code in the SMTP transport. Some hosts send humungous responses
6296 to HELO/EHLO, more than 1024 it seems.
6298 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6299 gives compiler warnings in some environments. I've now done it "properly",
6302 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6303 (because of problems with the built-in one) was declared to return uschar *
6304 instead of char *, causing compiler failure.
6306 25. Fixed a file descriptor leak when processing alias/forward files.
6308 26. Fixed a minor format string issue in dbfn.c.
6310 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6312 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6313 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6316 29. When a sender address is verified, it is cached, to save repeating the test
6317 when there is more than one recipient in a message. However, when the
6318 verification involves a callout, it is possible for different callout
6319 options to be set for different recipients. It is too complicated to keep
6320 track of this in the cache, so now Exim always runs a verification when a
6321 callout is required, relying on the callout cache for the optimization.
6322 The overhead is duplication of the address routing, but this should not be
6325 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6326 to be verified with callout=postmaster, and the main callout worked but the
6327 postmaster check failed, the verification correctly failed. However, if a
6328 subsequent RCPT command asked for sender verification *without* the
6329 postmaster check, incorrect caching caused this verification also to fail,
6332 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6333 it was not caching the DNS options (qualify_single, search_parents) that
6334 were used when the lookup failed. A subsequent lookup with different
6335 options therefore always gave the same answer, though there were cases
6336 where it should not have. (Example: a "domains = !$mx_any" option on a
6337 dnslookup router: the "domains" option is always processed without any
6338 widening, but the router might have qualify_single set.) Now Exim uses the
6339 cached value only when the same options are set.
6341 32. Added John Jetmore's "exipick" utility to the distribution.
6343 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6344 than a timeout (e.g. a certificate is required, and is not provided), an
6345 Exim server now closes the connection immediately. Previously it waited for
6346 the client to close - but if the client is SSL, it seems that they each
6347 wait for each other, leading to a delay before one of them times out.
6349 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6350 maintained 0.8.x compatibility because I don't think many are using it, and
6351 it is clearly obsolete.
6353 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6356 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6357 client certificate was expired. A simple patch fixes this, though I don't
6358 understand the full logic of why the verify callback is called multiple
6361 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6362 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6363 which causes problems with some clients (such as the Certicom SSL Plus
6364 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6365 disables the coutermeasure allowing Eudora to connect."
6367 38. Exim was not checking that a write() to a log file succeeded. This could
6368 lead to Bad Things if a log got too big, in particular if it hit a file
6369 size limit. Exim now panics and dies if it cannot write to a log file, just
6370 as it does if it cannot open a log file.
6372 39. Modified OS/Makefile-Linux so that it now contains
6374 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6376 The two -D definitions ensure that Exim is compiled with large file
6377 support, which makes it possible to handle log files that are bigger than
6380 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6381 instance) a domain was checked against a named list that involved a lookup,
6382 causing $domain_data to be set, then another domain was checked against the
6383 same list, then the first domain was re-checked, the value of $domain_data
6384 after the final check could be wrong. In particular, if the second check
6385 failed, it could be set empty. This bug probably also applied to
6388 41. The strip_trailing_dot option was not being applied to the address given
6389 with the -f command-line option.
6391 42. The code for reading a message's header from the spool was incrementing
6392 $received_count, but never initializing it. This meant that the value was
6393 incorrect (doubled) while delivering a message in the same process in which
6394 it was received. In the most common configuration of Exim, this never
6395 happens - a fresh exec is done - but it can happen when
6396 deliver_drop_privilege is set.
6398 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6399 it now includes up to 150 characters of the unexpected data in the log
6402 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6403 and building data strings. The size of both of these buffers was 10 000
6404 bytes - far larger than anybody would *ever* want, thought I. Needless to
6405 say, somebody hit the limit. I have increased the maximum line length to
6406 20 000 and the maximum data length of concatenated lines to 100 000. I have
6407 also fixed two bugs, because there was no checking on these buffers. Tsk,
6408 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6409 buffer is too small.
6411 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6412 lsearch lookups. Now it does.
6414 46. When parsing a route_list item in a manualroute router, a fixed-length
6415 buffer was used for the list of hosts. I made this 1024 bytes long,
6416 thinking that nobody would ever have a list of hosts that long. Wrong.
6417 Somebody had a whole pile of complicated expansion conditions, and the
6418 string was silently truncated, leading to an expansion error. It turns out
6419 that it is easier to change to an unlimited length (owing to other changes
6420 that have happened since this code was originally written) than to build
6421 structure for giving a limitation error. The length of the item that
6422 expands into the list of hosts is now unlimited.
6424 47. The lsearch lookup could not handle data where the length of text line was
6425 more than 4095 characters. Such lines were truncated, leading to shortened
6426 data being returned. It should now handle lines of any length.
6428 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6429 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6432 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6433 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6434 no longer happens in comment lines. A list of replacements is now placed
6435 at the head of all of the source files, except those whose only change is
6436 to replace PERL_COMMAND in the very first #! line.
6438 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6439 messages on the queue, with a bottom-up merge sort, using code contributed
6440 by Michael Haardt. This should make operations like -bp somewhat faster on
6441 large queues. It won't affect queue runners, except when queue_run_in_order
6444 51. Installed eximstats 1.31 in the distribution.
6446 52. Added support for SRV lookups to the dnslookup router.
6448 53. If an ACL referred to $message_body or $message_body_end, the value was not
6449 reset for any messages that followed in the same SMTP session.
6451 54. The store-handling optimization for building very long strings was not
6452 differentiating between the different store pools. I don't think this
6453 actually made any difference in practice, but I've tidied it.
6455 55. While running the routers to verify a sender address, $sender_address
6456 was still set to the sender address. This is wrong, because when routing to
6457 send a bounce to the sender, it would be empty. Therefore, I have changed
6458 it so that, while verifying a sender address, $sender_address is set to <>.
6459 (There is no change to what happens when verifying a recipient address.)
6461 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6462 target A or AAAA records (if not already returned) without resetting the
6463 qualify_single or search_parents options of the DNS resolver. These are
6464 inappropriate in this case because the targets of MX and SRV records must
6465 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6466 target that, when qualified, matched something in the local domain. These
6467 two options are now turned off when doing these lookups.
6469 57. It seems that at least some releases of Reiserfs (which does not have the
6470 concept of a fixed number of inodes) returns zero and not -1 for the
6471 number of available inodes. This interacted badly with check_spool_inodes,
6472 which assumed that -1 was the "no such thing" setting. What I have done is
6473 to check that the total number of inodes is greater than zero before doing
6474 the test of how many are available.
6476 58. When a "warn" ACL statement has a log_message modifier, the message is
6477 remembered, and not repeated. This is to avoid a lot of repetition when a
6478 message has many recipients that cause the same warning to be written.
6479 However, Exim was preserving the list of already written lines for an
6480 entire SMTP session, which doesn't seem right. The memory is now reset if a
6481 new message is started.
6483 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6484 debugging output unless log_rewrite was also set.
6486 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6487 of (int)(handle) when we know that handle contains (void *)(-1).
6489 61. The Exim daemon panic-logs an error return when it closes the incoming
6490 connection. However "connection reset by peer" seems to be common, and
6491 isn't really an error worthy of noting specially, so that particular error
6494 62. When Exim is trying to find all the local interfaces, it used to panic and
6495 die if the ioctl to get the interface flags failed. However, it seems that
6496 on at least one OS (Solaris 9) it is possible to have an interface that is
6497 included in the list of interfaces, but for which you get a failure error
6498 for this call. This happens when the interface is not "plumbed" into a
6499 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6500 failure of the "get flags" call assumes that the interface is down.
6502 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6503 makes life easier for people who are doing arithmetic on fields extracted
6504 from dates, where you often get leading zeros that should not be
6505 interpreted as octal.
6507 64. Added qualify_domain to the redirect router, to override the global
6510 65. If a pathologically long header line contained very many addresses (the
6511 report of this problem mentioned 10 000) and each of them was rewritten,
6512 Exim could use up a very large amount of memory. (It kept on making new
6513 copies of the header line as it rewrote, and never released the old ones.)
6514 At the expense of a bit more processing, the header rewriting function has
6515 been changed so that it no longer eats memory in this way.
6517 66. The generation of the Received: header has been moved from the time that a
6518 message starts to be received, to the time that it finishes. The timestamp
6519 in the Received: header should now be very close to that of the <= log
6520 line. There are two side-effects of this change:
6522 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6523 logged header lines no longer include the local Received: line, because
6524 it has not yet been created. The same applies to a copy of the message
6525 that is returned to a non-SMTP sender when a message is rejected.
6527 (b) When a filter file is tested using -bf, no additional Received: header
6528 is added to the test message. After some thought, I decided that this
6531 This change does not affect the value of $received_for. It is still set
6532 after address rewriting, but before local_scan() is called.
6534 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6536 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6537 gave an unhelpful panic error message, and a defer error. I have managed to
6538 change this behaviour so that it now rejects any supplied certificate,
6539 which seems right, as the list of acceptable certificates is empty.
6541 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6542 gave an unhelpful defer error. I have not managed to make this reject any
6543 supplied certificates, but the error message it gives is "no certificate
6544 supplied", which is not helpful.
6546 70. exigrep's output now also includes lines that are not associated with any
6547 message, but which match the given pattern. Implemented by a patch from
6548 Martin Sluka, which also tidied up the Perl a bit.
6550 71. Recipient callout verification, like sender verification, was using <> in
6551 the MAIL FROM command. This isn't really the right thing, since the actual
6552 sender may affect whether the remote host accepts the recipient or not. I
6553 have changed it to use the actual sender in the callout; this means that
6554 the cache record is now keyed on a recipient/sender pair, not just the
6555 recipient address. There doesn't seem to be a real danger of callout loops,
6556 since a callout by the remote host to check the sender would use <>.
6557 [SEE ABOVE: changed after hitting problems.]
6559 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6560 temporary errors. However, in the case of such a code being given after
6561 the end of a data transmission (i.e. after ".") Exim was failing to write
6562 a retry record for the message. (Yes, there was some broken host that was
6563 actually sending 8xx at this point.)
6565 73. An unknown lookup type in a host list could cause Exim to panic-die when
6566 the list was checked. (An example that provoked this was putting <; in the
6567 middle of a list instead of at the start.) If this happened during a DATA
6568 ACL check, a -D file could be left lying around. This kind of configuration
6569 error no longer causes Exim to die; instead it causes a defer error. The
6570 incident is still logged to the main and panic logs.
6572 74. Buglet left over from Exim 3 conversion. The message "too many messages
6573 in one connection" was written to the rejectlog but not the mainlog, except
6574 when address rewriting (yes!) was being logged.
6576 75. Added write_rejectlog option.
6578 76. When a system filter was run not as root (that is, when system_filter_user
6579 was set), the values of the $n variables were not being returned to the
6580 main process; thus, they were not subsequently available in the $sn
6583 77. Added +return_path_on_delivery log selector.
6585 78. A connection timeout was being treated differently from recipients deferred
6586 when testing hosts_max_try with a message that was older than the host's
6587 retry timeout. (The host should not be counted, thus allowing all hosts to
6588 be tried at least once before bouncing.) This may have been the cause of an
6589 occasionally reported bug whereby a message would remain on the queue
6590 longer than the retry timeout, but would be bounced if a delivery was
6591 forced. I say "may" because I never totally pinned down the problem;
6592 setting up timeout/retry tests is difficult. See also the next item.
6594 79. The ultimate address timeout was not being applied to errors that involved
6595 a combination of host plus message (for example, a timeout on a MAIL
6596 command). When an address resolved to a number of possible hosts, and they
6597 were not all tried for each delivery (e.g. because of hosts_max_try), a
6598 message could remain on the queue longer than the retry timeout.
6600 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6603 81. Fixed an obscure SMTP outgoing bug which required at least the following
6604 conditions: (a) there was another message waiting for the same server;
6605 (b) the server returned 5xx to all RCPT commands in the first message so
6606 that the message was not completed; (c) the server dropped the connection
6607 or gave a negative response to the RSET that Exim sends to abort the
6608 transaction. The observed case was a dropped connection after DATA that had
6609 been sent in pipelining mode. That is, the server had advertised PIPELINING
6610 but was not implementing it correctly. The effect of the bug was incorrect
6611 behaviour, such as trying another host, and this could lead to a crash.
6617 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6618 and daemon.c were passed as pointers to ints; they should have been
6619 pointers to socklen_t variables (which are typically unsigned ints).
6621 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6624 3. Fixed a really odd bug that affected only the testing scheme; patching a
6625 certain fixed string in the binary changed the value of another string that
6626 happened to be identical to the end of the original first string.
6628 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6629 name", it returns that address as the IP address. On some operating
6630 systems (e.g. Solaris), it also passes back the IP address string as the
6631 "host name". However, on others (e.g. Linux), it passes back an empty
6632 string. Exim wasn't checking for this, and was changing the host name to an
6633 empty string, assuming it had been canonicalized.
6635 5. Although rare, it is permitted to have more than one PTR record for a given
6636 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6637 all the names associated with an address, because they do in Solaris.
6638 However, it seems that they do not in Linux for data that comes from the
6639 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6640 I found this out when I moved to a new Linux workstation and tried to run
6641 the Exim test suite.
6643 To get round this problem I have changed the code so that it now does its
6644 own call to the DNS to look up PTR records when searching for a host name.
6645 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6646 addresses that are only in /etc/hosts are still found.
6648 This behaviour is, however, controlled by an option called host_lookup_
6649 order, which defaults to "bydns:byaddr". If people want to use the other
6650 order, or indeed, just use one or the other means of lookup, they can
6651 specify it in this variable.
6653 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6654 some operating systems, this comes back from gethostbyaddr() as an empty
6655 string, and this is what Exim used to test for. However, it seems that in
6656 other systems, "." is yielded. Exim now tests for this case too.
6658 7. The values of check_spool_space and check_log_space are now held internally
6659 as a number of kilobytes instead of an absolute number of bytes. If a
6660 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6661 kilobyte. This means that much larger values can be stored.
6663 8. Exim monitor: an attempt to get the action menu when not actually pointing
6664 at a message produces an empty menu entitled "No message selected". This
6665 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6666 no entries in it ("Shell widget menu has zero width and/or height"). So I
6667 have added a single, blank menu entry in this case.
6669 9. Added ${quote_local_part.
6671 10. MIME decoding is now applied to the contents of Subject: header lines when
6674 11. Now that a reference to $sender_host_address automatically causes a reverse
6675 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6676 host lookup before query-style lookups in lists that might use this
6677 variable. This has therefore been abolished, and the "net-" prefix is no
6678 longer necessary for query-style lookups.
6680 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6681 have been a typo for LFLAGS, so it has been changed.
6683 13. The install script calls Exim with "-C /dev/null" in order to find the
6684 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6685 to be output. However, since Exim outputs its version number before the
6686 error, it didn't break the script. It just looked ugly. I fixed this by
6687 always allowing "-C /dev/null" if the caller is root.
6689 14. Ignore overlarge ACL variable number when reading spool file - insurance
6690 against a later release with more variables having written the file.
6692 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6693 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6694 this, as well as the form without the "IPv6" on the front.
6696 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6697 OS/Makefile-Darwin file.
6699 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6700 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6702 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6703 variable. However, some OS use ints for inodes. I've added cast to long int
6704 to get rid of the compiler warning.
6706 19. I had forgotten to lock out "/../" in configuration file names when
6707 ALT_CONFIG_PREFIX was set.
6709 20. Routers used for verification do not need to specify transports. However,
6710 if such a router generated a host list, and callout was configured, Exim
6711 crashed, because it could not find a port number from the (non-existent)
6712 transport. It now assumes port 25 in this circumstance.
6714 21. Added the -t option to exigrep.
6716 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6717 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6718 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6719 them is set, LOOKUP_LSEARCH is forced.
6721 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6723 24. Added sender and host information to the "rejected by local_scan()" log
6724 line; previously there was no indication of these.
6726 25. Added .include_if_exists.
6728 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6729 Linux. It turns out that not all file systems support this. Apparently some
6730 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6731 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6732 not supported on the file descriptor, is now ignored when Exim is trying to
6733 sync a directory. This applies only to Linux.
6735 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6737 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6738 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6741 29. Added support for maildirsize files from supplied patch (modified a bit).
6743 30. The use of :fail: followed by an empty string could lead Exim to respond to
6744 sender verification failures with (e.g.):
6746 550 Verification failed for <xxx>
6747 550 Sender verify failed
6749 where the first response line was missing the '-' that indicates it is not
6750 the final line of the response.
6752 31. The loop for finding the name of the user that called Exim had a hardwired
6753 limit of 10; it now uses the value of finduser_retries, which is used for
6754 all other user lookups.
6756 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6759 33. Exim was neglecting to zero errno before one call of strtol() when
6760 expanding a string and expecting an integer value. On some systems this
6761 resulted in spurious "integer overflow" errors. Also, it was casting the
6762 result into an int without checking.
6764 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6765 did not work. The code looks as if it has *never* worked, though it appears
6766 to have been documented since at least release 1.62. I have made it work.
6768 35. The "timeout_DNS" error in retry rules, also documented since at least
6769 1.62, also never worked. As it isn't clear exactly what this means, and
6770 clearly it isn't a major issue, I have abolished the feature by treating it
6771 as "timeout", and writing a warning to the main and panic logs.
6773 36. The display of retry rules for -brt wasn't always showing the error code
6776 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6777 timeout_connect_A, timeout_connect_MX.
6779 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6780 to the empty sender.
6782 39. The daemon was not analysing the content of -oX till after it had closed
6783 stderr and disconnected from the controlling terminal. This meant that any
6784 syntax errors were only noted on the panic log, and the return code from
6785 the command was 0. By re-arranging the code a little, I've made the
6786 decoding happen first, so such errors now appear on stderr, and the return
6787 code is 1. However, the actual setting up of the sockets still happens in
6788 the disconnected process, so errors there are still only recorded on the
6791 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6792 connections (as happens on some IP stacks) was logged at start up time as
6793 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6794 it from "IPv6 and IPv4", which means that two separate sockets are being
6797 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6798 says whether AF_INET or AF_INET6 was passed as an argument.
6800 42. Exiwhat output was messed up when time zones were included in log
6803 43. Exiwhat now gives more information about the daemon's listening ports,
6804 and whether -tls-on-connect was used.
6806 44. The "port" option of the smtp transport is now expanded.
6808 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6809 silently ignored. Now an error message is written to the main and panic
6812 46. There's a new ACL modifier called "logwrite" which writes to a log file
6813 as soon as it is encountered.
6815 47. Added $local_user_uid and $local_user_gid at routing time.
6817 48. Exim crashed when trying to verify a sender address that was being
6820 49. Exim was recognizing only a space character after ".include". It now also
6821 recognizes a tab character.
6823 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6824 extracting the relevant information from the specification. The man page no
6825 longer contains scrambled data for the -d option, and I've added a section
6826 at the front about calling Exim under different names.
6828 51. Added "extra_headers" argument to the "mail" command in filter files.
6830 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6833 53. Installed eximstats 1.29.
6835 54. Added transport_filter_timeout as a generic transport option.
6837 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6838 Cc: header lines. This was required by RFC 822, but it not required by RFC
6841 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6842 incoming messages that did not have them. Now it does so only if the
6843 message originates locally, that is, if there is no associated remote host
6844 address. When Resent- header lines are present, this applies to the Resent-
6845 lines rather than the non-Resent- lines.
6847 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6848 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6850 58. Messages for configuration errors now include the name of the main
6851 configuration file - useful now that there may be more than one file in a
6852 list (.included file names were always shown).
6854 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6855 for those rare installations that do not start the daemon as root or run it
6856 setuid root. I've cut out the call to initgroups() if the daemon is not
6859 60. The Exim user and group can now be bound into the binary as text strings
6860 that are looked up at the start of Exim's processing.
6862 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6864 62. Added $mailstore_basename variable.
6866 63. Installed patch to sieve.c from Michael Haardt.
6868 64. When Exim failed to open the panic log after failing to open the main log,
6869 the original message it was trying to log was written to stderr and debug
6870 output, but if they were not available (the usual case in production), it
6871 was lost. Now it is written to syslog before the two lines that record the
6872 failures to open the logs.
6874 65. Users' Exim filters run in subprocesses under the user's uid. It is
6875 possible for a "deliver" command or an alias in a "personal" command to
6876 provoke an address rewrite. If logging of address rewriting is configured,
6877 this fails because the process is not running as root or exim. There may be
6878 a better way of dealing with this, but for the moment (because 4.30 needs
6879 to be released), I have disabled address rewrite logging when running a
6880 filter in a non-root, non-exim process.
6886 1. The buildconfig auxiliary program wasn't quoting the value set for
6887 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6888 not defined. This bug was masked in 4.22 by the effect that was fixed in
6891 2. Some messages that were rejected after a message id was allocated were
6892 shown as "incomplete" by exigrep. It no longer does this for messages that
6893 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6895 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6896 have allow_domain_literals set, the ID did not get logged in the <= line.
6897 Domain literals are now always recognized in Message-ID: header lines.
6899 4. The first argument for a ${extract expansion item is the key name or field
6900 number. Leading and trailing spaces in this item were not being ignored,
6901 causing some misleading effects.
6903 5. When deliver_drop_privilege was set, single queue runner processes started
6904 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6905 same command in the process it spins off) were not dropping privilege.
6907 6. When the daemon running as "exim" started a queue runner, it always
6908 re-executed Exim in the spun-off process. This is a waste of effort when
6909 deliver_drop_privilege is set. The new process now just calls the
6910 queue-runner function directly.
6916 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6919 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6920 ignored. Though the use of -oP was forcing the writing of a pid file, it
6921 was always written to the default place.
6923 3. If the message "no IP address found for host xxxx" is generated during
6924 incoming verification, it is now followed by identification of the incoming
6925 connection (so you can more easily find what provoked it).
6927 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6929 5. Added some features to "harden" Exim a bit more against certain attacks:
6931 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6932 be put in Local/Makefile. This is like the never_users runtime option,
6933 but it cannot be overridden. The default setting is "root".
6935 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6936 prefix string with which any file named in a -C command line option
6939 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6940 is retained for -C and -D only if the caller of Exim is root. Without
6941 it, the exim user may also use -C and -D and retain privilege.
6943 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6944 command line option is disabled.
6946 6. Macro names set by the -D option must start with an upper case letter, just
6947 like macro names defined in the configuration file.
6949 7. Added "dereference=" facility to LDAP.
6951 8. Two instances of the typo "uknown" in the source files are fixed.
6953 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6954 the Configure-Makefile script screwed up while processing it.
6956 10. Incorporated PCRE 4.4.
6958 11. The SMTP synchronization check was not operating right at the start of an
6959 SMTP session. For example, it could not catch a HELO sent before the client
6960 waited for the greeting. There is now a check for outstanding input at the
6961 point when the greeting is written. Because of the duplex, asynchronous
6962 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6963 way, but not yet received, when the check is performed.
6965 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6966 on TCP/IP sockets, because this apparently causes some broken clients to
6969 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6970 unchanged) from the Cygwin maintainer.
6972 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6973 was supported instead of testing for mbx. Effectively a typo.
6975 15. The spa authenticator server code was not checking that the input it
6976 received was valid base64.
6978 16. The debug output line for the "set" modifier in ACLs was not showing the
6979 name of the variable that was being set.
6981 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6983 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6984 lookup on its own. Something else was needed to trigger the lookup. For
6985 example, a match in host_lookup or the need for a host name in a host list.
6986 Now, if $sender_host_name is referenced and the host name has not yet been
6987 looked up, a lookup is performed. If the lookup fails, the variable remains
6988 empty, and $host_lookup_failed is set to "1".
6990 19. Added "eqi" as a case-independent comparison operator.
6992 20. The saslauthd authentication condition could segfault if neither service
6993 nor realm was specified.
6995 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6996 error message that was logged was misleading, and incoming SMTP
6997 connections were dropped. The message is now more accurate, and temporary
6998 errors are given to SMTP connections.
7000 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7001 (except RSET) with a 421 error, until QUIT is received. However, it was
7002 failing to send a response to QUIT.
7004 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7005 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7006 ACL is now run after the helo_try_verify_hosts code.
7008 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7009 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7010 case-independent, so other case variants are also recognized). Apparently
7011 some systems use these upper case variants.
7013 25. If more than two messages were waiting for the same host, and a transport
7014 filter was specified for the transport, Exim sent two messages over the
7015 same TCP/IP connection, and then failed with "socket operation on non-
7016 socket" when it tried to send the third.
7018 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7020 27. The extern definition of crypt16() in expand.c was not being excluded when
7021 the OS had its own crypt16() function.
7023 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7024 as a preferred synonym for return_size_limit, both as an option and as an
7027 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7029 30. Changed the default configuration ACL to relax the local part checking rule
7030 for addresses that are not in any local domains. For these addresses,
7031 slashes and pipe symbols are allowed within local parts, but the sequence
7032 /../ is explicitly forbidden.
7034 31. SPA server authentication was not clearing the challenge buffer before
7037 32. log_message in a "warn" ACL statement was writing to the reject log as
7038 well as to the main log, which contradicts the documentation and doesn't
7039 seem right (because no rejection is happening). So I have stopped it.
7041 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7042 However, I am unable to do any testing of this.
7044 34. Fixed an infelicity in the appendfile transport. When checking directories
7045 for a mailbox, to see if any needed to be created, it was accidentally
7046 using path names with one or more superfluous leading slashes; tracing
7047 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7049 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7050 contents are added to the log line that is written for every discarded
7051 recipient. (Previously a log_message setting was ignored.)
7053 36. The ${quote: operator now quotes the string if it is empty.
7055 37. The install script runs exim in order to find its version number. If for
7056 some reason other than non-existence or emptiness, which it checks, it
7057 could not run './exim', it was installing it with an empty version number,
7058 i.e. as "exim-". This error state is now caught, and the installation is
7061 38. An argument was missing from the function that creates an error message
7062 when Exim fails to connect to the socket for saslauthd authentication.
7063 This could cause Exim to crash, or give a corrupted message.
7065 39. Added isip, isip4, and isip6 to ${if conditions.
7067 40. The ACL variables $acl_xx are now saved with the message, and can be
7068 accessed later in routers, transports, and filters.
7070 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7071 strings in the file are not string-expanded.
7073 42. If a MAIL command specified a SIZE value that was too large to fit into an
7074 int variable, the check against message_size_limit failed. Such values are
7075 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7076 day this will have to be increased, but I don't think I want to be around
7077 when emails are that large.
7084 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7085 iconv() is not standard in FreeBSD.
7087 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7088 IPv6 enabled. The observed symptom was a segmentation fault on return from
7089 the function os_common_find_running_interfaces() in src/os.c.
7091 3. In the check_special_case() function in daemon.c I had used "errno" as an
7092 argument name, which causes warnings on some systems. This was basically a
7093 typo, since it was named "eno" in the comments!
7095 4. The code that waits for the clock to tick (at a resolution of some fraction
7096 of a second) so as to ensure message-id uniqueness was always waiting for
7097 at least one whole tick, when it could have waited for less. [This is
7098 almost certainly not relevant at current processor speeds, where it is
7099 unlikely to ever wait at all. But we try to future-proof.]
7101 5. The function that sleeps for a time interval that includes fractions of a
7102 second contained a race. It did not block SIGALRM between setting the
7103 timer, and suspending (a couple of lines later). If the interval was short
7104 and the sigsuspend() was delayed until after it had expired, the suspension
7105 never ended. On busy systems this could lead to processes getting stuck for
7108 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7109 process, before it forks any delivery processes. The open lookup caching
7110 mechanism meant that the open file or database connection was passed into
7111 the delivery process. The problem was that delivery processes always tidy
7112 up cached lookup data. This could cause a problem for the next delivery
7113 process started by the queue runner, because the external queue runner
7114 process does not know about the closure. So the next delivery process
7115 still has data in the lookup cache. In the case of a file lookup, there was
7116 no problem because closing a file descriptor in a subprocess doesn't affect
7117 the parent. However, if the lookup was caching a connection to a database,
7118 the connection was closed, and the second delivery process was likely to
7119 see errors such as "PGSQL: query failed: server closed the connection
7120 unexpectedly". The problem has been fixed by closing all cached lookups
7121 in a queue runner before running a delivery process.
7123 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7124 seem to have the "const" qualifier which it has on other OS. I've
7127 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7128 *of the same type* (client or server) with the same public name that an
7129 error should be diagnosed.
7131 9. When Exim looked up a host name for an IP address, but failed to find the
7132 original IP address when looking up the host name (a safety check), it
7133 output the message "<ip address> does not match any IP for NULL", which was
7134 confusing, to say the least. The bug was that the host name should have
7135 appeared instead of "NULL".
7137 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7138 user that is built into the binary, and the -C or -D options is used, root
7139 privilege is dropped before the configuration file is read. In addition,
7140 logging is switched to stderr instead of the normal log files. If the
7141 configuration then re-defines the Exim user, the unprivileged environment
7142 is probably not what is expected, so Exim logs a panic warning message (but
7145 However, if deliver_drop_privilege is set, the unprivileged state may well
7146 be exactly what is intended, so the warning has been cut out in that case,
7147 and Exim is allowed to try to write to its normal log files.
7153 1. smtp_return_error_details was not giving details for temporary sender
7154 or receiver verification errors.
7156 2. Diagnose a configuration error if two authenticators have the same public
7159 3. Exim used not to create the message log file for a message until the first
7160 delivery attempt. This could be confusing when incoming messages were held
7161 for policy or load reasons. The message log file is now created at the time
7162 the message is received, and an initial "Received" line is written to it.
7164 4. The automatically generated man page for command line options had a minor
7165 bug that caused no ill effects; however, a more serious problem was that
7166 the procedure for building the man page automatically didn't always
7167 operate. Consequently, release 4.20 contains an out-of-date version. This
7168 shouldn't happen again.
7170 5. When building Exim with embedded Perl support, the script that builds the
7171 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7172 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7174 6. The freeze_tell option was not being used for messages that were frozen on
7175 arrival, either by an ACL or by local_scan().
7177 7. Added the smtp_incomplete_transaction log selector.
7179 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7180 was accepting AUTH without a new EHLO.
7182 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7183 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7186 10. Logging of TCP/IP connections (when configured) now happens in the main
7187 daemon process instead of the child process, so that the TCP/IP connection
7188 count is more accurate (but it can never be perfect).
7190 11. The use of "drop" in a nested ACL was not being handled correctly in the
7191 outer ACL. Now, if condition failure induced by the nested "drop" causes
7192 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7193 or "require"), the connection is dropped.
7195 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7196 that yield "discard" can now be used with an "accept" or a "discard" verb,
7197 but an error is generated for any others (because I can't see a useful way
7198 to define what should happen).
7200 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7201 are now processed in the same way as lines in the Exim configuration file.
7202 In particular, continuation lines are supported.
7204 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7206 15. Added -ti meaning -t -i.
7208 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7209 domains, and warn by logging if others are found.
7211 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7212 returned from GIFCONF when Exim is trying to find the list of interfaces on
7213 a host. The code in os.c has been modified to copy each ifreq to an aligned
7214 structure in all cases.
7216 Also, in some cases, the returned ifreq's were being copied to a 'struct
7217 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7218 means the last couple of bytes of an IPv6 address could be chopped if the
7219 ifreq contained only a normal sockaddr (14 bytes storage).
7221 18. Named domain lists were not supported in the hosts_treat_as_local option.
7222 An entry such as +xxxx was not recognized, and was treated as a literal
7225 19. Ensure that header lines added by a DATA ACL are included in the reject log
7226 if the ACL subsequently rejects the message.
7228 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7229 MD5 (which is deprecated).
7231 21. When testing a filter file using -bf, Exim was writing a message when it
7232 took the sender from a "From " line in the message, but it was not doing so
7233 when it took $return_path from a Return-Path: header line. It now does.
7235 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7236 with a valid header line field name (a series of printing characters
7237 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7239 23. Changed "disc" in the source to "disk" to conform to the documentation and
7240 the book and for uniformity.
7242 24. Ignore Sendmail's -Ooption=value command line item.
7244 25. When execve() failed while trying to run a command in a pipe transport,
7245 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7246 could be confused with a return value of 69 from the command itself. This
7247 has been changed to 127, the value the shell returns if it is asked to run
7248 a non-existent command. The wording for the related log line suggests a
7249 non-existent command as the problem.
7251 26. If received_header_text expands to an empty string, do not add a Received:
7252 header line to the message. (Well, it adds a token one on the spool, but
7253 marks it "old" so that it doesn't get used or transmitted.)
7255 27. Installed eximstats 1.28 (addition of -nt option).
7257 28. There was no check for failure on the call to getsockname() in the daemon
7258 code. This can fail if there is a shortage of resources on the system, with
7259 ENOMEM, for example. A temporary error is now given on failure.
7261 29. Contrary to the C standard, it seems that in some environments, the
7262 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7263 program. Exim now does this explicitly; it affects the formatting of
7264 timestamps using strftime().
7266 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7267 complaints. I've now initialized all the variables, to avoid this.
7269 32. Header lines added by a system filter were not being "seen" during
7270 transport-time rewrites.
7272 33. The info_callback() function passed to OpenSSL is set up with type void
7273 (*)(SSL *, int, int), as described somewhere. However, when calling the
7274 function (actually a macro) that sets it up, the type void(*)() is
7275 expected. I've put in a cast to prevent warnings from picky compilers.
7277 34. If a DNS black list lookup found a CNAME record, but there were no A
7278 records associated with the domain it pointed at, Exim crashed.
7280 35. If a DNS black list lookup returned more than one A record, Exim ignored
7281 all but the first. It now scans all returned addresses if a particular IP
7282 value is being sought. In this situation, the contents of the
7283 $dnslist_value variable are a list of all the addresses, separated by a
7286 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7287 to accept a host name and all its aliases if the forward lookup for any of
7288 them yielded the IP address of the incoming connection. Now it accepts only
7289 those names whose forward lookup yields the correct IP address. Any other
7290 names are discarded. This closes a loophole whereby a rogue DNS
7291 administrator could create reverse DNS records to break through a
7292 wildcarded host restriction in an ACL.
7294 37. If a user filter or a system filter that ran in a subprocess used any of
7295 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7296 the wrong values were passed to the pipe command ($thisaddress had the
7297 value of $0, $0 had the value of $1, etc). This bug was introduced by
7298 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7300 38. Improved the line breaking for long SMTP error messages from ACLs.
7301 Previously, if there was no break point between 40 and 75 characters, Exim
7302 left the rest of the message alone. Two changes have been made: (a) I've
7303 reduced the minimum length to 35 characters; (b) if it can't find a break
7304 point between 35 and 75 characters, it looks ahead and uses the first one
7305 that it finds. This may give the occasional overlong line, but at least the
7306 remaining text gets split now.
7308 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7309 file descriptors might be low, and that setting 1000 would always raise it.
7310 It turns out that in some environments, the limit is already over 1000 and
7311 that lowering it causes trouble. So now Exim takes care not to decrease it.
7313 40. When delivering a message, the value of $return_path is set to $sender_
7314 address at the start of routing (routers may change the value). By an
7315 oversight, this default was not being set up when an address was tested by
7316 -bt or -bv, which affected the outcome if any router or filter referred to
7319 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7320 log only when "message" or "log_message" are set. However, if one of the
7321 conditions was an address verification, or a call to a nested ACL, the
7322 messages generated by the underlying test were being passed through. This
7323 no longer happens. The underlying message is available in $acl_verify_
7324 message for both "message" and "log_message" expansions, so it can be
7325 passed through if needed.
7327 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7328 new expansion $bh_ to give the encoded byte string without charset
7329 translation. Translation happens only if iconv() is available; HAVE_ICONV
7330 indicates this at build time. HEADERS_CHARSET gives the charset to
7331 translate to; headers_charset can change it in the configuration, and
7332 "headers charset" can change it in an individual filter file.
7334 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7335 that creates RFC 2047 encoded "words" labels them as that charset instead
7336 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7337 expansion operator; (ii) when Exim creates a From: line for a local
7338 message; (iii) when a header line is rewritten to include a "phrase" part.
7340 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7341 buggy, causing it to skip the first lines of messages whose message ID
7342 ended in 'D'. This would not have bitten before Exim release 4.14, because
7343 message IDs were unlikely to end in 'D' before then. The effect was to have
7344 incorrect size information for certain domains.
7346 45. #include "config.h" was missing at the start of the crypt16.c module. This
7347 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7350 46. If there was a timeout during a "random" callout check, Exim treated it as
7351 a failure of the random address, and carried on sending RSET and the real
7352 address. If the delay was just some slowness somewhere, the response to the
7353 original RCPT would be taken as a response to RSET and so on, causing
7354 mayhem of various kinds.
7356 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7357 when I implemented it. It didn't allow for the fact that some option values
7358 may legitimately be negative (e.g. size_addition), and it didn't even do
7359 the right test for positive values.
7361 48. Domain names in DNS records are case-independent. Exim always looks them up
7362 in lower case. Some resolvers return domain names in exactly the case they
7363 appear in the zone file, that is, they may contain uppercase letters. Not
7364 all resolvers do this - some return always lower case. Exim was treating a
7365 change of case by a resolver as a change of domain, similar to a widening
7366 of a domain abbreviation. This triggered its re-routing code and so it was
7367 trying to route what was effectively the same domain again. This normally
7368 caused routing to fail (because the router wouldn't handle the domain
7369 twice). Now Exim checks for this case specially, and just changes the
7370 casing of the domain that it ultimately uses when it transmits the message
7373 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7376 50. If a filter generated a file delivery with a non-absolute name (possible if
7377 no home directory exists for the router), the forbid_file option was not
7380 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7381 the existing equality matching.
7383 52. Exim was using ints instead of ino_t variables in some places where it was
7384 dealing with inode numbers.
7386 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7387 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7388 called TMPDIR, and if it finds it is different, it changes its value.
7390 54. The smtp_printf() function is now made available to local_scan() so
7391 additional output lines can be written before returning. There is also an
7392 smtp_fflush() function to enable the detection of a dropped connection.
7393 The variables smtp_input and smtp_batched_input are exported to
7396 55. Changed the default runtime configuration: the message "Unknown user"
7397 has been removed from the ACL, and instead placed on the localuser router,
7398 using the cannot_route_message feature. This means that any verification
7399 failures that generate their own messages won't get overridden. Similarly,
7400 the "Unrouteable address" message that was in the ACL for unverifiable
7401 relay addresses has also been removed.
7403 56. Added hosts_avoid_esmtp to the smtp transport.
7405 57. The exicyclog script was not checking for the esoteric option
7406 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7407 will work only if exicyclog is run under the appropriate euid.
7409 58. Following a discussion on the list, the rules by which Exim recognises line
7410 endings on incoming messages have been changed. The -dropcr and drop_cr
7411 options are now no-ops, retained only for backwards compatibility. The
7412 following line terminators are recognized: LF CRLF CR. However, special
7413 processing applies to CR:
7415 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7416 nor a local message in the state where . is a terminator.
7418 (ii) If a bare CR is encountered in a header line, an extra space is added
7419 after the line terminator so as not to end the header. The reasoning
7420 behind this is that bare CRs in header lines are most likely either
7421 to be mistakes, or people trying to play silly games.
7423 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7424 was being incorrectly given as 18 bytes larger than it should have been.
7425 This is a VOB (very old bug).
7427 60. This may never have affected anything current, but just in case it has:
7428 When the local host is found other than at the start of a list of hosts,
7429 the local host, those with the same MX, and any that follow, are discarded.
7430 When the list in question was part of a longer list of hosts, the following
7431 hosts (not currently being processed) were also being discarded. This no
7432 longer happens. I'm not sure if this situation could ever has previously
7435 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7438 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7439 header. This is something that is recommended in a new Internet Draft, and
7440 is something that is documented as being done by Sendmail. There are two
7441 possible values. For messages generated by the autoreply transport, Exim
7444 Auto-Submitted: auto-replied
7446 whereas for all other generated messages (e.g. bounces) it adds
7448 Auto-Submitted: auto-generated
7450 63. The "personal" condition in filters now includes a test for the
7451 Auto-Submitted: header. If it contains the string "auto-" the message it
7452 not considered personal.
7454 64. Added rcpt_include_affixes as a generic transport option.
7456 65. Added queue_only_override (default true).
7458 66. Added the syslog_duplication option.
7460 67. If what should have been the first header line of a message consisted of
7461 a space followed by a colon, Exim was mis-interpreting it as a header line.
7462 It isn't of course - it is syntactically invalid and should therefore be
7463 treated as the start of the message body. The misbehaviour could have
7464 caused a number of strange effects, including loss of data in subsequent
7465 header lines, and spool format errors.
7467 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7468 client host had authenticated. This control can now be exercised by an ACL
7469 for more flexibility.
7471 69. By default, callouts do not happen when testing with -bh. There is now a
7472 variant, -bhc, which does actually run the callout code, including
7473 consulting and updating the callout cache.
7475 70. Added support for saslauthd authentication, courtesy of Alexander
7478 71. If statvfs() failed on the spool or log directories while checking their
7479 size for availability, Exim confusingly gave the error "space shortage".
7480 Furthermore, in debugging mode it crashed with a floating point exception.
7481 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7482 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7483 serious problem, Exim now writes to the main and panic logs when this
7484 happens, with details of the failure. It then refuses to accept the
7485 incoming message, giving the message "spool directory problem" or "log
7486 directory problem" with a 421 code for SMTP messages.
7488 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7489 0, 1, and 2 exist, because some OS complain for execs without them (see
7490 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7491 descriptors. However, the code omitted to check that the open succeeded,
7492 causing mysterious errors if for some reason the permissions on /dev/null
7493 got screwed. Now Exim writes a message to the main and panic logs, and
7494 bombs out if it can't open /dev/null.
7496 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7497 interact so that it is all more flexible. It is supposed to remain
7498 backwards compatible. Also added extra_local_interfaces.
7500 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7501 to bomb out with an assertion failure - to the client this appears as a
7502 connection drop. This problem occurs in the part of the code that was taken
7503 from the Samba project. Fortunately, the assertion is in a very simple
7504 function, so I have fixed this by reproducing the function inline in the
7505 one place where it is called, and arranging for authentication to fail
7506 instead of killing the process with assert().
7508 75. The SPA client code was not working when the server requested OEM rather
7509 than Unicode encoding.
7511 76. Added code to make require_files with a specific uid setting more usable in
7512 the case where statting the file as root fails - usually a non-root-mounted
7513 NFS file system. When this happens and the failure is EACCES, Exim now
7514 forks a subprocess and does the per-uid checking as the relevant uid.
7516 77. Added process_log_path.
7518 78. If log_file_path was not explicitly set, a setting of check_log_space or
7519 check_log_inodes was ignored.
7521 79. If a space check for the spool or log partitions fails, the incident is now
7522 logged. Of course, in the latter case the data may get lost...
7524 80. Added the %p formatting code to string_format() so that it can be used to
7525 print addresses in debug_print(). Adjusted all the address printing in the
7526 debugging in store.c to use %p rather than %d.
7528 81. There was a concern that a line of code in smtp_in.c could overflow a
7529 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7530 initially expressed, the concern was not well-founded, because trailing
7531 spaces are removed early. However, if the trailing spaces were followed by
7532 a NULL, they did not get removed, so the overflow was possible. Two fixes
7535 (a) I re-wrote the offending code in a cleaner fashion.
7536 (b) If an incoming SMTP command contains a NULL character, it is rejected
7539 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7540 runs initgroups(), so that if the Exim user is in any additional groups,
7541 they will be used during message reception.
7547 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob