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
188 JH/01 Remove code calling the customisable local_scan function, unless a new
189 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
191 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
192 non-signal-safe functions being used.
194 JH/03 Bug 2269: When presented with a received message having a stupidly large
195 number of DKIM-Signature headers, disable DKIM verification to avoid
196 a resource-consumption attack. The limit is set at twenty.
198 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
199 report of oldest_pass in ${authres } in consequence, and separate out
200 some descriptions of reasons for verification fail.
202 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
203 files in the spool were present and unlocked. A queue-runner could spot
204 them, resulting in a duplicate delivery. Fix that by doing the unlock
205 after the unlink. Investigation by Tim Stewart. Take the opportunity to
206 add more error-checking on spoolfile handling while that code is being
209 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
210 No known attacks, no CVE, this is defensive hardening.
212 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
213 a queue-runner could start a delivery while other operations were ongoing.
214 Cutthrough delivery was a common victim, resulting in duplicate delivery.
215 Found and investigated by Tim Stewart. Fix by using the open message data
216 file handle rather than opening another, and not locally closing it (which
217 releases a lock) for that case, while creating the temporary .eml format
218 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
220 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
221 $sender_verify_failure/$recipient_verify_failure to "random".
223 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
226 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
227 Previously this would segfault.
229 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
232 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
233 like zero, since the resolver should be doing this for us, But we need one
234 as a CNAME but no MX presence gets the CNAME returned; we need to check
235 that doesn't point to an MX to declare it "no MX returned" rather than
236 "error, loop". A new main option is added so the older capability of
237 following some limited number of chain links is maintained.
239 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
241 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
242 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
243 now-deprecated earlier definitions used only specified the range up to TLS
244 1.2 (in the older-version library docs).
246 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
248 JH/15 Rework TLS client-side context management. Stop using a global, and
249 explicitly pass a context around. This enables future use of TLS for
250 connections to service-daemons (eg. malware scanning) while a client smtp
251 connection is using TLS; with cutthrough connections this is quite likely.
253 JH/16 Fix ARC verification to do AS checks in reverse order.
255 JH/17 Support a "tls" option on the ${readsocket } expansion item.
257 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
258 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
259 Previously the "utf8" would be re-prepended for every additional message.
261 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
262 Previously thery were accepted, resulting in issues when attempting to
263 forward messages to a non-supporting MTA.
265 PP/02 Let -n work with printing macros too, not just options.
267 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
268 one parent address was copied, and bogus data was used at delivery-logging
269 time. Either a crash (after delivery) or bogus log data could result.
270 Discovery and analysis by Tim Stewart.
272 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
273 Previously if the string ended mid-character, we did not insert the
274 promised '?' replacement.
276 PP/04 Documentation: current string operators work on bytes, not codepoints.
278 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
279 should pack well giving a smaller memory footprint so better caching and
280 therefore performance. Group the declarations where this can't be done so
281 that the byte-sized flag variables are not interspersed among pointer
282 variables, giving a better chance of good packing by the compiler.
284 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
285 non-null, to avoid issues with sites running BATV. Previously reports were
286 sent with an empty envelope sender so looked like bounces.
288 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
289 The ignore_error flag wasn't being returned from the filter subprocess so
290 was not set for later routers. Investigation and fix by Matthias Kurz.
292 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
293 and a msg:complete for the whole, when a message is manually removed using
294 -Mrm. Developement by Matthias Kurz, hacked on by JH.
296 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
297 a "Gnu special" function, asprintf() in the DB utility binary builds; I
298 hope that is portable enough.
300 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
301 requiring a known-CA anchor certificate; make it now rely entirely on the
302 TLSA as an anchor. Checking the name on the leaf cert against the name
303 on the A-record for the host is still done for TA (but not for EE mode).
305 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
306 included in delivery lines for non-proxied connections, when compiled with
307 SUPPORT_SOCKS and running with proxy logging enabled.
309 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
310 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
311 move the existing event to fire before the normal logging of message
312 failure so that custom logging is bracketed by normal logging.
314 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
315 msg:fail:internal event. Developement by Matthias Kurz.
317 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
318 far too small for todays use of crypto signatures stored there. Go all
319 the way to the max DNS message size of 64kB, even though this might be
320 overmuch for IOT constrained device use.
322 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
323 copy a string over itself. The library routine is documented as not
324 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
326 JH/32 For main options check_spool_space and check_inode_space, where the
327 platform supports 64b integers, support more than the previous 2^31 kB
328 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
329 the previous G, M, k.
331 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
332 $authenticated_fail_id variable on authentication failure. Previously
335 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
336 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
337 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
338 GNUTLS_SEC_PARAM_MEDIUM.
340 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
341 side. Previously we would continue as if no SNI had been received.
343 JH/36 Harden the handling of string-lists. When a list consisted of a sole
344 "<" character, which should be a list-separator specification, we walked
345 off past the nul-terimation.
347 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
348 causes) even when the retry time is not yet met. Previously they were
349 not, meaning that when (say) an account was over-quota and temp-rejecting,
350 and multiple senders' messages were queued, only one sender would get
351 notified on each configured delay_warning cycle.
353 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
355 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
356 cipher-suites, an error can be left on the stack even for a succeeding
357 accept; this results in impossible error messages when a later operation
360 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
361 return error codes indicating retry. Under TLS1.3 this becomes required.
363 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
364 it only wrote the new authenticators, resulting in a lack of tracking of
365 peer changes of ESMTP extensions until the next cache flush.
367 JH/41 Fix the loop reading a message header line to check for integer overflow,
368 and more-often against header_maxsize. Previously a crafted message could
369 induce a crash of the recive process; now the message is cleanly rejected.
371 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
372 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
378 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
379 When redis_servers is set to a list of > 1 element, and the Redis servers
380 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
381 case of MOVED into a DEFER case instead, thus moving the query onto the
382 next server in the list. For a cluster of N elements, all N servers must
383 be defined in redis_servers.
385 GF/02 Catch and remove uninitialized value warning in exiqsumm
386 Check for existence of @ARGV before looking at $ARGV[0]
388 JH/01 Replace the store_release() internal interface with store_newblock(),
389 which internalises the check required to safely use the old one, plus
390 the allocate and data copy operations duplicated in both (!) of the
391 extant use locations.
393 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
394 modifier. This matches the restriction on the commandline.
396 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
397 Previously only the last row was returned.
399 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
400 we assumed that tags in the header were well-formed, and parsed the
401 element content after inspecting only the first char of the tag.
402 Assumptions at that stage could crash the receive process on malformed
405 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
406 While running the DKIM ACL we operate on the Permanent memory pool so that
407 variables created with "set" persist to the DATA ACL. Also (at any time)
408 DNS lookups that fail create cache records using the Permanent pool. But
409 expansions release any allocations made on the current pool - so a dnsdb
410 lookup expansion done in the DKIM ACL releases the memory used for the
411 DNS negative-cache, and bad things result. Solution is to switch to the
412 Main pool for expansions.
413 While we're in that code, add checks on the DNS cache during store_reset,
414 active in the testsuite.
415 Problem spotted, and debugging aided, by Wolfgang Breyha.
417 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
418 When none of the hosts presented to a transport match an already-open
419 connection, close it and proceed with the list. Previously we would
420 queue the message. Spotted by Lena with Yahoo, probably involving
423 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
424 Previously a spurious "250 OK id=" response was appended to the proper
427 JH/08 The "support for" informational output now, which built with Content
428 Scanning support, has a line for the malware scanner interfaces compiled
429 in. Interface can be individually included or not at build time.
431 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
432 by the template makefile "src/EDITME". The "STREAM" support for an older
433 ClamAV interface method is removed.
435 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
436 rows affected is given instead).
438 JH/11 The runtime Berkeley DB library version is now additionally output by
439 "exim -d -bV". Previously only the compile-time version was shown.
441 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
442 SMTP connection. Previously, when one had more receipients than the
443 first, an abortive onward connection was made. Move to full support for
444 multiple onward connections in sequence, handling cutthrough connection
445 for all multi-message initiating connections.
447 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
448 routers. Previously, a multi-recipient message would fail to match the
449 onward-connection opened for the first recipient, and cause its closure.
451 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
452 a timeout on read on a GnuTLS initiating connection, resulting in the
453 initiating connection being dropped. This mattered most when the callout
454 was marked defer_ok. Fix to keep the two timeout-detection methods
457 JH/15 Relax results from ACL control request to enable cutthrough, in
458 unsupported situations, from error to silently (except under debug)
459 ignoring. This covers use with PRDR, frozen messages, queue-only and
462 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
464 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
465 metadata, resulting in a crash in free().
467 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
468 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
469 Broken also in d185889f4, with init system revamp.
471 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
472 Previously we abruptly closed the connection after reading a malware-
473 found indication; now we go on to read the "scan ok" response line,
476 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
477 ACL. Previously, a crash would result.
479 JH/19 Speed up macro lookups during configuration file read, by skipping non-
480 macro text after a replacement (previously it was only once per line) and
481 by skipping builtin macros when searching for an uppercase lead character.
483 JH/20 DANE support moved from Experimental to mainline. The Makefile control
484 for the build is renamed.
486 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
487 was allocated for every new TLS startup, meaning one per message. Fix
488 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
490 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
491 reported the original. Fix to report (as far as possible) the ACL
492 result replacing the original.
494 JH/23 Fix memory leak during multi-message connections using STARTTLS under
495 OpenSSL. Certificate information is loaded for every new TLS startup,
496 and the resources needed to be freed.
498 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
500 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
503 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
504 DATA response info to the (existing) per-recipient response info for
505 the "C=" log element. It can have useful tracking info from the
506 destination system. Patch from Simon Arlott.
508 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
509 length value. Previously this would segfault.
511 HS/02 Support Avast multiline protoocol, this allows passing flags to
512 newer versions of the scanner.
514 JH/28 Ensure that variables possibly set during message acceptance are marked
515 dead before release of memory in the daemon loop. This stops complaints
516 about them when the debug_store option is enabled. Discovered specifically
517 for sender_rate_period, but applies to a whole set of variables.
518 Do the same for the queue-runner and queue-list loops, for variables set
519 from spool message files. Do the same for the SMTP per-message loop, for
520 certain variables indirectly set in ACL operations.
522 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
523 as a multi-recipient message from a mailinglist manager). The coding had
524 an arbitrary cutoff number of characters while checking for more input;
525 enforced by writing a NUL into the buffer. This corrupted long / fast
526 input. The problem was exposed more widely when more pipelineing of SMTP
527 responses was introduced, and one Exim system was feeding another.
528 The symptom is log complaints of SMTP syntax error (NUL chars) on the
529 receiving system, and refused recipients seen by the sending system
530 (propating to people being dropped from mailing lists).
531 Discovered and pinpointed by David Carter.
533 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
534 replaced by the ${authresults } expansion.
536 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
538 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
539 allows proper process termination in container environments.
541 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
542 Previously the "final dot" had a newline after it; ensure it is CR,LF.
544 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
545 and "err_perm", deprecated since 4.83 when the RFC-defined words
546 "temperror" and "permerror" were introduced.
548 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
549 transport-filters or DKIM-signing. The restriction was lost in the
550 consolidation of verify-callout and delivery SMTP handling.
551 Extend the restriction to also cover ARC-signing.
553 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
554 in defer=pass mode supply a 450 to the initiator. Previously the message
557 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
558 tls_require_ciphers is used as before.
560 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
561 "pass_unscanned". Only tmpfails from the scanner are written to
562 the paniclog, as they may require admin intervention (permission
563 denied, license issues). Other scanner errors (like decompression
564 bombs) do not cause a paniclog entry.
566 JH/36 Fix reinitialisation of DKIM logging variable between messages.
567 Previously it was possible to log spurious information in receive log
570 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
571 triggered odd behaviour from Outlook Express clients.
573 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
576 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
577 since the IETF WG has not yet settled on that versus the original
578 "bare" representation.
580 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
581 Previously the millisecond value corrupted the output.
582 Fix also for syslog_pid=no and log_selector +pid, for which the pid
583 corrupted the output.
589 JH/01 Rework error string handling in TLS interface so that the caller in
590 more cases is responsible for logging. This permits library-sourced
591 string to be attached to addresses during delivery, and collapses
592 pairs of long lines into single ones.
594 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
595 during configuration. Wildcards are allowed and expanded.
597 JH/02 Rework error string handling in DKIM to pass more info back to callers.
598 This permits better logging.
600 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
601 do not close it down and have the child transport start it up again on
602 the passed-on TCP connection. Instead, proxy the child (and any
603 subsequent ones) for TLS via a unix-domain socket channel. Logging is
604 affected: the continued delivery log lines do not have any DNSSEC, TLS
605 Certificate or OCSP information. TLS cipher information is still logged.
607 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
608 identical IP addresses on different listening ports. Will also affect
611 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
612 add noisy ifdef guards to special-case this sillyness.
613 Patch from Bernd Kuhls.
615 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
616 than 255 are no longer allowed.
618 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
619 Disable the session-cache too, which might reduce our load. Since we
620 currrectly use a new context for every connection, both as server and
621 client, there is no benefit for these.
622 GnuTLS appears to not support tickets server-side by default (we don't
623 call gnutls_session_ticket_enable_server()) but client side is enabled
624 by default on recent versions (3.1.3 +) unless the PFS priority string
627 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
628 <https://reproducible-builds.org/specs/source-date-epoch/>.
630 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
631 the check for any unsuccessful recipients did not notice the limit, and
632 erroneously found still-pending ones.
634 JH/08 Pipeline CHUNKING command and data together, on kernels that support
635 MSG_MORE. Only in-clear (not on TLS connections).
637 JH/09 Avoid using a temporary file during transport using dkim. Unless a
638 transport-filter is involved we can buffer the headers in memory for
639 creating the signature, and read the spool data file once for the
640 signature and again for transmission.
642 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
643 4.77 as the kernel support then wasn't solid, having issues in 64bit
644 mode. Now, it's been long enough. Add support for FreeBSD also.
646 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
647 case where the routing stage had gathered several addresses to send to
648 a host before calling the transport for the first, we previously failed
649 to close down TLS in the old transport process before passing the TCP
650 connection to the new process. The new one sent a STARTTLS command
651 which naturally failed, giving a failed delivery and bloating the retry
652 database. Investigation and fix prototype from Wolfgang Breyha.
654 JH/12 Fix check on SMTP command input synchronisation. Previously there were
655 false-negatives in the check that the sender had not preempted a response
656 or prompt from Exim (running as a server), due to that code's lack of
657 awareness of the SMTP input buffering.
659 PP/04 Add commandline_checks_require_admin option.
660 Exim drops privileges sanely, various checks such as -be aren't a
661 security problem, as long as you trust local users with access to their
662 own account. When invoked by services which pass untrusted data to
663 Exim, this might be an issue. Set this option in main configuration
664 AND make fixes to the calling application, such as using `--` to stop
667 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
668 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
669 responses to those, into a single TLS record each way (this usually means
670 a single packet). As a side issue, smtp_enforce_sync now works on TLS
673 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
674 affects you only if you're dancing at the edge of the param size limits.
675 If you are, and this message makes sense to you, then: raise the
676 configured limit or use OpenSSL 1.1. Nothing we can do for older
679 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
680 cmdline element to get the documented default one. Previously it was
683 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
684 are unsuable for non-message-specific reasons - eg. connection timeout,
685 banner-time rejection.
687 JH/16 Fix logging of delivery remote port, when specified by router, under
690 PP/06 Repair manualroute's ability to take options in any order, even if one
691 is the name of a transport.
694 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
696 JH/17 Change the list-building routines interface to use the expanding-string
697 triplet model, for better allocation and copying behaviour.
699 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
700 Previously it was constructed the first time a possibly-matching string
701 was met in the configuration file input during startup; now it is done
704 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
705 compatible one, to avoid the (poorly documented) possibility of a config
706 file in the working directory redirecting the DB files, possibly correpting
707 some existing file. CVE-2017-10140 assigned for BDB.
709 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
710 cache-hot. Previously, although the result was properly cached, the
711 initial verify call returned a defer.
713 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
714 the main verify for receipient in uncached-mode.
716 JH/22 Retire historical build files to an "unsupported" subdir. These are
717 defined as "ones for which we have no current evidence of testing".
719 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
720 if present. Previously it was ignored.
722 JH/24 Start using specified-initialisers in C structure init coding. This is
723 a C99 feature (it's 2017, so now considered safe).
725 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
726 if was a fixed-sized field and bitmask ops via macros; it is now more
729 PP/07 GitHub PR 56: Apply MariaDB build fix.
730 Patch provided by Jaroslav Škarvada.
732 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
733 during Coverity cleanups [4.87 JH/47]
734 Diagnosis and fix provided by Michael Fischer v. Mollard.
736 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
737 the right size to place the terminating semicolon on its own folded
738 line, the header hash was calculated to an incorrect value thanks to
739 the (relaxed) space the fold became.
741 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
742 and confused the parent.
744 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
745 which could crash as a result. This could lead to undeliverable messages.
747 JH/28 Logging: "next input sent too soon" now shows where input was truncated
750 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
751 matters on fast-turnover and PID-randomising systems, which were getting
752 out-of-order delivery.
754 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
755 a possibly-overlapping copy. The symptom was that "Remote host closed
756 connection in response to HELO" was logged instead of the actual 4xx
759 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
760 Previously only that bufferd was discarded, resulting in SYMTP command
763 JH/32 DKIM: when a message has multiple signatures matching an identity given
764 in dkim_verify_signers, run the dkim acl once for each. Previously only
765 one run was done. Bug 2189.
767 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
768 "panic the current process" to "deliberately defer". The panic log is
769 still written with the problem list name; the mail and reject logs now
770 get a temp-reject line for the message that was being handled, saying
771 something like "domains check lookup or other defer". The SMTP 451
772 message is still "Temporary local problem".
774 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
775 A crafted sequence of BDAT commands could result in in-use memory beeing
776 freed. CVE-2017-16943.
778 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
779 from SMTP input. Previously it was always done; now only done for DATA
780 and not BDAT commands. CVE-2017-16944.
782 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
783 to the message (such as an overlong header line). Previously this was
784 not done and we did not exit BDAT mode. Followon from the previous item
785 though a different problem.
791 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
792 than -2003 did; needs libidn2 in addition to libidn.
794 JH/02 The path option on a pipe transport is now expanded before use.
796 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
797 Patch provided by "Björn", documentation fix added too.
799 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
800 missing a wire-to-host endian conversion.
802 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
803 close after a BDAT command line could be taken as a following command,
804 giving a synch failure. Fix by only checking for synch immediately
805 before acknowledging the chunk.
807 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
808 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
809 macro. Patches provided by Josh Soref.
811 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
812 Previously we did not; the RFC seems ambiguous and VRFY is not listed
813 by IANA as a service extension. However, John Klensin suggests that we
816 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
817 the dkim code may be unix-mode line endings rather than smtp wire-format
818 CRLF, so prepend a CR to any bare LF.
820 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
821 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
823 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
824 the first were themselves being wrongly included in the feed into dkim
825 processing; with most chunk sizes in use this resulted in an incorrect
826 body hash calculated value.
828 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
829 DKIM signature block, for verification. Although advised against by
830 standards it is specifically not ruled illegal.
832 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
834 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
835 missing a body hash (the bh= tag).
837 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
838 It seems that HAProxy sends the Proxy Protocol information in clear and
839 only then does a TLS startup, so do the same.
841 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
842 TCP connections (such as for Spamd) unless the daemon successfully set
843 Fast Open mode on its listening sockets. This fixes breakage seen on
844 too-old kernels or those not configured for Fast Open, at the cost of
845 requiring both directions being enabled for TFO, and TFO never being used
846 by non-daemon-related Exim processes.
848 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
849 endings, at least on the first header line. Try to canonify any that get
850 past that check, despite the cost.
852 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
853 now limited to an arbitrary five deep, while parsing addresses with the
854 strip_excess_angle_brackets option enabled.
856 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
857 instead leave the unprompted TLS handshake in socket buffer for the
858 TLS library to consume.
860 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
862 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
864 JH/16 Drop variables when they go out of scope. Memory management drops a whole
865 region in one operation, for speed, and this leaves assigned pointers
866 dangling. Add checks run only under the testsuite which checks all
867 variables at a store-reset and panics on a dangling pointer; add code
868 explicitly nulling out all the variables discovered. Fixes one known
869 bug: a transport crash, where a dangling pointer for $sending_ip_address
870 originally assigned in a verify callout, is re-used.
872 PP/06 Drop '.' from @INC in various Perl scripts.
874 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
876 PP/08 Reduce a number of compilation warnings under clang; building with
877 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
878 should be warning-free.
880 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
882 HS/01 Fix portability problems introduced by PP/08 for platforms where
883 realloc(NULL) is not equivalent to malloc() [SunOS et al].
885 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
886 chunk. This allows us to accept broken chunked messages. We need a more
887 general solution here.
889 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
890 already-broken messages in the queue.
892 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
894 JH/19 Fix reference counting bug in routing-generated-address tracking.
900 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
901 supports it and a size is available (ie. the sending peer gave us one).
903 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
904 years of being deprecated). The replacements are the ACLs
905 acl_smtp_mime and acl_not_smtp_mime.
907 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
908 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
909 or even an in-clear connection were permitted. Now, if the host lookup
910 was dnssec and dane was requested then the host is only used if the
911 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
912 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
913 if one fails this test.
914 This means that a poorly-configured remote DNS will make it incommunicado;
915 but it protects against a DNS-interception attack on it.
917 JH/04 Bug 1810: make continued-use of an open smtp transport connection
918 non-noisy when a race steals the message being considered.
920 JH/05 If main configuration option tls_certificate is unset, generate a
921 self-signed certificate for inbound TLS connections.
923 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
924 in rewrites and routers.
926 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
927 and logged a warning sing 4.83; now they are a configuration file error.
929 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
930 (lacking @domain). Apply the same qualification processing as RCPT.
932 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
934 JH/10 Support ${sha256:} applied to a string (as well as the previous
937 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
938 a cutthrough deliver is pending, as we always want to make a connection.
939 This also avoids re-routing the message when later placing the cutthrough
940 connection after a verify cache hit.
941 Do not update it with the verify result either.
943 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
944 when routing results in more than one destination address.
946 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
947 signing (which inhibits the cutthrough capability). Previously only
948 the presence of an option was tested; now an expansion evaluating as
949 empty is permissible (obviously it should depend only on data available
950 when the cutthrough connection is made).
952 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
953 the relevant preceding SMTP command did not note the pipelining mode.
955 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
956 Previously they were not counted.
958 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
959 as one having no matching records. Previously we deferred the message
960 that needed the lookup.
962 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
963 distinguished as "(=".
965 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
966 for missing MX records. Previously it only worked for missing A records.
968 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
970 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
971 after the data-go-ahead and data-ack. Patch from Jason Betts.
973 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
974 even for a "none" policy. Patch from Tony Meyer.
976 JH/22 Fix continued use of a connection for further deliveries. If a port was
977 specified by a router, it must also match for the delivery to be
980 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
981 When one of the recipients of a message was unsuitable for the connection
982 (has no matching addresses), we lost track of needing to mark it
983 deferred. As a result mail would be lost.
985 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
987 JH/25 Decoding ACL controls is now done using a binary search; the source code
988 takes up less space and should be simpler to maintain. Merge the ACL
989 condition decode tables also, with similar effect.
991 JH/26 Fix problem with one_time used on a redirect router which returned the
992 parent address unchanged. A retry would see the parent address marked as
993 delivered, so not attempt the (identical) child. As a result mail would
996 JH/27 Fix a possible security hole, wherein a process operating with the Exim
997 UID can gain a root shell. Credit to http://www.halfdog.net/ for
998 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1001 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1002 Main config options check_{log,spool}_{inodes,space} are now
1003 100 inodes, 10MB unless set otherwise in the configuration.
1005 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1006 Previously it only applied to the main-section connection policy
1009 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1011 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1012 by me. Added RFC7919 DH primes as an alternative.
1014 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1015 are not in the system include path.
1017 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1018 GnuTLS, when a session startup failed (eg because the client disconnected)
1019 Exim did stdio operations after fclose. This was exposed by a recent
1020 change which nulled out the file handle after the fclose.
1022 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1023 signed directly by the cert-signing cert, rather than an intermediate
1024 OCSP-signing cert. This is the model used by LetsEncrypt.
1026 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1028 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1029 an incoming connection.
1031 HS/02 Bug 1802: Do not half-close the connection after sending a request
1034 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1035 fallback to "prime256v1".
1037 JH/34 SECURITY: Use proper copy of DATA command in error message.
1038 Could leak key material. Remotely exploitable. CVE-2016-9963.
1044 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1045 and 3.4.4 - once the server is enabled to respond to an OCSP request
1046 it does even when not requested, resulting in a stapling non-aware
1047 client dropping the TLS connection.
1049 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1050 support variable-length bit vectors. No functional change.
1052 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1053 The I= interface field on outgoing lines is now after the H= remote
1054 host field, same as incoming lines. There is a separate
1055 outgoing_interface log selector which allows you to disable the
1058 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1059 If not running log_selector +smtp_connection the mainlog would be held
1060 open indefinitely after a "too many connections" event, including to a
1061 deleted file after a log rotate. Leave the per net connection logging
1062 leaving it open for efficiency as that will be quickly detected by the
1063 check on the next write.
1065 HS/01 Bug 1671: Fix post transport crash.
1066 Processing the wait-<transport> messages could crash the delivery
1067 process if the message IDs didn't exist for some reason. When
1068 using 'split_spool_directory=yes' the construction of the spool
1069 file name failed already, exposing the same netto behaviour.
1071 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1072 mime_regex ACL conditions.
1074 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1075 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1076 to HELO, local diagnostic string.
1078 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1079 log line to debug. Even when configured with a tls authenticator many
1080 client connections are expected to not authenticate in this way, so
1081 an authenticate fail is not an error.
1083 HS/02 Add the Exim version string to the process info. This way exiwhat
1084 gives some more detail about the running daemon.
1086 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1087 matter for fast-change records such as DNSBLs.
1089 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1090 retry record, even if constant. There may be multiple transports with
1091 different interface settings and the retry behaviour needs to be kept
1094 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1096 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1098 JH/10 Bug 840: fix log_defer_output option of pipe transport
1100 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1101 hosts_randomize. This matters a lot when combined with mua_wrapper.
1103 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1104 ${quote_pgsql:<string>} operator.
1106 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1108 JH/14 Change header file name for freeradius-client. Relevant if compiling
1109 with Radius support; from the Gentoo tree and checked under Fedora.
1111 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1113 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1114 option transport_filter, meaning no filtering.
1116 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1118 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1119 defaults to "*" (all hosts). The variable is now available when not built
1120 with TLS, default unset, mainly to enable keeping the testsuite sane.
1121 If a server certificate is not supplied (via tls_certificate) an error is
1122 logged, and clients will find TLS connections fail on startup. Presumably
1123 they will retry in-clear.
1124 Packagers of Exim are strongly encouraged to create a server certificate
1125 at installation time.
1127 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1128 with the $config_file variable.
1130 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1131 in transport context, after the attempt, and per-recipient. The latter type
1132 is per host attempted. The event data is the error message, and the errno
1133 information encodes the lookup type (A vs. MX) used for the (first) host,
1134 and the trailing two digits of the smtp 4xx response.
1136 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1137 to write to mainlog (or rejectlog, paniclog) in the window between file
1138 creation and permissions/ownership being changed. Particularly affects
1139 installations where exicyclog is run as root, rather than exim user;
1140 result is that the running daemon panics and dies.
1142 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1144 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1145 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1146 "pri" and "weight". Note that the previous implicit priority given by the
1147 list order is no longer honoured.
1149 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1150 for DKIM processing.
1152 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1153 by defining SUPPORT_SOCKS.
1155 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1156 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1157 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1158 variables are renamed to proxy_{local,external}_{address,port}.
1160 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1161 for a build by defining SUPPORT_I18N
1163 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1164 of the query string, and make ${quote_redis:} do that quoting.
1166 JH/29 Move Events support from Experimental to mainline, enabled by default
1167 and removable for a build by defining DISABLE_EVENT.
1169 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1171 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1172 cached by the daemon.
1174 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1175 by defining LOOKUP_REDIS. The libhiredis library is required.
1177 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1178 keys are given for lookup.
1180 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1181 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1182 only supported when built with TLS support. The PolarSSL SHA routines
1183 are still used when the TLS library is too old for convenient support.
1185 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1186 openssl_options), for security. OpenSSL forces this from version 1.1.0
1187 server-side so match that on older versions.
1189 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1190 allocation for $value could be released as the expansion processing
1191 concluded, but leaving the global pointer active for it.
1193 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1194 and to use the domains and local_parts ACL conditions.
1196 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1197 incorrectly not doubled on cutthrough transmission, hence seen as a
1198 body-termination at the receiving system - resulting in truncated mails.
1199 Commonly the sender saw a TCP-level error, and retransmitted the message
1200 via the normal store-and-forward channel. This could result in duplicates
1201 received - but deduplicating mailstores were liable to retain only the
1202 initial truncated version.
1204 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1206 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1208 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1209 we're in there, support oversigning also; bug 1309.
1211 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1213 HS/04 Add support for keep_environment and add_environment options.
1215 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1216 either intentional arithmetic overflow during PRNG, or testing config-
1219 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1220 delivery resulted in actual delivery. Cancel cutthrough before DATA
1223 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1224 rejects a recipient: pass the reject to the originator.
1226 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1227 Many were false-positives and ignorable, but it's worth fixing the
1230 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1231 for the new environment-manipulation done at startup. Move the routines
1232 from being local to tls.c to being global via the os.c file.
1234 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1235 an extract embedded as result-arg for a map, the first arg for extract
1236 is unavailable so we cannot tell if this is a numbered or keyed
1237 extraction. Accept either.
1243 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1246 JH/02 The smtp transport option "multi_domain" is now expanded.
1248 JH/03 The smtp transport now requests PRDR by default, if the server offers
1251 JH/04 Certificate name checking on server certificates, when exim is a client,
1252 is now done by default. The transport option tls_verify_cert_hostnames
1253 can be used to disable this per-host. The build option
1254 EXPERIMENTAL_CERTNAMES is withdrawn.
1256 JH/05 The value of the tls_verify_certificates smtp transport and main options
1257 default to the word "system" to access the system default CA bundle.
1258 For GnuTLS, only version 3.0.20 or later.
1260 JH/06 Verification of the server certificate for a TLS connection is now tried
1261 (but not required) by default. The verification status is now logged by
1262 default, for both outbound TLS and client-certificate supplying inbound
1265 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1268 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1269 Status Notification (bounce) messages are now MIME format per RFC 3464.
1270 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1271 under the control of the dsn_advertise_hosts option, and routers may
1272 have a dsn_lasthop option.
1274 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1275 default, modifiable by a malware= option. The list separator for
1276 the options can now be changed in the usual way. Bug 68.
1278 JH/10 The smtp_receive_timeout main option is now expanded before use.
1280 JH/11 The incoming_interface log option now also enables logging of the
1281 local interface on delivery outgoing connections.
1283 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1284 if the interface and destination host and port all match.
1286 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1289 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1290 Patch from Andrew Lewis.
1292 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1293 now supports optional time-restrictions, weighting, and priority
1294 modifiers per server. Patch originally by <rommer@active.by>.
1296 JH/16 The spamd_address main option now supports a mixed list of local
1297 and remote servers. Remote servers can be IPv6 addresses, and
1298 specify a port-range.
1300 JH/17 Bug 68: The spamd_address main option now supports an optional
1301 timeout value per server.
1303 JH/18 Bug 1581: Router and transport options headers_add/remove can
1304 now have the list separator specified.
1306 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1309 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1312 JH/21 Support for the A6 type of dns record is withdrawn.
1314 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1315 rather than the verbs used.
1317 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1318 from 255 to 1024 chars.
1320 JH/24 Verification callouts now attempt to use TLS by default.
1322 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1323 are generic router options now. The defaults didn't change.
1325 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1326 Original patch from Alexander Shikoff, worked over by JH.
1328 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1329 files if ZCAT_COMMAND is not executable.
1331 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1333 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1335 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1336 Normally benign, it bites when the pair was led to by a CNAME;
1337 modern usage is to not canonicalize the domain to a CNAME target
1338 (and we were inconsistent anyway for A-only vs AAAA+A).
1340 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1342 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1343 when evaluating $sender_host_dnssec.
1345 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1346 $sender_helo_dnssec variable.
1348 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1350 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1352 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1354 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1355 documented as working, but never had. Support all but $spam_report.
1357 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1358 added for tls authenticator.
1360 HS/03 Add perl_taintmode main config option
1366 TL/01 When running the test suite, the README says that variables such as
1367 no_msglog_check are global and can be placed anywhere in a specific
1368 test's script, however it was observed that placement needed to be near
1369 the beginning for it to behave that way. Changed the runtest perl
1370 script to read through the entire script once to detect and set these
1371 variables, reset to the beginning of the script, and then run through
1372 the script parsing/test process like normal.
1374 TL/02 The BSD's have an arc4random API. One of the functions to induce
1375 adding randomness was arc4random_stir(), but it has been removed in
1376 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1377 function when detected.
1379 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1380 cause callback expansion.
1382 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1383 syntax errors in an expansion can be treated as a string instead of
1384 logging or causing an error, due to the internal use of bool_lax
1385 instead of bool when processing it.
1387 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1388 server certificates when making smtp deliveries.
1390 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1392 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1394 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1396 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1397 Merged patch from Sebastian Wiedenroth.
1399 JH/05 Fix results-pipe from transport process. Several recipients, combined
1400 with certificate use, exposed issues where response data items split
1401 over buffer boundaries were not parsed properly. This eventually
1402 resulted in duplicates being sent. This issue only became common enough
1403 to notice due to the introduction of connection certificate information,
1404 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1406 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1407 size buffer was used, resulting in syntax errors when an expansion
1410 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1411 version 3.3.6 or later.
1413 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1414 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1415 both become "event_action", the variables become $event_name, $event_data
1416 and $event_defer_errno. There is a new variable $verify_mode, usable in
1417 routers, transports and related events. The tls:cert event is now also
1418 raised for inbound connections, if the main configuration event_action
1421 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1422 early OCSP support, but no stapling (appears to be less than 1.0.0).
1424 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1425 server certificate names available under the smtp transport option
1426 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1429 JH/10 Time-related extraction expansions from certificates now use the main
1430 option "timezone" setting for output formatting, and are consistent
1431 between OpenSSL and GnuTLS compilations. Bug 1541.
1433 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1434 encoded parameter in the incoming message. Bug 1558.
1436 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1437 include certificate info, eximon was claiming there were spoolfile
1440 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1442 JH/14 Log delivery-related information more consistently, using the sequence
1443 "H=<name> [<ip>]" wherever possible.
1445 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1446 are problematic for Debian distribution, omit them from the release
1449 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1451 JH/16 Fix string representation of time values on 64bit time_t architectures.
1454 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1455 output list separator was used.
1460 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1461 checkers that were complaining about end of non-void function with no
1464 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1465 This was a regression introduced in 4.83 by another bugfix.
1467 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1469 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1470 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1476 TF/01 Correctly close the server side of TLS when forking for delivery.
1478 When a message was received over SMTP with TLS, Exim failed to clear up
1479 the incoming connection properly after forking off the child process to
1480 deliver the message. In some situations the subsequent outgoing
1481 delivery connection happened to have the same fd number as the incoming
1482 connection previously had. Exim would try to use TLS and fail, logging
1483 a "Bad file descriptor" error.
1485 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1486 utilities have not been installed.
1488 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1489 temporary space as the ACL may create new global variables.
1491 TL/01 LDAP support uses per connection or global context settings, depending
1492 upon the detected version of the libraries at build time.
1494 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1495 to extract and use the src ip:port in logging and expansions as if it
1496 were a direct connection from the outside internet. PPv2 support was
1497 updated based on HAProxy spec change in May 2014.
1499 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1501 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1502 Properly escape header and check for NULL return.
1504 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1507 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1509 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1510 characters in header names, implemented as a verify condition.
1511 Contributed by Michael Fischer v. Mollard.
1513 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1514 results permerror and temperror. Previous values are deprecated but
1515 still accepted. In a future release, err_perm and err_temp will be
1516 completely removed, which will be a backward incompatibility if the
1517 ACL tests for either of these two old results. Patch contributed by
1518 user bes-internal on the mailing list.
1520 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1522 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1523 selectors, in both main and reject logs.
1525 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1528 JH/07 Add malware type "sock" for talking to simple daemon.
1530 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1532 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1533 routers/transports under cutthrough routing.
1535 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1536 numbers. Touch up "bool" conditional to keep the same definition.
1538 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1540 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1542 JH/12 Expand items in router/transport headers_add or headers_remove lists
1543 individually rather than the list as a whole. Bug 1452.
1545 Required for reasonable handling of multiple headers_ options when
1546 they may be empty; requires that headers_remove items with embedded
1547 colons must have them doubled (or the list-separator changed).
1549 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1550 view the policy declared in the DMARC record. Currently, $dmarc_status
1551 is a combined value of both the record presence and the result of the
1554 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1556 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1557 dnslookup router and the smtp transport (applying to the forward
1560 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1561 of ldap servers used for a specific lookup. Patch provided by Heiko
1564 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1565 New variable $lookup_dnssec_authenticated for observability.
1567 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1568 Patch submitted by Lars Timman.
1570 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1572 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1573 Requires trusted mode and valid format message id, aborts otherwise.
1574 Patch contributed by Heiko Schlichting.
1576 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1577 certextract with support for various fields. Bug 1358.
1579 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1580 is requested by default, modifiable by smtp transport option
1583 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1584 operate on certificate variables to give certificate fingerprints
1585 Also new ${sha256:cert_variable}.
1587 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1589 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1590 Christian Aistleitner.
1592 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1594 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1595 file. Patch from Wolfgang Breyha.
1597 JH/25 Expand the coverage of the delivery $host and $host_address to
1598 client authenticators run in verify callout. Bug 1476.
1600 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1601 align with daemon_smtp_ports. Bug 72.
1603 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1604 support and error reporting did not work properly.
1606 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1607 and is readable. Patch from Andrew Colin Kissa.
1609 TL/14 Enhance documentation of ${run expansion and how it parses the
1610 commandline after expansion, particularly in the case when an
1611 unquoted variable expansion results in an empty value.
1613 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1615 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1616 by renaming away from it. Addresses GH issue 15, reported by
1619 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1621 TL/15 SECURITY: prevent double expansion in math comparison functions
1622 (can expand unsanitized data). Not remotely exploitable.
1629 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1631 PP/02 Make -n do something, by making it not do something.
1632 When combined with -bP, the name of an option is not output.
1634 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1637 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1638 $sender_host_name and config options to manage this, and basic check
1641 PP/05 DSCP support for outbound connections and control modifier for inbound.
1643 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1644 (Only plugin which currently uses this is kerberos4, which nobody should
1645 be using, but we should make it available and other future plugins might
1646 conceivably use it, even though it would break NAT; stuff *should* be
1647 using channel bindings instead).
1649 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1650 name; added for Sendmail compatibility; requires admin caller.
1651 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1652 just ignore it); requires trusted caller.
1653 Also parse but ignore: -Ac -Am -X<logfile>
1656 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1658 TL/02 Add +smtp_confirmation as a default logging option.
1660 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1661 Patch by Magnus Holmgren from 2007-02-20.
1663 TL/04 Bugzilla 1281 - Spec typo.
1664 Bugzilla 1283 - Spec typo.
1665 Bugzilla 1290 - Spec grammar fixes.
1667 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1669 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1671 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1672 mailing list by by Dmitry Isaikin.
1674 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1676 JH/02 Support "G" suffix to numbers in ${if comparisons.
1678 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1680 NM/01 Bugzilla 1197 - Spec typo
1681 Bugzilla 1196 - Spec examples corrections
1683 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1685 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1686 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1689 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1690 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1692 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1693 "acl {{name}{arg}...}", and optional args on acl condition
1696 JH/05 Permit multiple router/transport headers_add/remove lines.
1698 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1700 JH/07 Avoid using a waiting database for a single-message-only transport.
1701 Performance patch from Paul Fisher. Bugzilla 1262.
1703 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1706 JH/09 Add $headers_added variable, with content from use of ACL modifier
1707 add_header (but not yet added to the message). Bugzilla 199.
1709 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1710 Pulled from Bugzilla 817 by Wolfgang Breyha.
1712 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1714 (nb: this is the same fix as in Exim 4.80.1)
1716 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1719 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1722 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1724 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1725 advertise SMTP AUTH mechanism to us, instead of a generic
1726 protocol violation error. Also, make Exim more robust to bad
1727 data from the Dovecot auth socket.
1729 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1731 When a queue runner is handling a message, Exim first routes the
1732 recipient addresses, during which it prunes them based on the retry
1733 hints database. After that it attempts to deliver the message to
1734 any remaining recipients. It then updates the hints database using
1737 So if a recipient address works intermittently, it can get repeatedly
1738 deferred at routing time. The retry hints record remains fresh so the
1739 address never reaches the final cutoff time.
1741 This is a fairly common occurrence when a user is bumping up against
1742 their storage quota. Exim had some logic in its local delivery code
1743 to deal with this. However it did not apply to per-recipient defers
1744 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1746 This change adds a proper retry rule check during routing so that the
1747 final cutoff time is checked against the message's age. We only do
1748 this check if there is an address retry record and there is not a
1749 domain retry record; this implies that previous attempts to handle
1750 the address had the retry_use_local_parts option turned on. We use
1751 this as an approximation for the destination being like a local
1752 delivery, as in LMTP.
1754 I suspect this new check makes the old local delivery cutoff check
1755 redundant, but I have not verified this so I left the code in place.
1757 TF/02 Correct gecos expansion when From: is a prefix of the username.
1759 Test 0254 submits a message to Exim with the header
1763 When I ran the test suite under the user fanf2, Exim expanded
1764 the header to contain my full name, whereas it should have added
1765 a Resent-Sender: header. It erroneously treats any prefix of the
1766 username as equal to the username.
1768 This change corrects that bug.
1770 GF/01 DCC debug and logging tidyup
1771 Error conditions log to paniclog rather than rejectlog.
1772 Debug lines prefixed by "DCC: " to remove any ambiguity.
1774 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1776 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1777 Bug spotted by Jeremy Harris; was flawed since initial commit.
1778 Would have resulted in OCSP responses post-SNI triggering an Exim
1779 NULL dereference and crash.
1781 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1783 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1784 Bug detection, analysis and fix by Samuel Thibault.
1785 Bugzilla 1331, Debian bug #698092.
1787 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1789 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1790 Server implementation by Todd Lyons, client by JH.
1791 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1792 config variable "prdr_enable" controls whether the server
1793 advertises the facility. If the client requests PRDR a new
1794 acl_data_smtp_prdr ACL is called once for each recipient, after
1795 the body content is received and before the acl_smtp_data ACL.
1796 The client is controlled by both of: a hosts_try_prdr option
1797 on the smtp transport, and the server advertisement.
1798 Default client logging of deliveries and rejections involving
1799 PRDR are flagged with the string "PRDR".
1801 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1802 fclose(). Diagnosis by Todd Lyons.
1804 PP/17 Update configure.default to handle IPv6 localhost better.
1805 Patch by Alain Williams (plus minor tweaks).
1808 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1809 This is now consistent with GnuTLS, and is now documented: the
1810 previous undocumented portable approach to treating the option as
1811 unset was to force an expansion failure. That still works, and
1812 an empty string is now equivalent.
1814 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1815 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1816 not performing validation itself.
1818 PP/20 Added force_command boolean option to pipe transport.
1819 Patch from Nick Koston, of cPanel Inc.
1821 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1824 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1826 PP/21 Fix eximon continuous updating with timestamped log-files.
1827 Broken in a format-string cleanup in 4.80, missed when I repaired the
1828 other false fix of the same issue.
1829 Report and fix from Heiko Schlichting.
1832 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1833 Report from Prashanth Katuri.
1835 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1836 It's SecureTransport, so affects any MacOS clients which use the
1837 system-integrated TLS libraries, including email clients.
1839 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1840 using a MIME ACL for non-SMTP local injection.
1841 Report and assistance in diagnosis by Warren Baker.
1843 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1845 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1847 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1848 last id that failed so it may be referenced in subsequent ACL's.
1850 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1853 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1854 ldap library initialization, allowing self-signed CA's to be
1855 used. Also properly sets require_cert option later in code by
1856 using NULL (global ldap config) instead of ldap handle (per
1857 session). Bug diagnosis and testing by alxgomz.
1859 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1860 the src/util/ subdirectory.
1862 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1863 renamed to Transport Post Delivery Action by Jeremy Harris, as
1866 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1867 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1868 redis_servers = needs to be configured which will be used by the redis
1869 lookup. Patch from Warren Baker, of The Packet Hub.
1871 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1873 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1874 hostname or reverse DNS when processing a host list. Used suggestions
1875 from multiple comments on this bug.
1877 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1879 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1880 Missed a few lines, added it to make the runtest require no keyboard
1883 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1884 contains upper case chars. Make router use caseful_local_part.
1886 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1887 support when GnuTLS has been built with p11-kit.
1893 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1895 This, or similar/improved, will also be change PP/11 of 4.82.
1901 PP/01 Handle short writes when writing local log-files.
1902 In practice, only affects FreeBSD (8 onwards).
1903 Bugzilla 1053, with thanks to Dmitry Isaikin.
1905 NM/01 Bugzilla 949 - Documentation tweak
1907 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1910 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1912 PP/02 Implemented gsasl authenticator.
1914 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1916 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1917 `pkg-config foo` for cflags/libs.
1919 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1920 with rest of GSASL and with heimdal_gssapi.
1922 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1923 `pkg-config foo` for cflags/libs for the TLS implementation.
1925 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1926 properties get this fed in as external SSF. A number of robustness
1927 and debugging improvements to the cyrus_sasl authenticator.
1929 PP/08 cyrus_sasl server now expands the server_realm option.
1931 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1932 Patch by Jeremy Harris.
1934 PP/10 Added dbmjz lookup type.
1936 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1938 PP/12 MAIL args handles TAB as well as SP, for better interop with
1939 non-compliant senders.
1940 Analysis and variant patch by Todd Lyons.
1942 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1943 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1944 Patch from Dirk Mueller <dmueller@suse.com>
1946 PP/13 tls_peerdn now print-escaped for spool files.
1947 Observed some $tls_peerdn in wild which contained \n, which resulted
1948 in spool file corruption.
1950 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1951 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1952 or write after TLS renegotiation, which otherwise led to messages
1955 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1956 as a tracking header (ie: a signed header comes before the signature).
1957 Patch from Wolfgang Breyha.
1959 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1960 comma-sep list; embedded commas doubled.
1962 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1964 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1966 Report and patch from Dmitry Banschikov.
1968 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1969 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1970 failures. We appear to now support TLS1.1+ with Exim.
1972 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1973 lets Exim select keys and certificates based upon TLS SNI from client.
1974 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1975 before an outbound SMTP session. New log_selector, +tls_sni.
1977 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1978 NULL dereference. Report and patch from Alun Jones.
1980 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1981 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1982 Not seeing resolver debug output on NetBSD, but suspect this is a
1983 resolver implementation change.
1985 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1986 Left warnings. Added "eximon gdb" invocation mode.
1988 PP/21 Defaulting "accept_8bitmime" to true, not false.
1990 PP/22 Added -bw for inetd wait mode support.
1992 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1993 locate the relevant includes and libraries. Made this the default.
1995 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1996 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1998 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1999 This may cause build issues on older platforms.
2001 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2002 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2003 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2004 Added SNI support via GnuTLS too.
2005 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2007 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2009 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2010 Applied second patch from Janne, implementing suggestion to default
2011 multiple-strings-in-record handling to match SPF spec.
2013 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2015 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2016 read-only, out of scope).
2017 Patch from Wolfgang Breyha, report from Stuart Northfield.
2019 PP/29 Fix three issues highlighted by clang analyser static analysis.
2020 Only crash-plausible issue would require the Cambridge-specific
2021 iplookup router and a misconfiguration.
2022 Report from Marcin Mirosław.
2024 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2026 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2027 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2028 As part of this, removing so much warning spew let me fix some minor
2029 real issues in debug logging.
2031 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2032 assignment on my part. Fixed.
2034 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2035 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2036 Janne Snabb (who went above and beyond: thank you).
2038 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2039 string otherwise requires a connection and a bunch more work and it's
2040 relatively easy to get wrong. Should also expose TLS library linkage
2043 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2044 64-bit ${eval} (JH/03).
2046 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2047 GNU libc to support some of the 64-bit stuff, should not lead to
2048 conflicts. Defined before os.h is pulled in, so if a given platform
2049 needs to override this, it can.
2051 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2052 protection layer was required, which is not implemented.
2053 Bugzilla 1254, patch from Wolfgang Breyha.
2055 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2056 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2057 tls_dhparam take prime identifiers. Also unbreak combination of
2058 OpenSSL+DH_params+TLSSNI.
2060 PP/39 Disable SSLv2 by default in OpenSSL support.
2066 PP/01 Solaris build fix for Oracle's LDAP libraries.
2067 Bugzilla 1109, patch from Stephen Usher.
2069 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2071 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2074 TF/02 Fix a couple more cases where we did not log the error message
2075 when unlink() failed. See also change 4.74-TF/03.
2077 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2078 lock up or crash if it happened to be inside a call to libc when it
2079 got a SIGUSR1 from exiwhat.
2081 The SIGUSR1 handler appends the current process status to the process
2082 log which is later printed by exiwhat. It used to use the general
2083 purpose logging code to do this, but several functions it calls are
2084 not safe for signals.
2086 The new output code in the SIGUSR1 handler is specific to the process
2087 log, and simple enough that it's easy to inspect for signal safety.
2088 Removing some special cases also simplifies the general logging code.
2089 Removing the spurious timestamps from the process log simplifies
2092 TF/04 Improved ratelimit ACL condition.
2094 The /noupdate option has been deprecated in favour of /readonly which
2095 has clearer semantics. The /leaky, /strict, and /readonly update modes
2096 are mutually exclusive. The update mode is no longer included in the
2097 database key; it just determines when the database is updated. (This
2098 means that when you upgrade Exim will forget old rate measurements.)
2100 Exim now checks that the per_* options are used with an update mode that
2101 makes sense for the current ACL. For example, when Exim is processing a
2102 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2103 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2104 must specify per_mail/readonly. If you omit the update mode it defaults to
2105 /leaky where that makes sense (as before) or /readonly where required.
2107 The /noupdate option is now undocumented but still supported for
2108 backwards compatibility. It is equivalent to /readonly except that in
2109 ACLs where /readonly is required you may specify /leaky/noupdate or
2110 /strict/noupdate which are treated the same as /readonly.
2112 A useful new feature is the /count= option. This is a generalization
2113 of the per_byte option, so that you can measure the throughput of other
2114 aggregate values. For example, the per_byte option is now equivalent
2115 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2117 The per_rcpt option has been generalized using the /count= mechanism
2118 (though it's more complicated than the per_byte equivalence). When it is
2119 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2120 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2121 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2122 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2123 Note that using per_rcpt with a non-readonly update mode in more than
2124 one ACL will cause the recipients to be double-counted. (The per_mail
2125 and per_byte options don't have this problem.)
2127 The handling of very low rates has changed slightly. If the computed rate
2128 is less than the event's count (usually one) then this event is the first
2129 after a long gap. In this case the rate is set to the same as this event's
2130 count, so that the first message of a spam run is counted properly.
2132 The major new feature is a mechanism for counting the rate of unique
2133 events. The new per_addr option counts the number of different
2134 recipients that someone has sent messages to in the last time period. It
2135 behaves like per_rcpt if all the recipient addresses are different, but
2136 duplicate recipient addresses do not increase the measured rate. Like
2137 the /count= option this is a general mechanism, so the per_addr option
2138 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2139 example, measure the rate that a client uses different sender addresses
2140 with the options per_mail/unique=$sender_address. There are further
2141 details in the main documentation.
2143 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2145 TF/06 Removed a few PCRE remnants.
2147 TF/07 Automatically extract Exim's version number from tags in the git
2148 repository when doing development or release builds.
2150 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2151 Bugzilla 879. Patch from Paul Fisher.
2153 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2154 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2157 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2159 PP/05 Variable $av_failed, true if the AV scanner deferred.
2160 Bugzilla 1078. Patch from John Horne.
2162 PP/06 Stop make process more reliably on build failure.
2163 Bugzilla 1087. Patch from Heiko Schlittermann.
2165 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2166 Bugzilla 1089. Patch from Heiko Schlittermann.
2168 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2169 Bugzilla 1131. Patch from Holger Weiß.
2171 PP/09 Handle IPv6 addresses with SPF.
2172 Bugzilla 860. Patch from Wolfgang Breyha.
2174 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2176 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2179 PP/11 match_* no longer expand right-hand-side by default.
2180 New compile-time build option, EXPAND_LISTMATCH_RHS.
2181 New expansion conditions, "inlist", "inlisti".
2183 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2185 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2187 PP/14 fix log_write() format string regression from TF/03.
2188 Bugzilla 1152. Patch from Dmitry Isaikin.
2194 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2196 PP/02 Harmonised TLS library version reporting; only show if debugging.
2197 Layout now matches that introduced for other libraries in 4.74 PP/03.
2199 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2201 PP/04 New "dns_use_edns0" global option.
2203 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2206 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2207 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2209 TK/01 Updated PolarSSL code to 0.14.2.
2210 Bugzilla 1097. Patch from Andreas Metzler.
2212 PP/07 Catch divide-by-zero in ${eval:...}.
2213 Fixes bugzilla 1102.
2215 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2218 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2219 format-string attack -- SECURITY: remote arbitrary code execution.
2221 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2222 time unintentionally subject to list matching rules, letting the header
2223 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2224 arbitrary string expansion). This allowed for information disclosure.
2226 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2227 INT_MIN/-1 -- value coerced to INT_MAX.
2233 NM/01 Workaround for PCRE version dependency in version reporting
2236 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2237 This fixes portability to compilers other than gcc, notably
2238 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2240 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2241 makefiles for portability to HP-UX and POSIX correctness.
2243 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2244 Also via indented variable definition in the Makefile.
2245 (Debugging by Oliver Heesakkers).
2247 PP/02 Restore caching of spamd results with expanded spamd_address.
2248 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2250 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2251 Improves build reliability. Fix from: Frank Elsner
2253 NM/02 Fix wide character breakage in the rfc2047 coding
2254 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2256 NM/03 Allow underscore in dnslist lookups
2257 Fixes bug 1026. Patch from Graeme Fowler
2259 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2260 Code patches from Adam Ciarcinski of NetBSD.
2262 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2265 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2266 is logged, to avoid truncation. Patch from John Horne.
2268 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2269 Patch from Jakob Hirsch.
2271 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2272 SQL string expansion failure details.
2273 Patch from Andrey Oktyabrski.
2275 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2276 Patch from Simon Arlott.
2278 PP/09 New lookups functionality failed to compile on old gcc which rejects
2279 extern declarations in function scope.
2280 Patch from Oliver Fleischmann
2282 PP/10 Use sig_atomic_t for flags set from signal handlers.
2283 Check getgroups() return and improve debugging.
2284 Fixed developed for diagnosis in bug 927 (which turned out to be
2287 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2288 Patch from Mark Zealey.
2290 PP/12 Bugzilla 1056: Improved spamd server selection.
2291 Patch from Mark Zealey.
2293 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2294 Based on patch from Heiko Schlittermann.
2296 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2297 Patch from Uwe Doering, sign-off by Michael Haardt.
2299 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2300 variable declaration deep within a block. Bug and patch from
2303 PP/15 lookups-Makefile IRIX compatibility coercion.
2305 PP/16 Make DISABLE_DKIM build knob functional.
2307 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2308 Patch by Simon Arlott
2310 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2311 variable argument macros. Our copy now differs from upstream.
2317 TF/01 Failure to get a lock on a hints database can have serious
2318 consequences so log it to the panic log.
2320 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2321 controlled using the smtp_confirmation log selector.
2323 TF/03 Include the error message when we fail to unlink a spool file.
2325 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2326 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2327 for maintaining out-of-tree patches for some time.
2329 PP/01 Bugzilla 139: Documentation and portability issues.
2330 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2331 Handle per-OS dynamic-module compilation flags.
2333 PP/02 Let /dev/null have normal permissions.
2334 The 4.73 fixes were a little too stringent and complained about the
2335 permissions on /dev/null. Exempt it from some checks.
2336 Reported by Andreas M. Kirchwitz.
2338 PP/03 Report version information for many libraries, including
2339 Exim version information for dynamically loaded libraries. Created
2340 version.h, now support a version extension string for distributors
2341 who patch heavily. Dynamic module ABI change.
2343 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2344 privilege escalation vulnerability whereby the Exim run-time user
2345 can cause root to append content of the attacker's choosing to
2348 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2351 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2352 If dropping privileges for untrusted macros, we disabled normal logging
2353 on the basis that it would fail; for the Exim run-time user, this is not
2354 the case, and it resulted in successful deliveries going unlogged.
2355 Fixed. Reported by Andreas Metzler.
2361 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2362 only prepend for the Resent-* case. Fixes regression introduced in
2363 Exim 4.70 by NM/22 for Bugzilla 607.
2365 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2366 increasing numbers of administrators be bitten by this.
2368 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2370 PP/03 Bugzilla 994: added openssl_options main configuration option.
2372 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2374 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2376 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2378 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2379 without a peer certificate, leading to a segfault because of an
2380 assumption that peers always have certificates. Be a little more
2381 paranoid. Problem reported by Martin Tscholak.
2383 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2384 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2385 NB: ClamAV planning to remove STREAM in "middle of 2010".
2386 CL also introduces -bmalware, various -d+acl logging additions and
2387 more caution in buffer sizes.
2389 PP/09 Implemented reverse_ip expansion operator.
2391 PP/10 Bugzilla 937: provide a "debug" ACL control.
2393 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2395 PP/12 Bugzilla 973: Implement --version.
2397 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2399 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2401 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2403 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2404 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2405 ignore trailing whitespace.
2407 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2409 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2412 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2413 Notification from Dr Andrew Aitchison.
2415 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2416 ExtendedDetectionInfo response format.
2417 Notification from John Horne.
2419 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2422 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2423 XSL and documented dependency on system catalogs, with examples of how
2426 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2429 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2430 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2431 configuration file which is writeable by the Exim user or group.
2433 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2434 of configuration files to cover files specified with the -C option if
2435 they are going to be used with root privileges, not just the default
2438 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2439 option (effectively making it always true).
2441 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2442 files to be used while preserving root privileges.
2444 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2445 that rogue child processes cannot use them.
2447 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2448 run-time user, instead of root.
2450 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2451 Exim run-time user without dropping privileges.
2453 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2454 result string, instead of calling string_vformat() twice with the same
2457 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2458 for other users. Others should always drop root privileges if they use
2459 -C on the command line, even for a whitelisted configure file.
2461 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2463 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2469 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2470 $data_path, and $header_path variables; fixed documentation bugs and
2473 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2474 exipick to access non-standard spools, including the "frozen" queue
2477 NM/01 Bugzilla 965: Support mysql stored procedures.
2478 Patch from Alain Williams
2480 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2482 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2483 Patch from Andreas Metzler
2485 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2486 Patch from Kirill Miazine
2488 NM/05 Bugzilla 671: Added umask to procmail example.
2490 JJ/03 installed exipick 20100323.0, fixing doc bug
2492 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2493 directory. Notification and patch from Dan Rosenberg.
2495 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2497 TK/02 Improve log output when DKIM signing operation fails.
2499 MH/01 Treat the transport option dkim_domain as a colon separated
2500 list, not as a single string, and sign the message with each element,
2501 omitting multiple occurences of the same signer.
2503 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2504 Bugzilla 985, 986. Patch by Simon Arlott
2506 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2507 Patch by Simon Arlott
2509 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2510 MBX locking. Notification from Dan Rosenberg.
2516 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2518 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2520 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2522 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2524 NM/04 Bugzilla 915: Flag broken perl installation during build.
2530 TK/01 Added patch by Johannes Berg that expands the main option
2531 "spamd_address" if it starts with a dollar sign.
2533 TK/02 Write list of recipients to X-Envelope-Sender header when building
2534 the mbox-format spool file for content scanning (suggested by Jakob
2537 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2538 (http://www.dcc-servers.net/) support via dccifd. Activated by
2539 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2541 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2542 by Mark Daniel Reidel <mr@df.eu>.
2544 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2545 When building exim an external PCRE library is now needed -
2546 PCRE is a system library on the majority of modern systems.
2547 See entry on PCRE_LIBS in EDITME file.
2549 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2550 conversation. Added nologin parameter to request.
2551 Patch contributed by Kirill Miazine.
2553 TF/01 Do not log submission mode rewrites if they do not change the address.
2555 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2557 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2558 log files in place. Contributed by Roberto Lima.
2560 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2562 TF/03 Bugzilla 615: When checking the local_parts router precondition
2563 after a local_part_suffix or local_part_prefix option, Exim now
2564 does not use the address's named list lookup cache, since this
2565 contains cached lookups for the whole local part.
2567 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2568 Robert Millan. Documentation is in experimental-spec.txt.
2570 TF/04 Bugzilla 668: Fix parallel build (make -j).
2572 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2574 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2575 Patch provided by Jan Srzednicki.
2577 TF/05 Leading white space used to be stripped from $spam_report which
2578 wrecked the formatting. Now it is preserved.
2580 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2581 that they are available at delivery time.
2583 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2585 TF/08 TLS error reporting now respects the incoming_interface and
2586 incoming_port log selectors.
2588 TF/09 Produce a more useful error message if an SMTP transport's hosts
2589 setting expands to an empty string.
2591 NM/06 Bugzilla 744: EXPN did not work under TLS.
2592 Patch provided by Phil Pennock.
2594 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2595 Patch provided by Richard Godbee.
2597 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2598 acl_smtp_notquit, added index entry.
2600 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2601 Patch provided by Eugene Bujak.
2603 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2604 accept(). Patch provided by Maxim Dounin.
2606 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2607 Patch provided by Phil Pennock.
2609 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2611 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2612 Patch provided by Brad "anomie" Jorsch.
2614 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2615 Patch provided by Dean Brooks.
2617 TK/05 Add native DKIM support (does not depend on external libraries).
2619 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2620 Patch provided by Graeme Fowler.
2622 NM/16 Bugzilla 851: Documentation example syntax fix.
2624 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2626 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2629 NM/19 Bugzilla 745: TLS version reporting.
2630 Patch provided by Phil Pennock.
2632 NM/20 Bugzilla 167: bool: condition support.
2633 Patch provided by Phil Pennock.
2635 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2636 clients. Patch provided by Phil Pennock.
2638 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2639 Patch provided by Brad "anomie" Jorsch.
2641 NM/23 Bugzilla 687: Fix misparses in eximstats.
2642 Patch provided by Heiko Schlittermann.
2644 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2645 Patch provided by Heiko Schlittermann.
2647 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2648 plus update to original patch.
2650 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2652 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2653 Patch provided by David Brownlee.
2655 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2657 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2659 NM/30 Bugzilla 888: TLS documentation bugfixes.
2661 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2663 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2664 Unlike the original bugzilla I have changed all shell scripts in src tree.
2666 NM/33 Bugzilla 898: Transport filter timeout fix.
2667 Patch by Todd Rinaldo.
2669 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2670 Patch by Serge Demonchaux.
2672 NM/35 Bugzilla 39: Base64 decode bug fixes.
2673 Patch by Jakob Hirsch.
2675 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2677 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2679 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2681 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2687 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2688 ALT-N's libdkim that I have put here:
2689 http://duncanthrax.net/exim-experimental/
2691 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2692 were called 'true' and it seems that C99 defines that as a reserved
2693 keyword to be used with 'bool' variable types. That means you could
2694 not include C99-style headers which use bools without triggering
2695 build errors in sieve.c.
2697 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2698 as mailq or other aliases. Changed the --help handling significantly
2699 to do whats expected. exim_usage() emits usage/help information.
2701 SC/01 Added the -bylocaldomain option to eximstats.
2703 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2705 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2707 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2713 PH/01 Another patch from the Sieve maintainer.
2715 PH/02 When an IPv6 address is converted to a string for single-key lookup
2716 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2717 dots are used instead of colons so that keys in lsearch files need not
2718 contain colons. This was done some time before quoting was made available
2719 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2720 (notated using the quote facility) so as to distinguish them from IPv4
2721 keys. This meant that lookups for IP addresses in host lists did not work
2722 for iplsearch lookups.
2724 This has been fixed by arranging for IPv6 addresses to be expressed with
2725 colons if the lookup type is iplsearch. This is not incompatible, because
2726 previously such lookups could never work.
2728 The situation is now rather anomalous, since one *can* have colons in
2729 ordinary lsearch keys. However, making the change in all cases is
2730 incompatible and would probably break a number of configurations.
2732 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2735 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2736 conversion specification without a maximum field width, thereby enabling
2737 a rogue spamd server to cause a buffer overflow. While nobody in their
2738 right mind would setup Exim to query an untrusted spamd server, an
2739 attacker that gains access to a server running spamd could potentially
2740 exploit this vulnerability to run arbitrary code as the Exim user.
2742 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2743 $primary_hostname instead of what libspf2 thinks the hosts name is.
2745 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2746 a directory entry by the name of the lookup key. Previously, if a
2747 symlink pointed to a non-existing file or a file in a directory that
2748 Exim lacked permissions to read, a lookup for a key matching that
2749 symlink would fail. Now it is enough that a matching directory entry
2750 exists, symlink or not. (Bugzilla 503.)
2752 PH/03 The body_linecount and body_zerocount variables are now exported in the
2755 PH/04 Added the $dnslist_matched variable.
2757 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2758 This means they are set thereafter only if the connection becomes
2761 PH/06 Added the client_condition to authenticators so that some can be skipped
2762 by clients under certain conditions.
2764 PH/07 The error message for a badly-placed control=no_multiline_responses left
2765 "_responses" off the end of the name.
2767 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2769 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2770 (without spaces) instead of just copying the configuration text.
2772 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2774 PH/11 Added $max_received_linelength.
2776 PH/12 Added +ignore_defer and +include_defer to host lists.
2778 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2779 way in which PCRE > 7.0 is built.
2781 PH/14 Implemented queue_only_load_latch.
2783 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2784 MAIL command. The effect was to mangle the value on 64-bit systems.
2786 PH/16 Another patch from the Sieve maintainer.
2788 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2790 PH/18 If a system quota error occurred while trying to create the file for
2791 a maildir delivery, the message "Mailbox is full" was not appended to the
2792 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2793 only to a quota excession during the actual writing of the file.
2795 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2796 characters?) which causes problems in log lines. The DN values are now
2797 passed through string_printing() before being added to log lines.
2799 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2800 and InterBase are left for another time.)
2802 PH/21 Added message_body_newlines option.
2804 PH/22 Guard against possible overflow in moan_check_errorcopy().
2806 PH/23 POSIX allows open() to be a macro; guard against that.
2808 PH/24 If the recipient of an error message contained an @ in the local part
2809 (suitably quoted, of course), incorrect values were put in $domain and
2810 $local_part during the evaluation of errors_copy.
2816 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2817 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2820 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2821 issue a MAIL command.
2823 PH/02 In an ACL statement such as
2825 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2827 if a client was not listed at all, or was listed with a value other than
2828 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2829 the condition was not true (as it should be), so access was not denied.
2830 The bug was that the ! inversion was incorrectly passed on to the second
2831 item. This has been fixed.
2833 PH/03 Added additional dnslists conditions == and =& which are different from
2834 = and & when the dns lookup returns more than one IP address.
2836 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2837 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2839 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2840 FSYNC, which compiles an option called disable_fsync that allows for
2841 bypassing fsync(). The documentation is heavily laced with warnings.
2843 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2845 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2846 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2847 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2848 including adding "make clean"; (3) Added -fPIC when compiling the test
2849 dynamically loaded module, to get rid of a warning.
2851 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2852 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2853 = 0s. The bug is otherwise harmless.
2855 PH/07 There was a bug in the dovecot authenticator such that the value of
2856 $auth1 could be overwritten, and so not correctly preserved, after a
2857 successful authentication. This usually meant that the value preserved by
2858 the server_setid option was incorrect.
2860 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2862 PH/09 Installed PCRE release 7.0.
2864 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2865 run for batched SMTP input. It is now run at the start of every message
2866 in the batch. While fixing this I discovered that the process information
2867 (output by running exiwhat) was not always getting set for -bs and -bS
2868 input. This is fixed, and it now also says "batched" for BSMTP.
2870 PH/11 Added control=no_pipelining.
2872 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2873 patch, slightly modified), and move the expansion of helo_data till after
2874 the connection is made in the smtp transport (so it can use these
2877 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2879 PH/14 Added log_selector = +pid.
2881 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2883 PH/16 Add ${if forany and ${if forall.
2885 PH/17 Added dsn_from option to vary the From: line in DSNs.
2887 PH/18 Flush SMTP output before performing a callout, unless control =
2888 no_callout_flush is set.
2890 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2891 was true (the default) a successful delivery failed to delete the retry
2892 item, thus causing premature timeout of the address. The bug is now
2895 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2897 PH/21 Long custom messages for fakedefer and fakereject are now split up
2898 into multiline responses in the same way that messages for "deny" and
2899 other ACL rejections are.
2901 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2902 with slight modification.
2904 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2905 draft, changing the syntax and factoring some duplicate code".
2907 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2908 for deliveries of the second and subsequent messages over the same SMTP
2911 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2912 ${reduce, with only minor "tidies".
2914 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2916 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2917 expansion side effects.
2919 PH/27 When a message times out after an over-quota error from an Exim-imposed
2920 quota, the bounce message says "mailbox is full". This message was not
2921 being given when it was a system quota that was exceeded. It now should
2924 MH/03 Made $recipients available in local_scan(). local_scan() already has
2925 better access to the recipient list through recipients_list[], but
2926 $recipients can be useful in postmaster-provided expansion strings.
2928 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2929 in the case of a MAIL command with additional options following the
2930 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2931 were accidentally chopped off.
2933 PH/29 SMTP synchronization checks are implemented when a command is read -
2934 there is a check that no more input is waiting when there shouldn't be
2935 any. However, for some commands, a delay in an ACL can mean that it is
2936 some time before the response is written. In this time, more input might
2937 arrive, invalidly. So now there are extra checks after an ACL has run for
2938 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2939 pipelining has not been advertised.
2941 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2943 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2944 returned in the SMTP response when smtp_return_error_details was false.
2945 This has been fixed.
2947 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2948 instead of the C I/O that was originally supplied, because problems were
2949 reported on Solaris.
2951 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2952 Exim which did not show up earlier: it was assuming that a call to
2953 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2954 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2955 was a macro that became an assignment, so it seemed to work. This has
2956 changed to a proper function call with a void return, hence the compile
2957 error. Exim's code has been fixed.
2959 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2962 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2964 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2966 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2967 because it turns out to be too big or there is a timeout) there is no
2968 "Completed" line in the log. When some messages of this type were
2969 selected by exigrep, they were listed as "not completed". Others were
2970 picked up by some special patterns. I have improved the selection
2971 criteria to be more general.
2973 PH/38 The host_find_failed option in the manualroute router can now be set
2974 to "ignore", to completely ignore a host whose IP address cannot be
2975 found. If all hosts are ignored, the behaviour is controlled by the new
2976 host_all_ignored option.
2978 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2979 homing or because of multiple MX records with /mx) generated more than
2980 one IP address, and the following item turned out to be the local host,
2981 all the secondary addresses of the first item were incorrectly removed
2982 from the list, along with the local host and any following hosts (which
2983 is what is supposed to happen).
2985 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2986 whoever called Exim into the -H file. In the case of the daemon it was
2987 behaving confusingly. When first started, it used values for whoever
2988 started the daemon, but after a SIGHUP it used the Exim user (because it
2989 calls itself on a restart). I have changed the code so that it now always
2992 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2993 message are rejected with the same error (e.g. no authentication or bad
2994 sender address), and a DATA command is nevertheless sent (as can happen
2995 with PIPELINING or a stupid MUA), the error message that was given to the
2996 RCPT commands is included in the rejection of the DATA command. This is
2997 intended to be helpful for MUAs that show only the final error to their
3000 PH/42 Another patch from the Sieve maintainer.
3002 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3003 Eximstats - Fixed some broken HTML links and added missing column headers
3005 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3006 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3008 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3010 PH/43 Yet another patch from the Sieve maintainer.
3012 PH/44 I found a way to check for a TCP/IP connection going away before sending
3013 the response to the final '.' that terminates a message, but only in the
3014 case where the client has not sent further data following the '.'
3015 (unfortunately, this is allowed). However, in many cases there won't be
3016 any further data because there won't be any more messages to send. A call
3017 to select() can be used: if it shows that the input is "ready", there is
3018 either input waiting, or the socket has been closed. An attempt to read
3019 the next input character can distinguish the two cases. Previously, Exim
3020 would have sent an OK response which the client would never have see.
3021 This could lead to message repetition. This fix should cure that, at
3022 least in a lot of common cases.
3024 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3025 advertised in response to EHLO.
3031 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3032 fixed by 4.65/MH/01 (is this a record?) are fixed:
3034 (i) An empty string was always treated as zero by the numeric comparison
3035 operators. This behaviour has been restored.
3037 (ii) It is documented that the numeric comparison operators always treat
3038 their arguments as decimal numbers. This was broken in that numbers
3039 starting with 0 were being interpreted as octal.
3041 While fixing these problems I realized that there was another issue that
3042 hadn't been noticed. Values of message_size_limit (both the global option
3043 and the transport option) were treated as octal if they started with 0.
3044 The documentation was vague. These values are now always treated as
3045 decimal, and I will make that clear in the documentation.
3051 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3052 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3055 MH/01 Don't check that the operands of numeric comparison operators are
3056 integers when their expansion is in "skipping" mode (fixes bug
3057 introduced by 4.64-PH/07).
3059 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3060 child addresses, Exim now panics and dies. Previously, because the count
3061 is held in a short int, deliveries were likely to be lost. As such a
3062 large number of recipients for a single message is ridiculous
3063 (performance will be very, very poor), I have chosen to impose a limit
3064 rather than extend the field.
3070 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3071 leftover -K file (the existence of which was triggered by #402).
3072 While we were at it, introduced process PID as part of the -K
3073 filename. This should rule out race conditions when creating
3076 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3077 processing considerably. Previous code took too long for large mails,
3078 triggering a timeout which in turn triggers #401.
3080 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3081 in the DK code in transports.c. sendfile() is not really portable,
3082 hence the _LINUX specificness.
3084 TF/01 In the add_headers option to the mail command in an Exim filter,
3085 there was a bug that Exim would claim a syntax error in any
3086 header after the first one which had an odd number of characters
3089 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3090 callout verification, Exim cached a "reject" for the entire domain. This
3091 is correct for most verifications, but it is not correct for a recipient
3092 verification with use_sender or use_postmaster set, because in that case
3093 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3094 case of MAIL FROM:<> rejection from other early rejections (e.g.
3095 rejection of HELO). When verifying a recipient using a non-null MAIL
3096 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3097 Whatever the result of the callout, the value of the domain cache is
3098 left unchanged (for any other kind of callout, getting as far as trying
3099 RCPT means that the domain itself is ok).
3101 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3104 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3105 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3106 session, the daemon ignores ECONNECT errors and logs others; it now
3107 ignores EPIPE as well.
3109 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3110 (quoted-printable decoding).
3112 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3113 later the small subsequent patch to fix an introduced bug.
3115 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3117 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3119 PH/08 An error is now given if message_size_limit is specified negative.
3121 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3122 to be given (somewhat) arbitrary names.
3124 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3127 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3128 miscellaneous code fixes
3130 PH/10 Added the log_reject_target ACL modifier to specify where to log
3133 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3134 hostname. This is wrong, because it relates to the incoming message (and
3135 probably the interface on which it is arriving) and not to the outgoing
3136 callout (which could be using a different interface). This has been
3137 changed to use the value of the helo_data option from the smtp transport
3138 instead - this is what is used when a message is actually being sent. If
3139 there is no remote transport (possible with a router that sets up host
3140 addresses), $smtp_active_hostname is used.
3142 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3143 tweaks were necessary in order to get it to work (see also 21 below):
3144 (a) The code assumed that strncpy() returns a negative number on buffer
3145 overflow, which isn't the case. Replaced with Exim's string_format()
3147 (b) There were several signed/unsigned issues. I just did the minimum
3148 hacking in of casts. There is scope for a larger refactoring.
3149 (c) The code used strcasecmp() which is not a standard C function.
3150 Replaced with Exim's strcmpic() function.
3151 (d) The code set only $1; it now sets $auth1 as well.
3152 (e) A simple test gave the error "authentication client didn't specify
3153 service in request". It would seem that Dovecot has changed its
3154 interface. Fortunately there's a specification; I followed it and
3155 changed what the client sends and it appears to be working now.
3157 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3160 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3161 address A is aliased to B and C, where B exists and C does not. Without
3162 -v the output is "A verified" because verification stops after a
3163 successful redirection if more than one address is generated. However,
3164 with -v the child addresses are also verified. Exim was outputting "A
3165 failed to verify" and then showing the successful verification for C,
3166 with its parentage. It now outputs "B failed to verify", showing B's
3167 parentage before showing the successful verification of C.
3169 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3170 look up a TXT record in a specific list after matching in a combined
3173 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3174 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3175 they consult the DNS. I had assumed they would set it the way they
3176 wanted; and indeed my experiments on Linux seem to show that in some
3177 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3178 To be on the safe side, however, I have now made the interface to
3179 host_find_byname() similar to host_find_bydns(), with an argument
3180 containing the DNS resolver options. The host_find_byname() function now
3181 sets these options at its start, just as host_find_bydns() does. The smtp
3182 transport options dns_qualify_single and dns_search_parents are passed to
3183 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3184 of host_find_byname() use the default settings of RES_DEFNAMES
3185 (qualify_single) but not RES_DNSRCH (search_parents).
3187 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3188 spool_read_header() do less string testing, by means of a preliminary
3189 switch on the second character of optional "-foo" lines. (This is
3190 overdue, caused by the large number of possibilities that now exist.
3191 Originally there were few.) While I was there, I also converted the
3192 str(n)cmp tests so they don't re-test the leading "-" and the first
3193 character, in the hope this might squeeze out yet more improvement.
3195 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3196 flag allowing group syntax was set by the header_syntax check but not
3197 turned off, possible causing trouble later; (2) The flag was not being
3198 set at all for the header_verify test, causing "group"-style headers to
3199 be rejected. I have now set it in this case, and also caused header_
3200 verify to ignore an empty address taken from a group. While doing this, I
3201 came across some other cases where the code for allowing group syntax
3202 while scanning a header line wasn't quite right (mostly, not resetting
3203 the flag correctly in the right place). These bugs could have caused
3204 trouble for malformed header lines. I hope it is now all correct.
3206 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3207 with the "reply" argument non-NULL. The code, however (which originally
3208 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3209 but it didn't always do it. This confused somebody who was copying the
3210 code for some other use. I have removed all the tests.
3212 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3213 feature that was used to support insecure browsers during the U.S. crypto
3214 embargo. It requires special client support, and Exim is probably the
3215 only MTA that supported it -- and would never use it because real RSA is
3216 always available. This code has been removed, because it had the bad
3217 effect of slowing Exim down by computing (never used) parameters for the
3218 RSA_EXPORT functionality.
3220 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3221 authenticator to fail if there's a tab character in the incoming data
3222 (there should never be unless someone is messing about, as it's supposed
3223 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3224 if the connection is using TLS or if the remote IP is the same as the
3225 local IP, and the "valid-client-cert option" if a client certificate has
3228 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3229 authenticators. This can be used for authorization after authentication
3230 succeeds. (In the case of plaintext, it servers for both authentication
3233 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3234 if any retry times were supplied.
3236 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3237 connection, where there is no client IP address to check. In this
3238 situation, the verify now always succeeds.
3240 PH/25 Applied John Jetmore's -Mset patch.
3242 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3244 PH/27 In a string expansion for a processed (not raw) header when multiple
3245 headers of the same name were present, leading whitespace was being
3246 removed from all of them, but trailing whitespace was being removed only
3247 from the last one. Now trailing whitespace is removed from each header
3248 before concatenation. Completely empty headers in a concatenation (as
3249 before) are ignored.
3251 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3252 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3254 PH/29 [Removed. This was a change that I later backed out, and forgot to
3255 correct the ChangeLog entry (that I had efficiently created) before
3256 committing the later change.]
3258 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3259 address errors (4xx response to RCPT) over the same connection as other
3260 messages routed to the same hosts. Such deliveries are always "forced",
3261 so retry times are not inspected. This resulted in far too many retries
3262 for the affected addresses. The effect occurred only when there were more
3263 hosts than the hosts_max_try setting in the smtp transport when it had
3264 the 4xx errors. Those hosts that it had tried were not added to the list
3265 of hosts for which the message was waiting, so if all were tried, there
3266 was no problem. Two fixes have been applied:
3268 (i) If there are any address or message errors in an SMTP delivery, none
3269 of the hosts (tried or untried) are now added to the list of hosts
3270 for which the message is waiting, so the message should not be a
3271 candidate for sending over the same connection that was used for a
3272 successful delivery of some other message. This seems entirely
3273 reasonable: after all the message is NOT "waiting for some host".
3274 This is so "obvious" that I'm not sure why it wasn't done
3275 previously. Hope I haven't missed anything, but it can't do any
3276 harm, as the worst effect is to miss an optimization.
3278 (ii) If, despite (i), such a delivery is accidentally attempted, the
3279 routing retry time is respected, so at least it doesn't keep
3280 hammering the server.
3282 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3283 in ${readsocket because some servers need this prod.
3285 PH/32 Added some extra debug output when updating a wait-xxx database.
3287 PH/33 The hint "could be header name not terminated by colon", which has been
3288 given for certain expansion errors for a long time, was not being given
3289 for the ${if def:h_colon_omitted{... case.
3291 PH/34 The spec says: "With one important exception, whenever a domain list is
3292 being scanned, $domain contains the subject domain." There was at least
3293 one case where this was not true.
3295 PH/35 The error "getsockname() failed: connection reset by peer" was being
3296 written to the panic log as well as the main log, but it isn't really
3297 panic-worthy as it just means the connection died rather early on. I have
3298 removed the panic log writing for the ECONNRESET error when getsockname()
3301 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3302 runs only) independently of the message's sender address. This meant
3303 that, if the 4xx error was in fact related to the sender, a different
3304 message to the same recipient with a different sender could confuse
3305 things. In particular, this can happen when sending to a greylisting
3306 server, but other circumstances could also provoke similar problems.
3307 I have changed the default so that the retry time for these errors is now
3308 based a combination of the sender and recipient addresses. This change
3309 can be overridden by setting address_retry_include_sender=false in the
3312 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3313 remote server are returned as part of bounce messages. This was not
3314 happening for LMTP over a pipe (the lmtp transport), but now it is the
3315 same for both kinds of LMTP.
3317 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3318 in header lines that were in fact CNAMEs. This is no longer the case.
3320 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3321 and queue runs started by the daemon processed all messages. This has
3322 been fixed so that -R and -S can now usefully be given with -q<time>.
3324 PH/40 Import PCRE release 6.7 (fixes some bugs).
3326 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3328 PH/42 Give an error if -q is specified more than once.
3330 PH/43 Renamed the variables $interface_address and $interface_port as
3331 $received_ip_address and $received_port, to make it clear that these
3332 values apply to message reception, and not to the outgoing interface when
3333 a message is delivered. (The old names remain recognized, of course.)
3335 PH/44 There was no timeout on the connect() call when using a Unix domain
3336 socket in the ${readsocket expansion. There now is.
3338 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3339 be meaningful with "accept".
3341 SC/01 Eximstats V1.43
3342 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3344 SC/02 Eximstats V1.44
3345 Use a glob alias rather than an array ref in the generated
3346 parser. This improves both readability and performance.
3348 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3349 Collect SpamAssassin and rejection statistics.
3350 Don't display local sender or destination tables unless
3351 there is data to show.
3352 Added average volumes into the top table text output.
3354 SC/04 Eximstats V1.46
3355 Collect data on the number of addresses (recipients)
3356 as well as the number of messages.
3358 SC/05 Eximstats V1.47
3359 Added 'Message too big' to the list of mail rejection
3360 reasons (thanks to Marco Gaiarin).
3362 SC/06 Eximstats V1.48
3363 Mainlog lines which have GMT offsets and are too short to
3364 have a flag are now skipped.
3366 SC/07 Eximstats V1.49 (Alain Williams)
3367 Added the -emptyok flag.
3369 SC/08 Eximstats V1.50
3370 Fixes for obtaining the IP address from reject messages.
3372 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3373 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3374 whitespace changes from 4.64-PH/27
3376 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3379 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3380 are found, allow negative numbers in numeric criteria)
3382 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3384 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3385 to match changes made in 4.64-PH/43
3387 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3389 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3390 "domain" is actually an IP address, and force a failure. This locks out
3391 those revolvers/nameservers that support "A-for-A" lookups, in
3392 contravention of the specifications.
3394 PH/48 When a host name was looked up from an IP address, and the subsequent
3395 forward lookup of the name timed out, the host name was left in
3396 $sender_host_name, contrary to the specification.
3398 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3399 restricted to single-key lookups, Exim was not diagnosing an error if
3400 * or *@ was used with a query-style lookup.
3402 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3404 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3405 long ago, but noone interested enough thought of it. Let's just say that
3406 the "1.1" means that there are some new functions that weren't there at
3407 some point in the past.
3409 PH/51 Error processing for expansion failure of helo_data from an smtp
3410 transport during callout processing was broken.
3412 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3413 tested/used via the -bh/-bhc/-bs options.
3415 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3416 bug, fixed in subsequent PCRE releases).
3418 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3419 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3421 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3427 SC/01 Use a glob alias rather than an array ref in eximstats generated
3428 parser. This improves both readability and performance.
3430 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3431 Don't display local sender or destination tables in eximstats unless
3432 there is data to show.
3433 Added average volumes into the eximstats top table text output.
3435 SC/03 Collect data on the number of addresses (recipients) as well
3436 as the number of messages in eximstats.
3438 TF/01 Correct an error in the documentation for the redirect router. Exim
3439 does (usually) call initgroups() when daemonizing.
3441 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3442 with consistent privilege compared to when running as a daemon.
3444 TF/03 Note in the spec that $authenticated_id is not set for local
3445 submissions from trusted users.
3447 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3448 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3450 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3451 by adding some example configuration directives to the default
3452 configuration file. A little bit of work is required to uncomment the
3453 directives and define how usernames and passwords are checked, but
3454 there is now a framework to start from.
3456 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3457 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3458 without this. I don't know how relevant this is to other LDAP libraries.
3460 PH/02 Add the verb name to the "unknown ACL verb" error.
3462 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3464 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3466 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3467 directory not expanded when it should be if an expanded home directory
3468 was set for the address (which is overridden by the transport).
3470 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3473 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3474 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3475 because it is too late at that time, and has no effect.
3477 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3478 security issue with \' (bugzilla #107). I could not use the
3479 PQescapeStringConn() function, because it needs a PGconn value as one of
3482 PH/08 When testing addresses using -bt, indicate those final addresses that
3483 are duplicates that would not cause an additional delivery. At least one
3484 person was confused, thinking that -bt output corresponded to deliveries.
3485 (Suppressing duplicates isn't a good idea as you lose the information
3486 about possibly different redirections that led to the duplicates.)
3488 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3489 systems where poll() doesn't work, in particular OS X.
3491 PH/10 Added more information to debugging output for retry time not reached.
3493 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3494 operations in malware.c.
3496 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3499 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3500 syslog_duplication set false, log lines that would normally be written
3501 both the the main log and to the reject log were not written to syslog at
3504 PH/14 In the default configuration, change the use of "message" in ACL warn
3505 statements to "add_header".
3507 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3508 not followed by a command (e.g. "seen endif").
3510 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3511 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3514 PH/17 Added extra conditions to the default value of delay_warning_condition
3518 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3519 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3520 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3523 The Auto-Submitted: and various List- headers are standardised, whereas I
3524 don't think Precedence: ever was.
3526 PH/18 Refactored debugging code in route_finduser() to show more information,
3527 in particular, the error code if getpwnam() issues one.
3529 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3530 This is apparently needed in addition to the PH/07 change above to avoid
3531 any possible encoding problems.
3533 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3534 but not after initializing Perl.
3536 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3537 output them only if debugging. By default they are written stderr,
3538 apparently, which is not desirable.
3540 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3543 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3546 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3548 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3549 authenticated or an ident call has been made. Suppress the default
3550 values for $authenticated_id and $authenticated_sender (but permit -oMai
3551 and -oMas) when testing with -bh.
3553 PH/24 Re-jigged the order of the tests in the default configuration so that the
3554 tests for valid domains and recipients precede the DNS black list and CSA
3555 tests, on the grounds that those ones are more expensive.
3557 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3558 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3559 This bug exists in every version of Exim that I still have, right back to
3562 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3563 However, an attempt to turn on case-sensitivity in a regex key by
3564 including (?-i) didn't work because the subject string was already
3565 lowercased, and the effects were non-intuitive. It turns out that a
3566 one-line patch can be used to allow (?-i) to work as expected.
3572 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3573 other effects) broke the use of negated acl sub-conditions.
3575 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3578 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3579 "Deny" causes Exim to reject the incoming connection with a 554 error.
3580 Unfortunately, if there is a major crisis, such as a disk failure,
3581 tcp-wrappers gives "deny", whereas what one would like would be some
3582 kind of temporary error. A kludge has been added to help with this.
3583 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3584 554 error is used if errno is still zero or contains ENOENT (which occurs
3585 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3588 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3590 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3591 errors. Otherwise a message that provokes a temporary error (when other
3592 messages do not) can cause a whole host to time out.
3594 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3595 addresses were routed directly to files or pipes from a redirect router.
3596 File deliveries just didn't batch; pipe deliveries might have suffered
3599 PH/06 A failure to get a lock for a hints database would erroneously always say
3600 "Failed to get write lock", even when it was really a read lock.
3602 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3603 of 0600. This has been changed to use the value of the lockfile_mode
3604 option (which defaults to 0600).
3606 PH/08 Applied small patch from the Sieve maintainer.
3608 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3609 folder from quota calculations, a direct delivery into this folder messed
3610 up the contents of the maildirsize file. This was because the regex was
3611 used only to exclude .Trash (or whatever) when the size of the mailbox
3612 was calculated. There was no check that a delivery was happening into an
3613 excluded directory. This bug has been fixed by ignoring all quota
3614 processing for deliveries into excluded directories.
3616 PH/10 Added the maildirfolder_create_regex option to appendfile.
3622 PH/01 The code for finding all the local interface addresses on a FreeBSD
3623 system running IPv6 was broken. This may well have applied to all BSD
3624 systems, as well as to others that have similar system calls. The broken
3625 code found IPv4 interfaces correctly, but gave incorrect values for the
3626 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3627 that it would not match correctly against @[] and not recognize the IPv6
3630 PH/02 The ipliteral router was not recognizing addresses of the form user@
3631 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3633 PH/03 Added disable_ipv6.
3635 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3636 decoded form, because it is most often used to construct To: headers
3637 lines in autoreplies, and the decoded form may well be syntactically
3638 invalid. However, $reply_address has leading white space removed, and all
3639 newlines turned into spaces so that the autoreply transport does not
3642 PH/05 If group was specified without a user on a router, and no group or user
3643 was specified on a transport, the group from the router was ignored.
3645 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3646 for visible compile-time settings that can be used to change these
3647 numbers, for those that want even more. Backwards compatibility with old
3648 spool files has been maintained. However, going back to a previous Exim
3649 release will lost any variables that are in spool files.
3651 PH/07 Two small changes when running in the test harness: increase delay when
3652 passing a TCP/IP connection to a new process, in case the original
3653 process has to generate a bounce, and remove special handling of
3654 127.0.0.2 (sic), which is no longer necessary.
3656 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3657 be the same on different OS.
3659 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3662 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3663 whether --show-vars was specified or not
3665 JJ/02 exipick: Added support for new ACL variable spool format introduced
3668 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3669 syntactically invalid From: or Reply-to: line, and a filter used this to
3670 generate an autoreply, and therefore failed to obtain an address for the
3671 autoreply, Exim could try to deliver to a non-existent relative file
3672 name, causing unrelated and misleading errors. What now happens is that
3673 it logs this as a hard delivery error, but does not attempt to create a
3676 PH/11 The exinext utility has a -C option for testing purposes, but although
3677 the given file was scanned by exinext itself; it wasn't being passed on
3678 when Exim was called.
3680 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3681 an end-of-file indication when reading a command response.
3683 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3684 compiled. In many other places in Exim, IPv6 addresses are always
3685 recognized, so I have changed this. It also means that IPv4 domain
3686 literals of the form [IPV4:n.n.n.n] are now always recognized.
3688 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3689 used if the router is not running as root, for example, when verifying at
3690 ACL time, or when using -bh. The debugging output from this situation was
3691 non-existent - all you got was a failure to exec. I have made two
3694 (a) Failures to set uid/gid, the current directory, or a process leader
3695 in a subprocess such as that created by queryprogram now generate
3696 suitable debugging output when -d is set.
3698 (b) The queryprogram router detects when it is not running as root,
3699 outputs suitable debugging information if -d is set, and then runs
3700 the subprocess without attempting to change uid/gid.
3702 PH/15 Minor change to Makefile for building test_host (undocumented testing
3705 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3706 additional section of a DNS packet that returns MX or SRV records.
3707 Instead, it always explicitly searches for A/AAAA records. This avoids
3708 major problems that occur when a DNS server includes only records of one
3709 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3710 fixed another bug: if SRV records were looked up and the corresponding
3711 address records were *not* found in the additional section, the port
3712 values from the SRV records were lost.
3714 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3715 using the correct key (the original address) when searching the retry
3716 rules in order to find which one to use for generating the retry hint.
3718 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3719 adding the default one. Similarly, if it contains a Reply-To: header, the
3720 errors_reply_to option, if set, is not used.
3722 PH/19 When calculating a retry time, Exim used to measure the "time since
3723 failure" by looking at the "first failed" field in the retry record. Now
3724 it does not use this if it is later than than the arrival time of the
3725 message. Instead it uses the arrival time. This makes for better
3726 behaviour in cases where some deliveries succeed, thus re-setting the
3727 "first failed" field. An example is a quota failure for a huge message
3728 when small messages continue to be delivered. Without this change, the
3729 "time since failure" will always be short, possible causing more frequent
3730 delivery attempts for the huge message than are intended.
3731 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3733 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3734 $1, $2, $3) because the numerical variables can be reset during some
3735 expansion items (e.g. "match"), thereby losing the authentication data.
3737 PH/21 Make -bV show the size of off_t variables so that the test suite can
3738 decide whether to run tests for quotas > 2G.
3740 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3741 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3742 filecount value is greater than 2G or if a quota value is greater than 2G
3743 on a system where the size of off_t is not greater than 4, a panic error
3746 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3747 never match. The debug and -bh output now contains an explicit error
3748 message indicating a malformed IPv4 address or mask.
3750 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3751 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3752 PH/23 above applies.
3754 PH/25 Do not write to syslog when running in the test harness. The only
3755 occasion when this arises is a failure to open the main or panic logs
3756 (for which there is an explicit test).
3758 PH/26 Added the /no_tell option to "control=freeze".
3760 PH/27 If a host name lookup failed very early in a connection, for example, if
3761 the IP address matched host_lookup and the reverse lookup yielded a name
3762 that did not have a forward lookup, an error message of the form "no IP
3763 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3764 could be logged. Now it outputs the IP address instead of "NULL".
3766 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3767 allows the sender and the authenticated sender to be set when
3768 submitting a message from within Exim. Since child_open_exim() is
3769 documented for local_scan(), the new function should be too.
3771 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3772 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3773 results in an empty string is now treated as unset.
3775 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3777 PH/31 Added sender_verify_fail logging option.
3779 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3780 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3781 tidied the source and removed it altogether.
3783 PH/33 When a queue run was abandoned because the load average was too high, a
3784 log line was always written; now it is written only if the queue_run log
3785 selector is set. In addition, the log line for abandonment now contains
3786 information about the queue run such as the pid. This is always present
3787 in "start" and "stop" lines but was omitted from the "abandon" line.
3789 PH/34 Omit spaces between a header name and the colon in the error message that
3790 is given when verify = headers_syntax fails (if there are lots of them,
3791 the message gets confusing).
3793 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3794 names, as there are now some PTR records that contain slashes. This check
3795 is only to protect against broken name servers that fall over on strange
3796 characters, so the fact that it applies to all lookups doesn't matter.
3798 PH/36 Now that the new test suite is complete, we can remove some of the
3799 special code in Exim that was needed for the old test suite. For example,
3800 sorting DNS records because real resolvers return them in an arbitrary
3801 order. The new test suite's fake resolver always returns records in the
3804 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3805 bounces) except when queue_only is set, to avoid logging races between
3806 the different processes.
3808 PH/38 Panic-die if .include specifies a non-absolute path.
3810 PH/39 A tweak to the "H" retry rule from its user.
3812 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3813 a label. They prevented compilation on older perls.
3815 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3816 a warning to be raised on newish perls.
3818 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3819 on queue. Changes to match documented behaviour of showing count of
3820 messages matching specified criteria.
3822 PH/40 Changed the default ident timeout from 30s to 5s.
3824 PH/41 Added support for the use of login_cap features, on those BSD systems
3825 that have them, for controlling the resources used by pipe deliveries.
3827 PH/42 The content-scanning code uses fopen() to create files in which to put
3828 message data. Previously it was not paying any attention to the mode of
3829 the files. Exim runs with umask(0) because the rest of the code creates
3830 files with open(), and sets the required mode explicitly. Thus, these
3831 files were ending up world-writeable. This was not a big issue, because,
3832 being within the spool directory, they were not world-accessible. I have
3833 created a function called modefopen, which takes an additional mode
3834 argument. It sets umask(777), creates the file, chmods it to the required
3835 mode, then resets the umask. All the relevant calls to fopen() in the
3836 content scanning code have been changed to use this function.
3838 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3839 to 24 hours. This avoids potential overflow problems when processing G
3840 and H retry rules. I suspect nobody ever tinkers with this value.
3842 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3844 PH/45 When the plaintext authenticator is running as a client, the server's
3845 challenges are checked to ensure they are valid base64 strings. By
3846 default, the authentication attempt is cancelled if an invalid string is
3847 received. Setting client_ignore_invalid_base64 true ignores these errors.
3848 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3849 they are received. Thus, the responses can be made to depend on the
3850 challenges. If an invalid string is ignored, an empty string is placed in
3853 PH/46 Messages that are created by the autoreply transport now contains a
3854 References: header, in accordance with RFCs 2822 and 3834.
3856 PH/47 Added authenticated_sender_force to the smtp transport.
3858 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3860 PH/49 Installed latest patch from the Sieve maintainer.
3862 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3863 was also set, the appendfile transport was unnecessarily scanning a
3864 directory of message files (e.g. for maildir delivery) to find the count
3865 of files (along with the size), even though it did not need this
3866 information. It now does the scan only if it needs to find either the
3867 size of the count of files.
3869 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3871 PH/52 Two bugs concerned with error handling when the smtp transport is
3874 (i) Exim was not creating retry information for temporary errors given
3875 for individual recipients after the DATA command when the smtp transport
3876 was used in LMTP mode. This meant that they could be retried too
3877 frequently, and not timed out correctly.
3879 (ii) Exim was setting the flag that allows error details to be returned
3880 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3881 recipients that were returned after the DATA command.
3883 PH/53 This is related to PH/52, but is more general: for any failing address,
3884 when detailed error information was permitted to be returned to the
3885 sender, but the error was temporary, then after the final timeout, only
3886 "retry timeout exceeded" was returned. Now it returns the full error as
3887 well as "retry timeout exceeded".
3889 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3890 do this, and (what is worse) MTAs that accept it.
3892 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3893 will now be deprecated.
3895 PH/56 New os.c-cygwin from the Cygwin maintainer.
3897 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3898 formats (previously only available in exim formats via -bpr, -bpru,
3899 and -bpra. Now also available in native and exiqgrep formats)
3901 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3902 with very large, slow to parse queues
3904 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3906 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3908 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3909 responses to authentication challenges, though it was showing the
3910 challenges; (ii) I've removed the CR characters from the debug output for
3913 PH/58 Allow for the insertion of a newline as well as a space when a string
3914 is turned into more than one encoded-word during RFC 2047 encoding. The
3915 Sieve code now uses this.
3917 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3918 data_4xx, lost_connection, tls_required.
3920 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3921 message was being sent as an SMTP response.
3923 PH/61 Add -l and -k options to exicyclog.
3925 PH/62 When verifying, if an address was redirected to one new address, so that
3926 verification continued, and the new address failed or deferred after
3927 having set something in $address_data, the value of $address_data was not
3928 passed back to the ACL. This was different to the case when no
3929 redirection occurred. The value is now passed back in both cases.
3931 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3932 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3933 use login_cap.h, so on its own it isn't the distinguishing feature. The
3934 new name refers directly to the setclassresources() function.
3936 PH/65 Added configuration files for NetBSD3.
3938 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3940 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3941 is preferred over IPv4.
3943 PH/68 The bounce_return_message and bounce_return_body options were not being
3944 honoured for bounces generated during the reception of non-SMTP messages.
3945 In particular, this applied to messages rejected by the ACL. This bug has
3946 been fixed. However, if bounce_return_message is true and bounce_return_
3947 body is false, the headers that are returned for a non-SMTP message
3948 include only those that have been read before the error was detected.
3949 (In the case of an ACL rejection, they have all been read.)
3951 PH/69 The HTML version of the specification is now built in a directory called
3952 spec_html instead of spec.html, because the latter looks like a path with
3953 a MIME-type, and this confuses some software.
3955 PH/70 Catch two compiler warnings in sieve.c.
3957 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3958 function verify_get_ident() calls ip_connect() to connect a socket, but
3959 if the "connect()" function timed out, ip_connect() used to close the
3960 socket. However, verify_get_ident() also closes the socket later, and in
3961 between Exim writes to the log, which may get opened at this point. When
3962 the socket was closed in ip_connect(), the log could get the same file
3963 descriptor number as the socket. This naturally causes chaos. The fix is
3964 not to close the socket in ip_connect(); the socket should be closed by
3965 the function that creates it. There was only one place in the code where
3966 this was missing, in the iplookup router, which I don't think anybody now
3967 uses, but I've fixed it anyway.
3969 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3970 well as to direct DNS lookups. Otherwise the handling of names in host
3971 lists is inconsistent and therefore confusing.
3977 PH/01 Two changes to the default runtime configuration:
3979 (1) Move the checks for relay_from_hosts and authenticated clients from
3980 after to before the (commented out) DNS black list checks.
3982 (2) Add control=submission to the relay_from_hosts and authenticated
3983 clients checks, on the grounds that messages accepted by these
3984 statements are most likely to be submissions.
3986 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3988 (1) Generate an error if the third argument for the ${prvs expansion is
3991 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3994 (3) Reset the variables that are obtained from the first argument of
3995 ${prvscheck and used in the second argument before leaving the code,
3996 because their memory is reclaimed, so using them afterwards may do
3999 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4000 one (it's much easier than Tom thought :-).
4002 (5) Because of (4), we can now allow for the use of $prvscheck_result
4003 inside the third argument.
4005 PH/03 For some reason, the default setting of PATH when running a command from
4006 a pipe transport was just "/usr/bin". I have changed it to
4009 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4010 anything to be listed in the output from -bV.
4012 PH/05 When a filter generated an autoreply, the entire To: header line was
4013 quoted in the delivery log line, like this:
4015 => >A.N.Other <ano@some.domain> <original@ddress> ...
4017 This has been changed so that it extracts the operative address. There
4018 may be more than one such address. If so, they are comma-separated, like
4021 => >ano@some.domain,ona@other.domain <original@ddress> ...
4023 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4024 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4025 not being looked up in the rDNS to get a host name, Exim was showing the
4026 IP address twice in Received: lines, even though the IP addresses were
4027 identical. For example:
4029 Received: from [1.2.3.4] (helo=[1.2.3.4])
4031 However, if the real host name was known, it was omitting the HELO data
4032 if it matched the actual IP address. This has been tidied up so that it
4033 doesn't show the same IP address twice.
4035 PH/07 When both +timestamp and +memory debugging was on, the value given by
4036 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4037 called by the debug printing, thereby overwriting the timestamp buffer.
4038 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4040 PH/08 When the original message was included in an autoreply transport, it
4041 always said "this is a copy of the message, including all the headers",
4042 even if body_only or headers_only was set. It now gives an appropriate
4045 PH/09 Applied a patch from the Sieve maintainer which:
4047 o fixes some comments
4048 o adds the (disabled) notify extension core
4049 o adds some debug output for the result of if/elsif tests
4050 o points to the current vacation draft in the documentation
4051 and documents the missing references header update
4055 o fixes a bug in processing the envelope test (when testing
4056 multiple envelope elements, the last element determined the
4059 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4060 Electronic Mail") by including:
4062 Auto-submitted: auto-generated
4064 in the messages that it generates (bounce messages and others, such as
4065 warnings). In the case of bounce messages for non-SMTP messages, there was
4066 also a typo: it was using "Auto_submitted" (underscore instead of
4067 hyphen). Since every message generated by Exim is necessarily in response
4068 to another message, thes have all been changed to:
4070 Auto-Submitted: auto-replied
4072 in accordance with these statements in the RFC:
4074 The auto-replied keyword:
4076 - SHOULD be used on messages sent in direct response to another
4077 message by an automatic process,
4079 - MUST NOT be used on manually-generated messages,
4081 - MAY be used on Delivery Status Notifications (DSNs) and Message
4082 Disposition Notifications (MDNs),
4084 - MUST NOT be used on messages generated by automatic or periodic
4085 processes, except for messages which are automatic responses to
4088 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4089 to the default Received: header definition.
4091 PH/12 Added log selector acl_warn_skipped (default on).
4093 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4094 variables because (a) they are in the wrong storage pool and (b) even if
4095 they were copied, it wouldn't work properly because of the caching.
4097 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4098 checking when decoding. Apparently there are clients that generate
4099 overlong encoded strings. Why am I not surprised?
4101 PH/15 If the first argument of "${if match_address" was not empty, but did not
4102 contain an "@" character, Exim crashed. Now it writes a panic log message
4103 and treats the condition as false.
4105 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4107 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4108 extension "envelope-auth". The code is finished and in agreement with
4109 other implementations, but there is no documentation so far and in fact,
4110 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4111 not changing the active code.
4113 Print executed "if" and "elsif" statements when debugging is used. This
4114 helps a great deal to understand what a filter does.
4116 Document more things not specified clearly in RFC3028. I had all this
4117 sorted out, when out of a sudden new issues came to my mind. Oops."
4119 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4122 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4123 canonical form (as documented). However, after a host name lookup from
4124 the IP address, check_host() was doing a simple string comparison with
4125 addresses acquired from the DNS when checking that the found name did
4126 have the original IP as one of its addresses. Since any found IPv6
4127 addresses are likely to be in abbreviated form, the comparison could
4128 fail. Luckily, there already exists a function for doing the comparison
4129 by converting both addresses to binary, so now that is used instead of
4130 the text comparison.
4132 PH/20 There was another similar case to PH/19, when a complete host name was
4133 given in a host list; looking up its IP address could give an abbreviated
4134 form, whereas the current host's name might or might not be abbreviated.
4135 The same fix has been applied.
4141 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4142 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4145 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4146 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4148 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4150 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4151 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4152 Also, for "Auto-Submitted", treat anything other than "no" as
4153 non-personal, in accordance with RFC 3834. (Previously it treated
4154 anything starting "auto-" as non-personal.)
4156 TF/01 The control=submission/name=... option had a problem with syntax
4157 errors if the name included a slash character. The /name= option
4158 now slurps the rest of the string, so it can include any characters
4159 but it must come last in the list of options (after /sender_retain
4162 PH/05 Some modifications to the interface to the fake nameserver for the new
4170 TK/01 Added the "success_on_redirect" address verification option. See
4171 NewStuff for rationale and an example.
4173 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4175 PH/02 Patch to exigrep to allow it to work on syslog lines.
4177 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4178 fread() to skip over the body file's header line, because in Cygwin the
4179 header line is locked and is inaccessible.
4181 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4182 co-exist for some time) to make it clear that it is the Exim ID that is
4183 referenced, not the Message-ID: header line.
4185 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4186 string_format() function, because snprintf() does not exist on all
4189 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4190 ${stat: expansion item.
4192 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4193 protocol synchronization error", to keep the pedants happy.
4195 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4196 well as for IRIX systems, when gcc is being used. See the host.c source
4199 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4201 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4204 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4205 transport and to the smtp transport in LMTP mode.
4207 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4209 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4210 run a filter in a subprocess. This could lead to confusion in subsequent
4211 lookups in the parent process. There should also be a search_tidyup() at
4212 the end of the subprocess.
4214 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4215 only if the host matched helo_try_verify_hosts, which caused the
4216 verification to occur when the EHLO/HELO command was issued. The ACL just
4217 tested the remembered result. Now, if a previous verification attempt has
4218 not happened, "verify = helo" does it there and then.
4220 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4222 TK/03 Fix log output including CR from clamd.
4224 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4225 exist provoked a memory error which could cause a segfault.
4227 PH/15 Installed PCRE 6.2
4229 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4231 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4232 of the problem. Specifically, suggested +O2 rather than +O1 for the
4235 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4237 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4238 record, the port was not correctly logged when the outgoing_port log
4239 selector was set (it logged the transort's default port).
4241 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4242 fallback_hosts, and "hosts" in the smtp transport.
4244 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4245 host errors such as "Connection refused".
4247 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4248 authentication with radiusclient 0.4.9:
4250 - Error returned from rc_read_config was caught wrongly
4251 - Username/password not passed on to radius server due to wrong length.
4253 The presumption is that some radiusclient API changes for 4.51/PH/17
4254 were not taken care of correctly. The code is still untested by me (my
4255 Linux distribution still has 0.3.2 of radiusclient), but it was
4256 contributed by a Radius user.
4258 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4259 expanding the "port" option of the smtp transport.
4261 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4262 while reading a MIME header. Thanks to Tom Hughes for a patch.
4264 PH/24 Include config.h inside local_scan.h so that configuration settings are
4267 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4268 that in an ACL for RCPT (for example), you can examine exactly what was
4271 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4272 commands, but it was not correctly comparing the address with the actual
4273 client host address. Thus, it would show the EHLO address in Received:
4274 header lines when this was not necessary.
4276 PH/27 Added the % operator to ${eval:}.
4278 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4279 it should be ignoring failures (because with -C, for example, it has lost
4280 privilege). It wasn't ignoring creation failures other than "already
4283 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4286 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4287 because an input line was too long, either on its own, or by virtue of
4288 too many continuations, the temporary file was not being removed, and the
4289 return code was incorrect.
4291 PH/31 Missing "BOOL" in function definition in filtertest.c.
4293 PH/32 Applied Sieve patches from the maintainer.
4295 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4297 PH/33 Added "verify = not_blind".
4299 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4300 Local/Makefile (with some defaults set). These are used in built scripts
4301 such as exicyclog, but they have never been used in the exim_install
4302 script (though there are many overriding facilities there). I have
4303 arranged that the exim_install script now takes note of these two
4306 PH/35 Installed configuration files for Dragonfly.
4308 PH/36 When a locally submitted message by a trusted user did not contain a
4309 From: header, and the sender address was obtained from -f or from an SMTP
4310 MAIL command, and the trusted user did not use -F to supply a sender
4311 name, $originator_name was incorrectly used when constructing a From:
4312 header. Furthermore, $originator_name was used for submission mode
4313 messages from external hosts without From: headers in a similar way,
4314 which is clearly wrong.
4316 PH/37 Added control=suppress_local_fixups.
4318 PH/38 When log_selector = +received_sender was set, and the addition of the
4319 sender made the log line's construction buffer exactly full, or one byte
4320 less than full, an overflow happened when the terminating "\n" was
4323 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4324 when the result of a list match is failure because a DNS lookup failed.
4326 PH/40 RM_COMMAND is now used in the building process.
4328 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4329 the "build-* directories that it finds.
4331 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4332 address in a domain literal was a prefix of an interface address.
4334 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4335 when verifying a sender address, unless rewrite_headers is false.
4337 PH/44 Wrote a long comment about why errors_to addresses are verified as
4338 recipients, not senders.
4340 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4341 the ratelimit ACL was added.
4343 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4345 PH/46 Added extra information about PostgreSQL errors to the error string.
4347 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4348 avoiding the need to install special zones in a real server. This is
4349 backwards compatible; if it can't find the fake resolver, it drops back.
4350 Thus, both old and new test suites can be run.
4352 TF/02 Added util/ratelimit.pl
4354 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4355 clock is set back in time.
4357 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4358 Candler <B.Candler@pobox.com>.
4360 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4361 OK for addresses that are the result of redirections.
4363 PH/48 A number of further additions for the benefit of the new test suite,
4364 including a fake gethostbyname() that interfaces to the fake DNS resolver
4367 TF/06 The fix for widen_domains has also been applied to qualify_single and
4368 search_parents which are the other dnslookup options that can cause
4371 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4374 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4376 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4377 service" if it encountered an MX record with an empty target hostname.
4378 The message is now "an MX or SRV record indicated no SMTP service".
4380 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4381 if the DNS of the sending site is misconfigured. This is quite a
4382 common situation. This change restores the behaviour of treating a
4383 helo verification defer as a failure.
4385 PH/51 If self=fail was set on a router, the bounce message did not include the
4386 actual error message.
4392 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4394 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4395 command itself ended in error, the underlying message about the transport
4396 filter timeout was being overwritten with the pipe command error. Now the
4397 underlying error message should be appended to the second error message.
4399 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4401 PH/02 Reduce the amount of output that "make" produces by default. Full output
4402 can still be requested.
4404 PH/03 The warning log line about a condition test deferring for a "warn" verb
4405 was being output only once per connection, rather than after each
4406 occurrence (because it was using the same function as for successful
4407 "warn" verbs). This seems wrong, so I have changed it.
4409 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4410 it should not have, which might have caused a crash in the right
4411 circumstances, but probably never did.
4413 PH/04 Installed a modified version of Tony Finch's patch to make submission
4414 mode fix the return path as well as the Sender: header line, and to
4415 add a /name= option so that you can make the user's friendly name appear
4418 TF/03 Added the control = fakedefer ACL modifier.
4420 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4421 Mark Lowes for thorough testing.
4423 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4425 TK/03 Merged latest SRS patch from Miles Wilton.
4427 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4428 with the definition in sysexits.h (which is #included earlier).
4429 Fortunately, Exim does not actually use EX_OK. The code used to try to
4430 preserve the sysexits.h value, by assuming that macro definitions were
4431 scanned for macro replacements. I have been disabused of this notion,
4432 so now the code just undefines EX_OK before #including unistd.h.
4434 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4435 in the smtp transport. When a block could not be written in a single
4436 write() function, the timeout was being re-applied to each part-write.
4437 This seems wrong - if the receiver was accepting one byte at a time it
4438 would take for ever. The timeout is now adjusted when this happens. It
4439 doesn't have to be particularly precise.
4441 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4442 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4444 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4445 without a domain if the check to <postmaster@domain> fails.
4447 SC/01 Eximstats: added -xls and the ability to specify output files
4448 (patch written by Frank Heydlauf).
4450 SC/02 Eximstats: use FileHandles for outputting results.
4452 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4454 SC/04 Eximstats: fixed display of large numbers with -nvr option
4456 SC/05 Eximstats: fixed merging of reports with empty tables.
4458 SC/06 Eximstats: added the -include_original_destination flag
4460 SC/07 Eximstats: removed tabs and trailing whitespace.
4462 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4464 TK/06 MBOX spool code: Add real "From " MBOX separator line
4465 so the .eml file is really in mbox format (even though
4466 most programs do not really care). Patch from Alex Miller.
4468 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4469 The latter is generated from $received_to and is only set if the
4470 message has one envelope recipient. SA can use these headers,
4471 obviously out-of-the-box. Patch from Alex Miller.
4473 PH/08 The ${def test on a variable was returning false if the variable's
4474 value was "0", contrary to what the specification has always said!
4475 The result should be true unless the variable is empty.
4477 PH/09 The syntax error of a character other than { following "${if
4478 def:variable_name" (after optional whitespace) was not being diagnosed.
4479 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4480 accidental colon was present, for example, could give incorrect results.
4482 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4483 result is used (not including appendfile, where other changes are about
4486 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4487 This involved changing a lot of size variables from int to off_t. It
4488 should work with maildirs and everything.
4490 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4491 spamd dying while we are connected to it.
4493 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4496 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4497 function for BDB 4.3.
4499 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4501 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4502 the sources for PCRE have been re-arranged and more files are now
4505 PH/15 The code I had for printing potentially long long variables in PH/11
4506 above was not the best (it lost precision). The length of off_t variables
4507 is now inspected at build time, and an appropriate printing format (%ld
4508 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4509 to be "long long int" or "long int". This is needed for the internal
4510 formatting function string_vformat().
4512 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4513 the configuration file to be ":syslog", then the script "guesses" where
4514 the logs files are, rather than using the compiled in default. In our
4515 case the guess is not the same as the compiled default, so the script
4516 suddenly stopped working when I started to use syslog. The patch checks
4517 to see if log_file_path is "". If so, it attempts to read it from exim
4518 with no configuration file to get the compiled in version, before it
4519 falls back to the previous guessing code."
4521 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4522 implementing BATV in an Exim configuration. See NewStuff for the gory
4525 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4526 Makefile that are specific to HP-UX.
4528 PH/18 If the "use_postmaster" option was set for a recipient callout together
4529 with the "random" option, the postmaster address was used as the MAIL
4530 FROM address for the random test, but not for the subsequent recipient
4531 test. It is now used for both.
4533 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4534 patch removes a few documentation additions to RFC 3028, because the
4535 latest draft now contains them. It adds the new en;ascii-case comparator
4536 and a new error check for 8bit text in MIME parts. Comparator and
4537 require names are now matched exactly. I enabled the subaddress
4538 extension, but it is not well tested yet (read: it works for me)."
4540 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4541 rework some of the code of TK/09 above to avoid the hardwired use of
4542 "%lld" and "long long". Replaced the call to snprintf() with a call to
4545 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4546 records point to non-existent hosts", "retry timeout exceeded", and
4547 "retry time not reached for any host after a long failure period".
4549 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4550 experimental DomainKeys support:
4552 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4553 (2) On an error such as an illegally used "control", the wrong name for
4554 the control was given.
4556 These problems did NOT occur unless DomainKeys support was compiled.
4558 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4560 PH/24 Added ${if match_ip condition.
4562 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4563 databases so that it will be absolutely obvious if a crash occurs in the
4564 DB library. This is a regular occurrence (often caused by mis-matched
4567 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4568 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4569 picked up on a user's system that detects such things. There doesn't seem
4570 to be a gcc warning option for this - only an attribute that has to be
4571 put on the function's prototype. It seems that in Fedora Core 4 they have
4572 set this on a number of new functions. No doubt there will be more in due
4575 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4576 specify a transport. However, if an address that was verified by such a
4577 router was the subject of a callout, Exim crashed because it tried to
4578 read the rcpt_include_affixes from the non-existent transport. Now it
4579 just assumes that the setting of that option is false. This bug was
4580 introduced by 4.51/PH/31.
4582 PH/28 Changed -d+all to exclude +memory, because that information is very
4583 rarely of interest, but it makes the output a lot bigger. People tend to
4584 do -d+all out of habit.
4586 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4587 code in os-type was giving problems when libc.so lives in lib64, like on
4590 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4591 aren't the modern standard, and it seems that some systems' include files
4592 don't always have them. Exim was already checking for some of the newer
4593 ones like T_AAAA, and defining it itself. I've added checks for all the
4594 record types that Exim uses.
4596 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4597 not automatically generating a new one, as it is supposed to. This
4598 prevented TLS from working. If the file did exist, but contained invalid
4599 data, a new version was generated, as expected. It was only the case of a
4600 non-existent file that was broken.
4602 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4603 with a change in libdomainkeys > 0.64.
4605 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4606 from DNS. If the selector record carries the flag, it now has
4607 precedence over the domain-wide flag.
4609 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4611 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4612 the use of a port name that isn't defined in /etc/services) occurred, the
4613 message was deferred as in a normal delivery, and thus remained on the
4614 spool, instead of being failed because of the mua_wrapper setting. This
4615 is now fixed, and I tidied up some of the mua_wrapper messages at the
4618 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4619 the messages in a hash of arrays rather than using individual hashes.
4620 This is a bit cleaner and results in dramatic memory savings, albeit
4621 at a slight CPU cost.
4623 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4624 as requested by Marc Sherman.
4626 SC/10 Eximstats: added histograms for user specified patterns as requested
4629 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4631 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4632 fopen() in the content-scanning modules that did not already have it.
4638 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4639 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4641 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4643 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4645 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4646 to test Sieve filters that use "vacation".
4648 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4649 that changes the way the GnuTLS parameters are stored in the cache file.
4650 The new format can be generated externally. For backward compatibility,
4651 if the data in the cache doesn't make sense, Exim assumes it has read an
4652 old-format file, and it generates new data and writes a new file. This
4653 means that you can't go back to an older release without removing the
4656 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4657 work if there are any delivery delays because "one_time" forces the
4658 parent to be marked "delivered", so its unseen clone is never tried
4659 again. For this reason, Exim now forbids the simultaneous setting of
4662 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4663 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4664 entry if you want to know the details. The fix, however, neglected to
4665 consider the case when local delivery batching is involved. The test for
4666 "previously delivered" was not happening when checking to see if an
4667 address could be batched with a previous (undelivered) one; under
4668 certain circumstances this could lead to multiple deliveries to the same
4671 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4672 in its include files, and this causes problems building Exim.
4674 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4675 header_syntax) but Exim was just ignoring anything given after a slash.
4676 In particular, this caused confusion with an attempt to use "verify =
4677 reverse_host_lookup/defer_ok". An error is now given when options are
4678 supplied for verify items that do not have them. (Maybe reverse_host_
4679 lookup should have a defer_ok option, but that's a different point.)
4681 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4682 defined by RFC 821) to 2048, because there were problems with some AUTH
4683 commands, and RFC 1869 says the size should be increased for extended
4684 SMTP commands that take arguments.
4686 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4689 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4690 "unknown" error; now it says that the functionality isn't in the binary.
4692 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4693 an address' error message when a string expansion fails (syntax or
4694 whatever). Otherwise the password may appear in the log. Following change
4695 PH/42 below, there is no longer a chance of it appearing in a bounce
4698 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4700 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4701 of its addresses was ever tried. (Bugzilla bug #2.)
4703 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4704 the result incorrectly in the debug output. (It correctly added a newline
4705 to what was transported.)
4707 TF/01 Added $received_time.
4709 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4710 commented out examples of how to interface to a virus scanner and to
4711 SpamAssassin. Also added commented examples of av_scanner and
4712 spamd_address settings.
4714 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4715 and controls are allowed in which ACLs. There were a couple of minor
4716 errors. Some of the entries in the conditions table (which is a table of
4717 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4718 negation of where the condition IS allowed.
4720 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4722 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4723 header file does not have a version number, so I've had to invent a new
4724 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4725 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4726 radiusclient), but it was contributed by a Radius user.
4728 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4729 files or directories, for OpenSSL.
4731 PH/19 When an Exim process that is running as root has to create an Exim log
4732 file, it does so in a subprocess that runs as exim:exim so as to get the
4733 ownership right at creation (otherwise, other Exim processes might see
4734 the file with the wrong ownership). There was no test for failure of this
4735 fork() call, which would lead to the process getting stuck as it waited
4736 for a non-existent subprocess. Forks do occasionally fail when resources
4737 run out. I reviewed all the other calls to fork(); they all seem to check
4740 PH/20 When checking for unexpected SMTP input at connect time (before writing
4741 the banner), Exim was not dealing correctly with a non-positive return
4742 from the read() function. If the client had disconnected by this time,
4743 the result was a log entry for a synchronization error with an empty
4744 string after "input=" when read() returned zero. If read() returned -1
4745 (an event I could not check), uninitialized data bytes were printed.
4746 There were reports of junk text (parts of files, etc) appearing after
4749 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4751 PH/22 Added support for macro redefinition, and (re)definition in between
4752 driver and ACL definitions.
4754 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4755 forgetting to use the resulting value; it was using the unexpanded value.
4757 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4758 hadn't been configured. The fix is from Juergen Kreileder, who
4759 understands it better than I do:
4761 "Here's what I see happening with three configured cyrus_sasl
4762 authenticators configured (plain, login, cram-md5):
4764 On startup auth_cyrus_sasl_init() gets called for each of these.
4765 This means three calls to sasl_listmech() without a specified mech_list.
4766 => SASL tests which mechs of all available mechs actually work
4767 => three warnings about OTP not working
4768 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4770 With the patch, sasl_listmech() also gets called three times. But now
4771 SASL's mech_list option is set to the server_mech specified in the the
4772 authenticator. Or in other words, the answer from sasl_listmech()
4773 gets limited to just the mech you're testing for (which is different
4775 => the return list contains just 'plain' or 'login', 'cram-md5' or
4776 nothing depending on the value of ob->server_mech.
4778 I've just tested the patch: Authentication still works fine,
4779 unavailable mechs specified in the exim configuration are still
4780 caught, and the auth.log warnings about OTP are gone."
4782 PH/25 When debugging is enabled, the contents of the command line are added
4783 to the debugging output, even when log_selector=+arguments is not
4786 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4787 answer is "GNU", and only if the return is "GNU/something" is the answer
4790 PH/27 $acl_verify_message is now set immediately after the failure of a
4791 verification in an ACL, and so is available in subsequent modifiers. In
4792 particular, the message can be preserved by coding like this:
4794 warn !verify = sender
4795 set acl_m0 = $acl_verify_message
4797 Previously, $acl_verify_message was set only while expanding "message"
4798 and "log_message" when a very denied access.
4800 PH/28 Modified OS/os.c-Linux with
4802 -#ifndef OS_LOAD_AVERAGE
4803 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4805 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4806 nomenclature these days.)
4808 PH/29 Installed patch from the Sieve maintainer that adds the options
4809 sieve_useraddress and sieve_subaddress to the redirect router.
4811 PH/30 In these circumstances:
4812 . Two addresses routed to the same list of hosts;
4813 . First host does not offer TLS;
4814 . First host accepts first address;
4815 . First host gives temporary error to second address;
4816 . Second host offers TLS and a TLS session is established;
4817 . Second host accepts second address.
4818 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4819 and peerdn, if requested) that were in fact used only for the second
4822 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4823 attention to any local part prefix or suffix that was matched by the
4824 router that accepted the address. It now behaves in the same way as it
4825 does for delivery: the affixes are removed from the local part unless
4826 rcpt_include_affixes is set on the transport.
4828 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4829 timeout during the DATA phase of an incoming message.
4831 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4832 applied a patch sanctioned by the Sieve maintainer.
4834 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4835 the uid or gid is negative. A case of a negative gid caused this to be
4836 noticed. The fix allows for either to be negative.
4838 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4839 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4842 PH/36 The change PH/12 above was broken. Fixed it.
4844 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4845 the grounds that routing the same address twice would always produce the
4846 same answer. This might have been true once, but it is certainly no
4847 longer true now. Routing a child address may depend on the previous
4848 routing that produced that child. Some complicated redirection strategies
4849 went wrong when messages had multiple recipients, and made Exim's
4850 behaviour dependent on the order in which the addresses were given.
4852 I have moved the duplicate checking until after the routing is complete.
4853 Exim scans the addresses that are assigned to local and remote
4854 transports, and removes any duplicates. This means that more work will be
4855 done, as duplicates will always all be routed, but duplicates are
4856 presumably rare, so I don't expect this is of any significance.
4858 For deliveries to pipes, files, and autoreplies, the duplicate checking
4859 still happens during the routing process, since they are not going to be
4862 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4863 It corrects a timeout issue with spamd. This is Ian's comment: "The
4864 background is that sometimes spamd either never reads data from a
4865 connection it has accepted, or it never writes response data. The exiscan
4866 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4867 blindly assumes that writes won't block so it may never time out."
4869 PH/39 Allow G after quota size as well as K and M.
4871 PH/40 The value set for $authenticated_id in an authenticator may not contain
4872 binary zeroes or newlines because the value is written to log lines and
4873 to spool files. There was no check on this. Now the value is run through
4874 the string_printing() function so that such characters are converted to
4875 printable escape sequences.
4877 PH/41 $message_linecount is a new variable that contains the total number of
4878 lines in the message. Compare $body_linecount, which is the count for the
4881 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4882 bounce and delay warning messages, except in certain special cases, which
4885 (a) An SMTP error message from a remote host;
4886 (b) A message specified in a :fail: redirection;
4887 (c) A message specified in a "fail" command in a system filter;
4888 (d) A message specified in a FAIL return from the queryprogram router;
4889 (e) A message specified by the cannot_route_message router option.
4891 In these cases only, Exim does include the error details in bounce and
4892 warning messages. There are also a few cases where bland messages such
4893 as "unrouteable address" or "local delivery error" are given.
4895 PH/43 $value is now also set for the "else" part of a ${run expansion.
4897 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4898 being worked on, but at least Exim now implements the latest version to
4901 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4902 process to complete was treated as a delivery failure, a timeout while
4903 writing the message to the pipe was logged, but erroneously treated as a
4904 successful delivery. Such timeouts include transport filter timeouts. For
4905 consistency with the overall process timeout, these timeouts are now
4906 treated as errors, giving rise to delivery failures by default. However,
4907 there is now a new Boolean option for the pipe transport called
4908 timeout_defer, which, if set TRUE, converts the failures into defers for
4909 both kinds of timeout. A transport filter timeout is now identified in
4912 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4913 systems where "make" and "gmake" are different, calling "gmake" at top
4914 level broke things. I've arranged for the value of $(MAKE) to be passed
4915 from the Makefile to this script so that it can call the same version of
4919 A note about Exim versions 4.44 and 4.50
4920 ----------------------------------------
4922 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4923 changes of various kinds. As a consequence, a big documentation update was
4924 needed. This delayed the release for rather longer than seemed good, especially
4925 in the light of a couple of (minor) security issues. Therefore, the changes
4926 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4927 release. So 4.44 and 4.50 are in effect two different branches that both start
4930 I have left the 4.50 change log unchanged; it contains all the changes since
4931 4.43. The change log for 4.44 is below; many of its items are identical to
4932 those for 4.50. This seems to be the most sensible way to preserve the
4933 historical information.
4939 1. Minor wording change to the doc/README.SIEVE file.
4941 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4942 computation of the current number of files was incorrect.
4944 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4945 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4948 4. Give more explanation in the error message when the command for a transport
4949 filter fails to execute.
4951 5. There are several places where Exim runs a non-Exim command in a
4952 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4953 was being done only for the command run by the queryprogram router. It is
4954 now done for all such subprocesses. The other cases are: ${run, transport
4955 filters, and the commands run by the lmtp and pipe transports.
4957 6. Added CONFIGURE_GROUP build-time option.
4959 7. Some older OS have a limit of 256 on the maximum number of file
4960 descriptors. Exim was using setrlimit() to set 1000 as a large value
4961 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4962 systems. I've change it so that if it can't get 1000, it tries for 256.
4964 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4965 was an oversight, and furthermore, ever since the addition of extra
4966 controls (e.g. 4.43/32), the checks on when to allow different forms of
4967 "control" were broken. There should now be diagnostics for all cases when a
4968 control that does not make sense is encountered.
4970 9. Added the /retain_sender option to "control=submission".
4972 10. $recipients is now available in the predata ACL (oversight).
4974 11. Tidy the search cache before the fork to do a delivery from a message
4975 received from the command line. Otherwise the child will trigger a lookup
4976 failure and thereby defer the delivery if it tries to use (for example) a
4977 cached ldap connection that the parent has called unbind on.
4979 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4980 of $address_data from the recipient verification was clobbered by the
4981 sender verification.
4983 13. The value of address_data from a sender verification is now available in
4984 $sender_address_data in subsequent conditions in the ACL statement.
4986 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4988 15. Added a new option "connect=<time>" to callout options, to set a different
4991 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4992 was its contents. (It was OK if the option was not defined at all.)
4994 17. A "Completed" log line is now written for messages that are removed from
4995 the spool by the -Mrm option.
4997 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4998 information about exactly what failed.
5000 19. Added -dd to debug only the daemon process.
5002 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5003 handles timeouts, both on the server side and network timeouts. Renamed the
5004 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5006 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5007 It is now set to "smtps".
5009 22. $host_address is now set to the target address during the checking of
5010 ignore_target_hosts.
5012 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5013 being passed; this would have caused $sender_host_name to have been used if
5014 matching the list had actually called for a host name (not very likely,
5015 since this list is usually IP addresses). A host name is now passed as
5018 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5019 code that specifies a non-restarting handler (typically sigaction() in
5020 modern systems) in an attempt to fix a rare and obscure crash bug.
5022 25. Narrowed the window for a race in the daemon that could cause it to ignore
5023 SIGCHLD signals. This is not a major problem, because they are used only to
5024 wake it up if nothing else does.
5026 26. A malformed maildirsize file could cause Exim to calculate negative values
5027 for the mailbox size or file count. Odd effects could occur as a result.
5028 The maildirsize information is now recalculated if the size or filecount
5031 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5032 support for a long time. Removed HAVE_SYS_VFS_H.
5034 28. Installed the latest version of exipick from John Jetmore.
5036 29. In an address list, if the pattern was not a regular expression, an empty
5037 subject address (from a bounce message) matched only if the pattern was an
5038 empty string. Non-empty patterns were not even tested. This was the wrong
5039 because it is perfectly reasonable to use an empty address as part of a
5040 database query. An empty address is now tested by patterns that are
5041 lookups. However, all the other forms of pattern expect the subject to
5042 contain a local part and a domain, and therefore, for them, an empty
5043 address still always fails if the pattern is not itself empty.
5045 30. Exim went into a mad DNS loop when attempting to do a callout where the
5046 host was specified on an smtp transport, and looking it up yielded more
5047 than one IP address.
5049 31. Re-factored the code for checking spool and log partition space into a
5050 function that finds that data and another that does the check. The former
5051 is then used to implement four new variables: $spool_space, $log_space,
5052 $spool_inodes, and $log_inodes.
5054 32. The RFC2047 encoding function was originally intended for short strings
5055 such as real names; it was not keeping to the 75-character limit for
5056 encoded words that the RFC imposes. It now respects the limit, and
5057 generates multiple encoded words if necessary. To be on the safe side, I
5058 have increased the buffer size for the ${rfc2047: expansion operator from
5061 33. It is now permitted to omit both strings after an "if" condition; if the
5062 condition is true, the result is "true". As before, when the second string
5063 is omitted, a false condition yields an empty string. This makes it less
5064 cumbersome to write custom ACL and router conditions.
5066 34. Failure to deliver a bounce message always caused it to be frozen, even if
5067 there was an errors_to setting on the router. The errors_to setting is now
5070 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5071 canonical form (fully expanded) before being placed in
5072 $sender_host_address.
5074 36. The table in the code that translates DNS record types into text (T_A to
5075 "A" for instance) was missing entries for NS and CNAME. It is just possible
5076 that this could have caused confusion if both these types were looked up
5077 for the same domain, because the text type is used as part of Exim's
5078 per-process caching. But the chance of anyone hitting this buglet seems
5081 37. The dnsdb lookup has been extended in a number of ways.
5083 (1) There is a new type, "zns", which walks up the domain tree until it
5084 finds some nameserver records. It should be used with care.
5086 (2) There is a new type, "mxh", which is like "mx" except that it returns
5087 just the host names, not the priorities.
5089 (3) It is now possible to give a list of domains (or IP addresses) to be
5090 looked up. The behaviour when one of the lookups defers can be
5091 controlled by a keyword.
5093 (4) It is now possible to specify the separator character for use when
5094 multiple records are returned.
5096 38. The dnslists ACL condition has been extended: it is now possible to supply
5097 a list of IP addresses and/or domains to be looked up in a particular DNS
5100 39. Added log_selector=+queue_time_overall.
5102 40. When running the queue in the test harness, wait just a tad after forking a
5103 delivery process, to get repeatability of debugging output.
5105 41. Include certificate and key file names in error message when GnuTLS fails
5106 to set them up, because the GnuTLS error message doesn't include the name
5107 of the failing file when there is a problem reading it.
5109 42. Allow both -bf and -bF in the same test run.
5111 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5113 44. The "Exiscan patch" is now merged into the mainline Exim source.
5115 45. Sometimes the final signoff response after QUIT could fail to get
5116 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5117 < 0 before doing a fflush(). This bug looks as though it goes back to the
5118 introduction of TLS in release 3.20, but "sometimes" must have been rare
5119 because the tests only now provoked it.
5121 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5122 (this can affect the format of dates).
5124 47. exim_tidydb, when checking for the continued existence of a message for
5125 which it has found a message-specific retry record, was not finding
5126 messages that were in split spool directories. Consequently, it was
5127 deleting retry records that should have stayed in existence.
5129 48. Steve fixed some bugs in eximstats.
5131 49. The SPA authentication driver was not abandoning authentication and moving
5132 on to the next authenticator when an expansion was forced to fail,
5133 contradicting the general specification for all authenticators. Instead it
5134 was generating a temporary error. It now behaves as specified.
5136 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5137 (the order specifies the preference for clients). The order is now AES256,
5138 AES128, 3DES, ARCFOUR128.
5140 51. Small patch to Sieve code - explicitly set From: when generating an
5143 52. Exim crashed if a remote delivery caused a very long error message to be
5144 recorded - for instance if somebody sent an entire SpamAssassin report back
5145 as a large number of 550 error lines. This bug was coincidentally fixed by
5146 increasing the size of one of Exim's internal buffers (big_buffer) that
5147 happened as part of the Exiscan merge. However, to be on the safe side, I
5148 have made the code more robust (and fixed the comments that describe what
5151 53. Now that there can be additional text after "Completed" in log lines (if
5152 the queue_time_overall log selector is set), a one-byte patch to exigrep
5153 was needed to allow it to recognize "Completed" as not the last thing in
5156 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5157 patch that reportedly fixes this has been added. I am not expert enough to
5158 create a test for it. This is what the patch creator wrote:
5160 "I found a little strange behaviour of ldap code when working with
5161 Windows 2003 AD Domain, where users was placed in more than one
5162 Organization Units. When I tried to give exim partial DN, the exit code
5163 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5164 But simultaneously result of request was absolutely normal ldap result,
5165 so I produce this patch..."
5167 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5168 REFERENCE, so I have modified the code to exclude the patch when that macro
5171 55. Some experimental protocols are using DNS PTR records for new purposes. The
5172 keys for these records are domain names, not reversed IP addresses. The
5173 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5174 leaves it alone. Component reversal etc. now happens only for IP addresses.
5177 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5179 57. Double the size of the debug message buffer (to 2048) so that more of very
5180 long debug lines gets shown.
5182 58. The exicyclog utility now does better if the number of log files to keep
5183 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5185 59. Two changes related to the smtp_active_hostname option:
5187 (1) $smtp_active_hostname is now available as a variable.
5188 (2) The default for smtp_banner uses $smtp_active_hostname instead
5189 of $primary_hostname.
5191 60. The host_aton() function is supposed to be passed a string that is known
5192 to be a valid IP address. However, in the case of IPv6 addresses, it was
5193 not checking this. This is a hostage to fortune. Exim now panics and dies
5194 if the condition is not met. A case was found where this could be provoked
5195 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5196 components; fortuitously, this particular loophole had already been fixed
5197 by change 4.50/55 above.
5199 If there are any other similar loopholes, the new check in host_aton()
5200 itself should stop them being exploited. The report I received stated that
5201 data on the command line could provoke the exploit when Exim was running as
5202 exim, but did not say which command line option was involved. All I could
5203 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5204 running as the user.
5207 61. There was a buffer overflow vulnerability in the SPA authentication code
5208 (which came originally from the Samba project). I have added a test to the
5209 spa_base64_to_bits() function which I hope fixes it.
5212 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5213 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5215 63. The daemon start-up calls getloadavg() while still root for those OS that
5216 need the first call to be done as root, but it missed one case: when
5217 deliver_queue_load_max is set with deliver_drop_privilege. This is
5218 necessary for the benefit of the queue runner, because there is no re-exec
5219 when deliver_drop_privilege is set.
5221 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5222 This has been fixed.
5224 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5225 was in use, was not putting the data itself into the right store pool;
5226 consequently, it could be overwritten for a subsequent message in the same
5227 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5230 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5232 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5233 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5234 the function were treating the return as a boolean value, which happened to
5235 work because 0=false and not-0=true, but is not correct code.
5237 68. The host_aton() function was not handling scoped IPv6 addresses (those
5238 with, for example, "%eth0" on the end) correctly.
5240 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5241 negated items (that is, ~something) in unsigned ints. Some compilers
5242 apparently mutter when there is no cast.
5244 70. If an address verification called from an ACL failed, and did not produce a
5245 user-specific message (i.e. there was only a "system" message), nothing was
5246 put in $acl_verify_message. In this situation, it now puts the system
5249 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5250 session; change 4.31/43 added the unwanted input to the log line - except
5251 that it did not do this in the start of session case. It now does.
5253 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5254 This is wrong and can cause the other end to generate a synchronization
5255 error if it is another Exim or anything else that does the synchronization
5256 check. A QUIT command is no longer sent after a timeout.
5258 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5259 during host lookups.
5261 74. The defer_ok option of callout verification was not working if it was used
5262 when verifying addresses in header lines, that is, for this case:
5264 verify = header_sender/callout=defer_ok
5266 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5267 those file descriptors could be used for SMTP connections. If anything
5268 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5269 could be sent to the SMTP client, causing chaos. The daemon now opens
5270 stdin, stdout, and stderr to /dev/null when it puts itself into the
5273 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5274 log by default. The user can override this with suitable Perl magic.
5276 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5277 the log message when discard triggers, was not working for the DATA ACL or
5278 for the non-SMTP ACL.
5280 78. Error message wording change in sieve.c.
5282 79. If smtp_accept_max_per_host was set, the number of connections could be
5283 restricted to fewer than expected, because the daemon was trying to set up
5284 a new connection before checking whether the processes handling previous
5285 connections had finished. The check for completed processes is now done
5286 earlier. On busy systems, this bug wouldn't be noticed because something
5287 else would have woken the daemon, and it would have reaped the completed
5290 80. If a message was submitted locally by a user whose login name contained one
5291 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5292 It caused a spool format error. I have fixed the spool reading code. A
5293 related problem was that the "from" clause in the Received: line became
5294 illegal because of the space(s). It is now covered by ${quote_local_part.
5296 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5299 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5301 83. Patch from Sieve maintainer to fix unterminated string problem in
5302 "vacation" handling.
5304 84. Some minor changes to the Linux configuration files to help with other
5305 OS variants using glibc.
5307 85. One more patch for Sieve to update vacation handling to latest spec.
5310 ----------------------------------------------------
5311 See the note above about the 4.44 and 4.50 releases.
5312 ----------------------------------------------------
5318 1. Change 4.43/35 introduced a bug that caused file counts to be
5319 incorrectly computed when quota_filecount was set in an appendfile
5322 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5323 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5326 3. Give more explanation in the error message when the command for a transport
5327 filter fails to execute.
5329 4. There are several places where Exim runs a non-Exim command in a
5330 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5331 was being done only for the command run by the queryprogram router. It is
5332 now done for all such subprocesses. The other cases are: ${run, transport
5333 filters, and the commands run by the lmtp and pipe transports.
5335 5. Some older OS have a limit of 256 on the maximum number of file
5336 descriptors. Exim was using setrlimit() to set 1000 as a large value
5337 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5338 systems. I've change it so that if it can't get 1000, it tries for 256.
5340 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5341 was an oversight, and furthermore, ever since the addition of extra
5342 controls (e.g. 4.43/32), the checks on when to allow different forms of
5343 "control" were broken. There should now be diagnostics for all cases when a
5344 control that does not make sense is encountered.
5346 7. $recipients is now available in the predata ACL (oversight).
5348 8. Tidy the search cache before the fork to do a delivery from a message
5349 received from the command line. Otherwise the child will trigger a lookup
5350 failure and thereby defer the delivery if it tries to use (for example) a
5351 cached ldap connection that the parent has called unbind on.
5353 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5354 of $address_data from the recipient verification was clobbered by the
5355 sender verification.
5357 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5358 was its contents. (It was OK if the option was not defined at all.)
5360 11. A "Completed" log line is now written for messages that are removed from
5361 the spool by the -Mrm option.
5363 12. $host_address is now set to the target address during the checking of
5364 ignore_target_hosts.
5366 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5367 being passed; this would have caused $sender_host_name to have been used if
5368 matching the list had actually called for a host name (not very likely,
5369 since this list is usually IP addresses). A host name is now passed as
5372 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5373 code that specifies a non-restarting handler (typically sigaction() in
5374 modern systems) in an attempt to fix a rare and obscure crash bug.
5376 15. Narrowed the window for a race in the daemon that could cause it to ignore
5377 SIGCHLD signals. This is not a major problem, because they are used only to
5378 wake it up if nothing else does.
5380 16. A malformed maildirsize file could cause Exim to calculate negative values
5381 for the mailbox size or file count. Odd effects could occur as a result.
5382 The maildirsize information is now recalculated if the size or filecount
5385 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5386 support for a long time. Removed HAVE_SYS_VFS_H.
5388 18. Updated exipick to current release from John Jetmore.
5390 19. Allow an empty sender to be matched against a lookup in an address list.
5391 Previously the only cases considered were a regular expression, or an
5394 20. Exim went into a mad DNS lookup loop when doing a callout where the
5395 host was specified on the transport, if the DNS lookup yielded more than
5398 21. The RFC2047 encoding function was originally intended for short strings
5399 such as real names; it was not keeping to the 75-character limit for
5400 encoded words that the RFC imposes. It now respects the limit, and
5401 generates multiple encoded words if necessary. To be on the safe side, I
5402 have increased the buffer size for the ${rfc2047: expansion operator from
5405 22. Failure to deliver a bounce message always caused it to be frozen, even if
5406 there was an errors_to setting on the router. The errors_to setting is now
5409 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5410 canonical form (fully expanded) before being placed in
5411 $sender_host_address.
5413 24. Updated eximstats to version 1.33
5415 25. Include certificate and key file names in error message when GnuTLS fails
5416 to set them up, because the GnuTLS error message doesn't include the name
5417 of the failing file when there is a problem reading it.
5419 26. Expand error message when OpenSSL has problems setting up cert/key files.
5422 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5423 (this can affect the format of dates).
5425 28. exim_tidydb, when checking for the continued existence of a message for
5426 which it has found a message-specific retry record, was not finding
5427 messages that were in split spool directories. Consequently, it was
5428 deleting retry records that should have stayed in existence.
5430 29. eximstats updated to version 1.35
5431 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5432 1.35 - bugfix such that pie charts by volume are generated correctly
5434 30. The SPA authentication driver was not abandoning authentication and moving
5435 on to the next authenticator when an expansion was forced to fail,
5436 contradicting the general specification for all authenticators. Instead it
5437 was generating a temporary error. It now behaves as specified.
5439 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5440 (the order specifies the preference for clients). The order is now AES256,
5441 AES128, 3DES, ARCFOUR128.
5443 31. Small patch to Sieve code - explicitly set From: when generating an
5446 32. Exim crashed if a remote delivery caused a very long error message to be
5447 recorded - for instance if somebody sent an entire SpamAssassin report back
5448 as a large number of 550 error lines. This bug was coincidentally fixed by
5449 increasing the size of one of Exim's internal buffers (big_buffer) that
5450 happened as part of the Exiscan merge. However, to be on the safe side, I
5451 have made the code more robust (and fixed the comments that describe what
5454 33. Some experimental protocols are using DNS PTR records for new purposes. The
5455 keys for these records are domain names, not reversed IP addresses. The
5456 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5457 leaves it alone. Component reversal etc. now happens only for IP addresses.
5460 34. The host_aton() function is supposed to be passed a string that is known
5461 to be a valid IP address. However, in the case of IPv6 addresses, it was
5462 not checking this. This is a hostage to fortune. Exim now panics and dies
5463 if the condition is not met. A case was found where this could be provoked
5464 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5465 components; fortuitously, this particular loophole had already been fixed
5466 by change 4.50/55 or 4.44/33 above.
5468 If there are any other similar loopholes, the new check in host_aton()
5469 itself should stop them being exploited. The report I received stated that
5470 data on the command line could provoke the exploit when Exim was running as
5471 exim, but did not say which command line option was involved. All I could
5472 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5473 running as the user.
5476 35. There was a buffer overflow vulnerability in the SPA authentication code
5477 (which came originally from the Samba project). I have added a test to the
5478 spa_base64_to_bits() function which I hope fixes it.
5481 36. The daemon start-up calls getloadavg() while still root for those OS that
5482 need the first call to be done as root, but it missed one case: when
5483 deliver_queue_load_max is set with deliver_drop_privilege. This is
5484 necessary for the benefit of the queue runner, because there is no re-exec
5485 when deliver_drop_privilege is set.
5487 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5488 was in use, was not putting the data itself into the right store pool;
5489 consequently, it could be overwritten for a subsequent message in the same
5490 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5493 38. Sometimes the final signoff response after QUIT could fail to get
5494 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5495 < 0 before doing a fflush(). This bug looks as though it goes back to the
5496 introduction of TLS in release 3.20, but "sometimes" must have been rare
5497 because the tests only now provoked it.
5503 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5504 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5505 Now it returns an integer. A number of calls were still expecting a T/F
5506 return. Fortuitously, in all cases, the tests worked in OK situations,
5507 which is the norm. However, things would have gone wrong on any write
5508 failures on the smtp file descriptor. This function is used when sending
5509 messages over SMTP and also when doing verify callouts.
5511 2. When Exim is called to do synchronous delivery of a locally submitted
5512 message (the -odf or -odi options), it no longer closes stderr before doing
5515 3. Implemented the mua_wrapper option.
5517 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5519 5. Implemented the functions header_remove(), header_testname(),
5520 header_add_at_position(), and receive_remove_recipient(), and exported them
5523 6. If an ACL "warn" statement specified the addition of headers, Exim already
5524 inserted X-ACL-Warn: at the start if there was no header name. However, it
5525 was not making this test for the second and subsequent header lines if
5526 there were newlines in the string. This meant that an invalid header could
5527 be inserted if Exim was badly configured.
5529 7. Allow an ACL "warn" statement to add header lines at the start or after all
5530 the Received: headers, as well as at the end.
5532 8. Added the rcpt_4xx retry error code.
5534 9. Added postmaster_mailfrom=xxx to callout verification option.
5536 10. Added mailfrom=xxxx to the callout verification option, for verify=
5539 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5540 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5542 12. Inserted some casts to stop certain compilers warning when using pointer
5543 differences as field lengths or precisions in printf-type calls (mostly
5544 affecting debugging statements).
5546 13. Added optional readline() support for -be (dynamically loaded).
5548 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5549 same clock tick as a message's arrival, so that its received time was the
5550 same as the "first fail" time on the retry record, and that message
5551 remained on the queue past the ultimate address timeout, every queue runner
5552 would try a delivery (because it was past the ultimate address timeout) but
5553 after another failure, the ultimate address timeout, which should have then
5554 bounced the address, did not kick in. This was a "< instead of <=" error;
5555 in most cases the first failure would have been in the next clock tick
5556 after the received time, and all would be well.
5558 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5559 being recognized when the domain list was tested by the match_domain
5560 condition in an expansion string.
5562 16. Added the ${str2b64: operator.
5564 17. Exim was always calling setrlimit() to set a large limit for the number of
5565 processes, without checking whether the existing limit was already
5566 adequate. (It did check for the limit on file descriptors.) Furthermore,
5567 errors from getrlimit() and setrlimit() were being ignored. Now they are
5568 logged to the main and panic logs, but Exim does carry on, to try to do its
5569 job under whatever limits there are.
5571 18. Imported PCRE 5.0.
5573 19. Trivial typo in log message " temporarily refused connection" (the leading
5576 20. If the log selector return_path_on_delivery was set and an address was
5577 redirected to /dev/null, the delivery process crashed because it assumed
5578 that a return path would always be set for a "successful" delivery. In this
5579 case, the whole delivery is bypassed as an optimization, and therefore no
5582 21. Internal re-arrangement: the function for sending a challenge and reading
5583 a response while authentication was assuming a zero-terminated challenge
5584 string. It's now changed to take a pointer and a length, to allow for
5585 binary data in such strings.
5587 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5589 23. Exim was not respecting finduser_retries when seeking the login of the
5590 uid under which it was called; it was always trying 10 times. (The default
5591 setting of finduser_retries is zero.) Also, it was sleeping after the final
5592 failure, which is pointless.
5594 24. Implemented tls_on_connect_ports.
5596 25. Implemented acl_smtp_predata.
5598 26. If the domain in control=submission is set empty, Exim assumes that the
5599 authenticated id is a complete email address when it generates From: or
5600 Sender: header lines.
5602 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5603 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5604 chown and chgrp in /bin and hostname in /usr/bin.
5606 28. Exim was keeping the "process log" file open after each use, just as it
5607 does for the main log. This opens the possibility of it remaining open for
5608 long periods when the USR1 signal hits a daemon. Occasional processlog
5609 errors were reported, that could have been caused by this. Anyway, it seems
5610 much more sensible not to leave this file open at all, so that is what now
5613 29. The long-running daemon process does not normally write to the log once it
5614 has entered its main loop, and it closes the log before doing so. This is
5615 so that log files can straightforwardly be renamed and moved. However,
5616 there are a couple of unusual error situations where the daemon does write
5617 log entries, and I had neglected to close the log afterwards.
5619 30. The text of an SMTP error response that was received during a remote
5620 delivery was being truncated at 512 bytes. This is too short for some of
5621 the long messages that one sometimes sees. I've increased the limit to
5624 31. It is now possible to make retry rules that apply only when a message has a
5625 specific sender, in particular, an empty sender.
5627 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5628 it possible to be selective about when SMTP synchronization is enforced.
5630 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5632 32. Implemented hosts_connection_nolog.
5634 33. Added an ACL for QUIT.
5636 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5639 35. Added mailbox_size and mailbox_filecount to appendfile.
5641 36. Added control = no_multiline_responses to ACLs.
5643 37. There was a bug in the logic of the code that waits for the clock to tick
5644 in the case where the clock went backwards by a substantial amount such
5645 that the microsecond fraction of "now" was more than the microsecond
5646 fraction of "then" (but the whole seconds number was less).
5648 38. Added support for the libradius Radius client library this is found on
5649 FreeBSD (previously only the radiusclient library was supported).
5655 1. When certain lookups returned multiple values in the form name=value, the
5656 quoting of the values was not always being done properly. Specifically:
5657 (a) If the value started with a double quote, but contained no whitespace,
5659 (b) If the value contained whitespace other than a space character (i.e.
5660 tabs or newlines or carriage returns) it was not quoted.
5661 This fix has been applied to the mysql and pgsql lookups by writing a
5662 separate quoting function and calling it from the lookup code. The fix
5663 should probably also be applied to nisplus, ibase and oracle lookups, but
5664 since I cannot test any of those, I have not disturbed their existing code.
5666 2. A hit in the callout cache for a specific address caused a log line with no
5667 reason for rejecting RCPT. Now it says "Previous (cached) callout
5668 verification failure".
5670 3. There was an off-by-one bug in the queryprogram router. An over-long
5671 return line was truncated at 256 instead of 255 characters, thereby
5672 overflowing its buffer with the terminating zero. As well as fixing this, I
5673 have increased the buffer size to 1024 (and made a note to document this).
5675 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5676 when Exim is waiting for an SMTP response from a remote server, Exim
5677 restarts its select() call on the socket, thereby resetting its timeout.
5678 This is not a problem when such interrupts are rare. Somebody set up a cron
5679 job to run exiwhat every 2 minutes, which is less than the normal select()
5680 timeout (5 or 10 minutes). This meant that the select() timeout never
5681 kicked in because it was always reset. I have fixed this by comparing the
5682 time when an interrupt arrives with the time at the start of the first call
5683 to select(). If more time than the timeout has elapsed, the interrupt is
5684 treated as a timeout.
5686 5. Some internal re-factoring in preparation for the addition of Sieve
5687 extensions (by MH). In particular, the "personal" test is moved to a
5688 separate function, and given an option for scanning Cc: and Bcc: (which is
5689 not set for Exim filters).
5691 6. When Exim created an email address using the login of the caller as the
5692 local part (e.g. when creating a From: or Sender: header line), it was not
5693 quoting the local part when it contained special characters such as @.
5695 7. Installed new OpenBSD configuration files.
5697 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5698 try to make them clearer.
5700 9. Callout options, other than the timeout value, were being ignored when
5701 verifying sender addresses in header lines. For example, when using
5703 verify = header_sender/callout=no_cache
5705 the cache was (incorrectly) being used.
5707 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5708 only the Cygwin environment.
5710 11. When return_path_on_delivery was set as a log selector, if different remote
5711 addresses in the same message used different return paths and parallel
5712 remote delivery occurred, the wrong values would sometimes be logged.
5713 (Whenever a remote delivery process finished, the return path value from
5714 the most recently started remote delivery process was logged.)
5716 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5717 lines when AUTH and/or TLS are in use. This is the "received protocol"
5718 field. Exim used to use "asmtp" for authenticated SMTP, without any
5719 indication (in the protocol name) for TLS use. Now it follows the RFC and
5720 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5721 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5722 names appear in log lines as well as in Received: header lines.
5724 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5725 extensions, and comparison tests, and to fix some bugs.
5727 14. Changes to the "personal" filter test:
5729 (1) The test was buggy in that it was just doing the equivalent of
5730 "contains" tests on header lines. For example, if a user's address was
5731 anne@some.where, the "personal" test would incorrectly be true for
5733 To: susanne@some.where
5735 This test is now done by extracting each address from the header in turn,
5736 and checking the entire address. Other tests that are part of "personal"
5737 are now done using regular expressions (for example, to check local parts
5738 of addresses in From: header lines).
5740 (2) The list of non-personal local parts in From: addresses has been
5741 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5742 taken from the Sieve specification recommendations.
5744 (3) If the message contains any header line starting with "List-" it is
5745 treated as non-personal.
5747 (4) The test for "circular" in the Subject: header line has been removed
5748 because it now seems ill-conceived.
5750 15. Minor typos in src/EDITME comments corrected.
5752 16. Installed latest exipick from John Jetmore.
5754 17. If headers_add on a router specified a text string that was too long for
5755 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5756 of string_sprintf() is now avoided.
5758 18. $message_body_size was not set (it was always zero) when running the DATA
5759 ACL and the local_scan() function.
5761 19. For the "mail" command in an Exim filter, no default was being set for
5762 the once_repeat time, causing a random time value to be used if "once" was
5763 specified. (If the value happened to be <= 0, no repeat happened.) The
5764 default is now 0s, meaning "never repeat". The "vacation" command was OK
5765 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5766 (I found it when inspecting the code).
5768 20. There is now an overall timeout for performing a callout verification. It
5769 defaults to 4 times the callout timeout, which applies to individual SMTP
5770 commands during the callout. The overall timeout applies when there is more
5771 than one host that can be tried. The timeout is checked before trying the
5772 next host. This prevents very long delays if there are a large number of
5773 hosts and all are timing out (e.g. when the network connections are timing
5774 out). The value of the overall timeout can be changed by specifying an
5775 additional sub-option for "callout", called "maxwait". For example:
5777 verify = sender/callout=5s,maxwait=20s
5779 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5780 to the end before writing, but this should make it even safer).
5782 22. Exim was forgetting that it had advertised PIPELINING for the second and
5783 subsequent messages on an SMTP connection. It was also not resetting its
5784 memory on STARTTLS and an internal HELO.
5786 23. When Exim logs an SMTP synchronization error within a session, it now
5787 records whether PIPELINING has been advertised or not.
5789 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5790 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5791 rather than long int.
5793 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5795 26. Added the never_mail option to autoreply.
5801 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5802 crash if the getsockname() call failed; this can happen if a connection is
5803 closed very soon after it is established. The problem was simply in the
5804 order in which certain operations were done, causing Exim to try to write
5805 to the SMTP stream before it had set up the file descriptor. The bug has
5806 been fixed by making things happen in the correct order.
5812 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5813 before the connection was closed, thus losing the rejection response.
5815 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5816 some early Solaris releases, but causes trouble in current releases where
5817 socklen_t is defined.
5819 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5822 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5825 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5826 that releasing the top of it at the end releases what was used for sub-
5827 expansions (unless the block got too big). However, discard this block if
5828 the first thing is a variable or header, so that we can use its block when
5829 it is dynamic (useful for very large $message_headers, for example).
5831 6. Lookups now cache *every* query, not just the most recent. A new, separate
5832 store pool is used for this. It can be recovered when all lookup caches are
5833 flushed. Lookups now release memory at the end of their result strings.
5834 This has involved some general refactoring of the lookup sources.
5836 7. Some code has been added to the store_xxx() functions to reduce the amount
5837 of flapping under certain conditions.
5839 8. log_incoming_interface used to affect only the <= reception log lines. Now
5840 it causes the local interface and port to be added to several more SMTP log
5841 lines, for example "SMTP connection from", and rejection lines.
5843 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5845 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5847 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5848 could be overwritten at the end of the current message (or the start of a
5849 new message if it was set in a HELO ACL). The value is now preserved for
5850 the duration of the SMTP connection.
5852 12. If a transport had a headers_rewrite setting, and a matching header line
5853 contained an unqualified address, that address was qualified, even if it
5854 did not match any rewriting rules. The underlying bug was that the values
5855 of the flags that permit the existence of unqualified sender and recipient
5856 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5857 non-local messages, and by -bnq for local messages) were not being
5858 preserved with the message after it was received.
5860 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5861 "next input=" as part of the text comprising the host identity instead of
5862 the correct text. The code was using the same buffer for two different
5863 strings. However, depending on which order the printing function evaluated
5864 its arguments, the bug did not always show up. Under Linux, for example, my
5865 test suite worked just fine.
5867 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5868 doesn't work with some older versions of Perl. It has been changed to "my",
5869 which in any case is probably the better facility to use.
5871 15. A really picky compiler found some instances of statements for creating
5872 error messages that either had too many or two few arguments for the format
5875 16. The size of the buffer for calls to the DNS resolver has been increased
5876 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5877 for addresses that have a lot of PTR records. This alleviates a problem; it
5878 does not fully solve it.
5880 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5881 buffer now truncates the list and logs the incident, which is the same
5882 action as happens when Exim is looking up a host name and its aliases.
5883 Previously in this situation something unpredictable would happen;
5884 sometimes it was "internal error: store_reset failed".
5886 18. If a server dropped the connection unexpectedly when an Exim client was
5887 using GnuTLS and trying to read a response, the client delivery process
5888 crashed while trying to generate an error log message.
5890 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5891 string, for example:
5893 warn message = H1: something\nH2: something
5895 the text was added as a single header line from Exim's point of view
5896 though it ended up OK in the delivered message. However, searching for the
5897 second and subsequent header lines using $h_h2: did not work. This has been
5898 fixed. Similarly, if a system filter added multiple headers in this way,
5899 the routers could not see them.
5901 20. Expanded the error message when iplsearch is called with an invalid key to
5902 suggest using net-iplsearch in a host list.
5904 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5905 ACLs are no longer actually imposed (and a message to that effect is
5908 22. If a "gecos" field in a passwd entry contained escaped characters, in
5909 particular, if it contained a \" sequence, Exim got it wrong when building
5910 a From: or a Sender: header from that name. A second bug also caused
5911 incorrect handling when an unquoted " was present following a character
5912 that needed quoting.
5914 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5915 was not being matched caselessly.
5917 24. Arranged for all hyphens in the exim.8 source to be escaped with
5920 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5921 properly. Recipient callout cache records were still being keyed to include
5922 the sender, even when use_sender was set false. This led to far more
5923 callouts that were necessary. The sender is no longer included in the key
5924 when use_sender is false.
5926 26. Added "control = submission" modifier to ACLs.
5928 27. Added the ${base62d: operator to decode base 62 numbers.
5930 28. dnsdb lookups can now access SRV records.
5932 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5933 the configuration file.
5935 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5936 (-v) mode. This makes the output for a verbose queue run more intelligible.
5938 31. Added a use_postmaster feature to recipient callouts.
5940 32. Added the $body_zerocount variable, containing the number of binary zero
5941 bytes in the message body.
5943 33. The time of last modification of the "new" subdirectory is now used as the
5944 "mailbox time last read" when there is a quota error for a maildir
5947 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5949 35. Added +ignore_unknown as a special item in host lists.
5951 36. Code for decoding IPv6 addresses in host lists is now included, even if
5952 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5953 address was recognized as an IP address, but was then not correctly decoded
5954 into binary, causing unexpected and incorrect effects when compared with
5961 1. Very minor rewording of debugging text in manualroute to say "list of
5962 hosts" instead of "hostlist".
5964 2. If verify=header_syntax was set, and a header line with an unqualified
5965 address (no domain) and a large number of spaces between the end of the
5966 name and the colon was received, the reception process suffered a buffer
5967 overflow, and (when I tested it) crashed. This was caused by some obsolete
5968 code that should have been removed. The fix is to remove it!
5970 3. When running in the test harness, delay a bit after writing a bounce
5971 message to get a bit more predictability in the log output.
5973 4. Added a call to search_tidyup() just before forking a reception process. In
5974 theory, someone could use a lookup in the expansion of smtp_accept_max_
5975 per_host which, without the tidyup, could leave open a database connection.
5977 5. Added the variables $recipient_data and $sender_data which get set from a
5978 lookup success in an ACL "recipients" or "senders" condition, or a router
5979 "senders" option, similar to $domain_data and $local_part_data.
5981 6. Moved the writing of debug_print from before to after the "senders" test
5984 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5985 problems for message scanning, either using a data ACL, or using
5986 local_scan() because the Received: header was not generated till after they
5987 were called (in order to set the time as the time of reception completion).
5988 I have revised the way this works. The header is now generated after the
5989 body is received, but before the ACL or local_scan() are called. After they
5990 are run, the timestamp in the header is updated.
5996 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5997 before starting a queue runner without re-exec. This happened only when
5998 deliver_drop_privilege was set or when the Exim user was set to root. The
5999 effect of the bug was that timeouts during subsequent deliveries caused
6000 crashes instead of being properly handled. The handler is now left at its
6001 default (and expected) setting.
6003 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6004 message, again when deliver_drop_privilege is set or Exim is run as root.
6005 The bug described in (1) was not present in this case, but the tidying up
6006 of the other signals was missing. I have made the two cases consistent.
6008 3. The ignore_target_hosts setting on a manualroute router was being ignored
6009 for hosts that were looked up using the /MX notation.
6011 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6014 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6015 operated on the sender address. After changing the $sender_address to <>
6016 for the sender address verify, Exim was re-instated it as the original
6017 (before rewriting) address, but remembering that it had rewritten it, so it
6018 wasn't rewriting it again. This bug also had the effect of breaking the
6019 sender address verification caching when the sender address was rewritten.
6021 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6022 This has been changed so that if the ip literal address matches
6023 ignore_target_hosts, the router declines.
6025 7. Added expansion conditions match_domain, match_address, and match_local_
6026 part (NOT match_host).
6028 8. The placeholder for the Received: header didn't have a length field set.
6030 9. Added code to Exim itself and to exim_lock to test for a specific race
6031 condition that could lead to file corruption when using MBX delivery. The
6032 issue is with the lockfile that is created in /tmp. If this file is removed
6033 after a process has opened it but before that process has acquired a lock,
6034 there is the potential for a second process to recreate the file and also
6035 acquire a lock. This could lead to two Exim processes writing to the file
6036 at the same time. The added code performs the same test as UW imapd; it
6037 checks after acquiring the lock that its file descriptor still refers to
6038 the same named file.
6040 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6041 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6044 11. Added the smtp_active_hostname option. If used, this will typically be made
6045 to depend on the incoming interface address. Because $interface_address is
6046 not set up until the daemon has forked a reception process, error responses
6047 that can happen earlier (such as "too many connections") no longer contain
6050 12. If an expansion in a condition on a "warn" statement fails because a lookup
6051 defers, the "warn" statement is abandoned, and the next ACL statement is
6052 processed. Previously this caused the whole ACL to be aborted.
6054 13. Added the iplsearch lookup type.
6056 14. Added ident_timeout as a log selector.
6058 15. Added tls_certificate_verified as a log selector.
6060 16. Added a global option tls_require_ciphers (compare the smtp transport
6061 option of the same name). This controls incoming TLS connections.
6063 17. I finally figured out how to make tls_require_ciphers do a similar thing
6064 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6065 before starting the TLS session.
6067 18. Tabs are now shown as \t in -bP output.
6069 19. If the log selector return_path_on_delivery was set, Exim crashed when
6070 bouncing a message because it had too many Received: header lines.
6072 20. If two routers both had headers_remove settings, and the first one included
6073 a superfluous trailing colon, the final name in the first list and the
6074 first name in the second list were incorrectly joined into one item (with a
6075 colon in the middle).
6081 1. Added -C and -D options to the exinext utility, mainly to make it easier
6082 to include in the automated testing, but these could be helpful when
6083 multiple configurations are in use.
6085 2. The exinext utility was not formatting the output nicely when there was
6086 an alternate port involved in the retry record key, nor when there was a
6087 message id as well (for retries that were specific to a specific message
6088 and a specific host). It was also confused by IPv6 addresses, because of
6089 the additional colons they contain. I have fixed the IPv4 problem, and
6090 patched it up to do a reasonable job for IPv6.
6092 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6093 delivery, the log line now contains "pipelined" if PIPELINING was used.
6095 4. An SMTP transport process used to panic and die if the bind() call to set
6096 an explicit outgoing interface failed. This has been changed; it is now
6097 treated in the same way as a connect() failure.
6099 5. A reference to $sender_host_name in the part of a conditional expansion
6100 that was being skipped was still causing a DNS lookup. This no longer
6103 6. The def: expansion condition was not recognizing references to header lines
6104 that used bh_ and bheader_.
6106 7. Added the _cache feature to named lists.
6108 8. The code for checking quota_filecount in the appendfile transport was
6109 allowing one more file than it should have been.
6111 9. For compatibility with Sendmail, the command line option
6119 and sets the incoming protocol and host name (for trusted callers). The
6120 host name and its colon can be omitted when only the protocol is to be set.
6121 Note the Exim already has two private options, -pd and -ps, that refer to
6122 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6123 "s", but I don't think that's a major issue.
6125 10. A number of refactoring changes to the code, none of which should affect
6128 (a) The number of logging options was getting close to filling up the
6129 32-bit word that was used as a bit map. I have split them into two classes:
6130 those that are passed in the argument to log_write(), and those that are
6131 only ever tested independently outside of that function. These are now in
6132 separate 32-bit words, so there is plenty of room for expansion again.
6133 There is no change in the user interface or the logging behaviour.
6135 (b) When building, for example, log lines, the code previously used a
6136 macro that called string_cat() twice, in order to add two strings. This is
6137 not really sufficiently general. Furthermore, there was one instance where
6138 it was actually wrong because one of the argument was used twice, and in
6139 one call a function was used. (As it happened, calling the function twice
6140 did not affect the overall behaviour.) The macro has been replaced by a
6141 function that can join an arbitrary number of extra strings onto a growing
6144 (c) The code for expansion conditions now uses a table and a binary chop
6145 instead of a serial search (which was left over from when there were very
6146 few conditions). Also, it now recognizes conditions like "pam" even when
6147 the relevant support is not compiled in: a suitably worded error message is
6148 given if an attempt is made to use such a condition.
6150 11. Added ${time_interval:xxxxx}.
6152 12. A bug was causing one of the ddress fields not to be passed back correctly
6153 from remote delivery subprocesses. The field in question was not being
6154 subsequently used, so this caused to problems in practice.
6156 13. Added new log selectors queue_time and deliver_time.
6158 14. Might have fixed a bug in maildirsizefile handling that threw up
6159 "unexpected character" debug warnings, and recalculated the data
6160 unnecessarily. In any case, I expanded the warning message to give more
6163 15. Added the message "Restricted characters in address" to the statements in
6164 the default ACL that block characters like @ and % in local parts.
6166 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6167 Three changes have been made:
6169 (a) There was a serious bug; a negative response to MAIL caused the whole
6170 recipient domain to be cached as invalid, thereby blocking all messages
6171 to all local parts at the same domain, from all senders. This bug has
6172 been fixed. The domain is no longer cached after a negative response to
6173 MAIL if the sender used is not empty.
6175 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6178 (c) A new callout option, "use_sender" has been added for people who want
6179 the modified behaviour.
6185 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6188 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6189 indeed breaks things for older releases.
6191 3. Added additional logging to the case where there is a problem reading data
6192 from a filter that is running in a subprocess using a pipe, in order to
6193 try to track down a specific problem.
6195 4. Testing facility fudge: when running in the test harness and attempting
6196 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6197 getting "No route to host". Convert this to a timeout.
6199 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6202 6. Some OS don't have socklen_t but use size_t instead. This affects the
6203 fifth argument of getsockopt() amongst other things. This is now
6204 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6205 can be set for individual OS. I have set it for SunOS5, OSF1, and
6206 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6207 some earlier ones do not.
6209 7. Change 4.30/15 was not doing the test caselessly.
6211 8. The standard form for an IPv6 address literal was being rejected by address
6212 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6213 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6214 this, as well as the form without the "IPv6" on the front (but only when
6215 address literals are enabled, of course).
6217 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6219 10. Exim crashed if a message with an empty sender address specified by -f
6220 encountered a router with an errors_to setting. This could be provoked only
6221 by a command such as
6225 where an empty string was supplied; "<>" did not hit this bug.
6227 11. Installed PCRE release 4.5.
6229 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6230 remained set. It is now erased.
6232 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6233 times from message ids (which are base 36 rather than the normal 62).
6235 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6236 were being counted as actual protocol errors, and logged if the log
6237 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6238 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6239 and DATA following a set of rejected RCPTs do not count as protocol errors.
6240 In other words, Exim assumes they were pipelined, though this may not
6241 actually be the case. Of course, in all cases the client gets an
6242 appropriate error code.
6244 15. If a lookup fails in an ACL condition, a message about the failure may
6245 be available; it is used if testing the ACL cannot continue, because most
6246 such messages specify what the cause of the deferral is. However, some
6247 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6248 that caused an old message to be retained and used if a later statement
6249 caused a defer, replacing the real cause of the deferral.
6251 16. If an IP address had so many PTR records that the DNS lookup buffer
6252 was not large enough to hold them, Exim could crash while trying to process
6253 the truncated data. It now detects and logs this case.
6255 17. Further to 4.21/58, another change has been made: if (and only if) the
6256 first line of a message (the first header line) ends with CRLF, a bare LF
6257 in a subsequent header line has a space inserted after it, so as not to
6258 terminate the header.
6260 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6261 unnecessarily, used atoi() instead of strtol(), and didn't check the
6262 termination when getting file sizes from file names by regex.
6264 19. Completely re-implemented the support for maildirsize files, in the light
6265 of a number of problems with the previous contributed implementation
6266 (4.30/29). In particular:
6268 . If the quota is zero, the maildirsize file is maintained, but no quota is
6271 . If the maildir directory does not exist, it is created before any attempt
6272 to write a maildirsize file.
6274 . The quota value in the file is just a cache; if the quota is changed in
6275 the transport, the new value overrides.
6277 . A regular expression is available for excluding directories from the
6280 20. The autoreply transport checks the characters in options that define the
6281 message's headers; it allows continued headers, but it was checking with
6282 isspace() after an embedded newline instead of explicitly looking for a
6285 21. If all the "regular" hosts to which an address was routed had passed their
6286 expiry times, and had not reached their retry times, the address was
6287 bounced, even if fallback hosts were defined. Now Exim should go on to try
6290 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6291 equivalent code in the SMTP transport. Some hosts send humungous responses
6292 to HELO/EHLO, more than 1024 it seems.
6294 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6295 gives compiler warnings in some environments. I've now done it "properly",
6298 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6299 (because of problems with the built-in one) was declared to return uschar *
6300 instead of char *, causing compiler failure.
6302 25. Fixed a file descriptor leak when processing alias/forward files.
6304 26. Fixed a minor format string issue in dbfn.c.
6306 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6308 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6309 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6312 29. When a sender address is verified, it is cached, to save repeating the test
6313 when there is more than one recipient in a message. However, when the
6314 verification involves a callout, it is possible for different callout
6315 options to be set for different recipients. It is too complicated to keep
6316 track of this in the cache, so now Exim always runs a verification when a
6317 callout is required, relying on the callout cache for the optimization.
6318 The overhead is duplication of the address routing, but this should not be
6321 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6322 to be verified with callout=postmaster, and the main callout worked but the
6323 postmaster check failed, the verification correctly failed. However, if a
6324 subsequent RCPT command asked for sender verification *without* the
6325 postmaster check, incorrect caching caused this verification also to fail,
6328 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6329 it was not caching the DNS options (qualify_single, search_parents) that
6330 were used when the lookup failed. A subsequent lookup with different
6331 options therefore always gave the same answer, though there were cases
6332 where it should not have. (Example: a "domains = !$mx_any" option on a
6333 dnslookup router: the "domains" option is always processed without any
6334 widening, but the router might have qualify_single set.) Now Exim uses the
6335 cached value only when the same options are set.
6337 32. Added John Jetmore's "exipick" utility to the distribution.
6339 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6340 than a timeout (e.g. a certificate is required, and is not provided), an
6341 Exim server now closes the connection immediately. Previously it waited for
6342 the client to close - but if the client is SSL, it seems that they each
6343 wait for each other, leading to a delay before one of them times out.
6345 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6346 maintained 0.8.x compatibility because I don't think many are using it, and
6347 it is clearly obsolete.
6349 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6352 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6353 client certificate was expired. A simple patch fixes this, though I don't
6354 understand the full logic of why the verify callback is called multiple
6357 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6358 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6359 which causes problems with some clients (such as the Certicom SSL Plus
6360 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6361 disables the coutermeasure allowing Eudora to connect."
6363 38. Exim was not checking that a write() to a log file succeeded. This could
6364 lead to Bad Things if a log got too big, in particular if it hit a file
6365 size limit. Exim now panics and dies if it cannot write to a log file, just
6366 as it does if it cannot open a log file.
6368 39. Modified OS/Makefile-Linux so that it now contains
6370 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6372 The two -D definitions ensure that Exim is compiled with large file
6373 support, which makes it possible to handle log files that are bigger than
6376 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6377 instance) a domain was checked against a named list that involved a lookup,
6378 causing $domain_data to be set, then another domain was checked against the
6379 same list, then the first domain was re-checked, the value of $domain_data
6380 after the final check could be wrong. In particular, if the second check
6381 failed, it could be set empty. This bug probably also applied to
6384 41. The strip_trailing_dot option was not being applied to the address given
6385 with the -f command-line option.
6387 42. The code for reading a message's header from the spool was incrementing
6388 $received_count, but never initializing it. This meant that the value was
6389 incorrect (doubled) while delivering a message in the same process in which
6390 it was received. In the most common configuration of Exim, this never
6391 happens - a fresh exec is done - but it can happen when
6392 deliver_drop_privilege is set.
6394 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6395 it now includes up to 150 characters of the unexpected data in the log
6398 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6399 and building data strings. The size of both of these buffers was 10 000
6400 bytes - far larger than anybody would *ever* want, thought I. Needless to
6401 say, somebody hit the limit. I have increased the maximum line length to
6402 20 000 and the maximum data length of concatenated lines to 100 000. I have
6403 also fixed two bugs, because there was no checking on these buffers. Tsk,
6404 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6405 buffer is too small.
6407 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6408 lsearch lookups. Now it does.
6410 46. When parsing a route_list item in a manualroute router, a fixed-length
6411 buffer was used for the list of hosts. I made this 1024 bytes long,
6412 thinking that nobody would ever have a list of hosts that long. Wrong.
6413 Somebody had a whole pile of complicated expansion conditions, and the
6414 string was silently truncated, leading to an expansion error. It turns out
6415 that it is easier to change to an unlimited length (owing to other changes
6416 that have happened since this code was originally written) than to build
6417 structure for giving a limitation error. The length of the item that
6418 expands into the list of hosts is now unlimited.
6420 47. The lsearch lookup could not handle data where the length of text line was
6421 more than 4095 characters. Such lines were truncated, leading to shortened
6422 data being returned. It should now handle lines of any length.
6424 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6425 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6428 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6429 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6430 no longer happens in comment lines. A list of replacements is now placed
6431 at the head of all of the source files, except those whose only change is
6432 to replace PERL_COMMAND in the very first #! line.
6434 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6435 messages on the queue, with a bottom-up merge sort, using code contributed
6436 by Michael Haardt. This should make operations like -bp somewhat faster on
6437 large queues. It won't affect queue runners, except when queue_run_in_order
6440 51. Installed eximstats 1.31 in the distribution.
6442 52. Added support for SRV lookups to the dnslookup router.
6444 53. If an ACL referred to $message_body or $message_body_end, the value was not
6445 reset for any messages that followed in the same SMTP session.
6447 54. The store-handling optimization for building very long strings was not
6448 differentiating between the different store pools. I don't think this
6449 actually made any difference in practice, but I've tidied it.
6451 55. While running the routers to verify a sender address, $sender_address
6452 was still set to the sender address. This is wrong, because when routing to
6453 send a bounce to the sender, it would be empty. Therefore, I have changed
6454 it so that, while verifying a sender address, $sender_address is set to <>.
6455 (There is no change to what happens when verifying a recipient address.)
6457 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6458 target A or AAAA records (if not already returned) without resetting the
6459 qualify_single or search_parents options of the DNS resolver. These are
6460 inappropriate in this case because the targets of MX and SRV records must
6461 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6462 target that, when qualified, matched something in the local domain. These
6463 two options are now turned off when doing these lookups.
6465 57. It seems that at least some releases of Reiserfs (which does not have the
6466 concept of a fixed number of inodes) returns zero and not -1 for the
6467 number of available inodes. This interacted badly with check_spool_inodes,
6468 which assumed that -1 was the "no such thing" setting. What I have done is
6469 to check that the total number of inodes is greater than zero before doing
6470 the test of how many are available.
6472 58. When a "warn" ACL statement has a log_message modifier, the message is
6473 remembered, and not repeated. This is to avoid a lot of repetition when a
6474 message has many recipients that cause the same warning to be written.
6475 However, Exim was preserving the list of already written lines for an
6476 entire SMTP session, which doesn't seem right. The memory is now reset if a
6477 new message is started.
6479 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6480 debugging output unless log_rewrite was also set.
6482 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6483 of (int)(handle) when we know that handle contains (void *)(-1).
6485 61. The Exim daemon panic-logs an error return when it closes the incoming
6486 connection. However "connection reset by peer" seems to be common, and
6487 isn't really an error worthy of noting specially, so that particular error
6490 62. When Exim is trying to find all the local interfaces, it used to panic and
6491 die if the ioctl to get the interface flags failed. However, it seems that
6492 on at least one OS (Solaris 9) it is possible to have an interface that is
6493 included in the list of interfaces, but for which you get a failure error
6494 for this call. This happens when the interface is not "plumbed" into a
6495 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6496 failure of the "get flags" call assumes that the interface is down.
6498 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6499 makes life easier for people who are doing arithmetic on fields extracted
6500 from dates, where you often get leading zeros that should not be
6501 interpreted as octal.
6503 64. Added qualify_domain to the redirect router, to override the global
6506 65. If a pathologically long header line contained very many addresses (the
6507 report of this problem mentioned 10 000) and each of them was rewritten,
6508 Exim could use up a very large amount of memory. (It kept on making new
6509 copies of the header line as it rewrote, and never released the old ones.)
6510 At the expense of a bit more processing, the header rewriting function has
6511 been changed so that it no longer eats memory in this way.
6513 66. The generation of the Received: header has been moved from the time that a
6514 message starts to be received, to the time that it finishes. The timestamp
6515 in the Received: header should now be very close to that of the <= log
6516 line. There are two side-effects of this change:
6518 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6519 logged header lines no longer include the local Received: line, because
6520 it has not yet been created. The same applies to a copy of the message
6521 that is returned to a non-SMTP sender when a message is rejected.
6523 (b) When a filter file is tested using -bf, no additional Received: header
6524 is added to the test message. After some thought, I decided that this
6527 This change does not affect the value of $received_for. It is still set
6528 after address rewriting, but before local_scan() is called.
6530 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6532 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6533 gave an unhelpful panic error message, and a defer error. I have managed to
6534 change this behaviour so that it now rejects any supplied certificate,
6535 which seems right, as the list of acceptable certificates is empty.
6537 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6538 gave an unhelpful defer error. I have not managed to make this reject any
6539 supplied certificates, but the error message it gives is "no certificate
6540 supplied", which is not helpful.
6542 70. exigrep's output now also includes lines that are not associated with any
6543 message, but which match the given pattern. Implemented by a patch from
6544 Martin Sluka, which also tidied up the Perl a bit.
6546 71. Recipient callout verification, like sender verification, was using <> in
6547 the MAIL FROM command. This isn't really the right thing, since the actual
6548 sender may affect whether the remote host accepts the recipient or not. I
6549 have changed it to use the actual sender in the callout; this means that
6550 the cache record is now keyed on a recipient/sender pair, not just the
6551 recipient address. There doesn't seem to be a real danger of callout loops,
6552 since a callout by the remote host to check the sender would use <>.
6553 [SEE ABOVE: changed after hitting problems.]
6555 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6556 temporary errors. However, in the case of such a code being given after
6557 the end of a data transmission (i.e. after ".") Exim was failing to write
6558 a retry record for the message. (Yes, there was some broken host that was
6559 actually sending 8xx at this point.)
6561 73. An unknown lookup type in a host list could cause Exim to panic-die when
6562 the list was checked. (An example that provoked this was putting <; in the
6563 middle of a list instead of at the start.) If this happened during a DATA
6564 ACL check, a -D file could be left lying around. This kind of configuration
6565 error no longer causes Exim to die; instead it causes a defer error. The
6566 incident is still logged to the main and panic logs.
6568 74. Buglet left over from Exim 3 conversion. The message "too many messages
6569 in one connection" was written to the rejectlog but not the mainlog, except
6570 when address rewriting (yes!) was being logged.
6572 75. Added write_rejectlog option.
6574 76. When a system filter was run not as root (that is, when system_filter_user
6575 was set), the values of the $n variables were not being returned to the
6576 main process; thus, they were not subsequently available in the $sn
6579 77. Added +return_path_on_delivery log selector.
6581 78. A connection timeout was being treated differently from recipients deferred
6582 when testing hosts_max_try with a message that was older than the host's
6583 retry timeout. (The host should not be counted, thus allowing all hosts to
6584 be tried at least once before bouncing.) This may have been the cause of an
6585 occasionally reported bug whereby a message would remain on the queue
6586 longer than the retry timeout, but would be bounced if a delivery was
6587 forced. I say "may" because I never totally pinned down the problem;
6588 setting up timeout/retry tests is difficult. See also the next item.
6590 79. The ultimate address timeout was not being applied to errors that involved
6591 a combination of host plus message (for example, a timeout on a MAIL
6592 command). When an address resolved to a number of possible hosts, and they
6593 were not all tried for each delivery (e.g. because of hosts_max_try), a
6594 message could remain on the queue longer than the retry timeout.
6596 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6599 81. Fixed an obscure SMTP outgoing bug which required at least the following
6600 conditions: (a) there was another message waiting for the same server;
6601 (b) the server returned 5xx to all RCPT commands in the first message so
6602 that the message was not completed; (c) the server dropped the connection
6603 or gave a negative response to the RSET that Exim sends to abort the
6604 transaction. The observed case was a dropped connection after DATA that had
6605 been sent in pipelining mode. That is, the server had advertised PIPELINING
6606 but was not implementing it correctly. The effect of the bug was incorrect
6607 behaviour, such as trying another host, and this could lead to a crash.
6613 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6614 and daemon.c were passed as pointers to ints; they should have been
6615 pointers to socklen_t variables (which are typically unsigned ints).
6617 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6620 3. Fixed a really odd bug that affected only the testing scheme; patching a
6621 certain fixed string in the binary changed the value of another string that
6622 happened to be identical to the end of the original first string.
6624 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6625 name", it returns that address as the IP address. On some operating
6626 systems (e.g. Solaris), it also passes back the IP address string as the
6627 "host name". However, on others (e.g. Linux), it passes back an empty
6628 string. Exim wasn't checking for this, and was changing the host name to an
6629 empty string, assuming it had been canonicalized.
6631 5. Although rare, it is permitted to have more than one PTR record for a given
6632 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6633 all the names associated with an address, because they do in Solaris.
6634 However, it seems that they do not in Linux for data that comes from the
6635 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6636 I found this out when I moved to a new Linux workstation and tried to run
6637 the Exim test suite.
6639 To get round this problem I have changed the code so that it now does its
6640 own call to the DNS to look up PTR records when searching for a host name.
6641 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6642 addresses that are only in /etc/hosts are still found.
6644 This behaviour is, however, controlled by an option called host_lookup_
6645 order, which defaults to "bydns:byaddr". If people want to use the other
6646 order, or indeed, just use one or the other means of lookup, they can
6647 specify it in this variable.
6649 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6650 some operating systems, this comes back from gethostbyaddr() as an empty
6651 string, and this is what Exim used to test for. However, it seems that in
6652 other systems, "." is yielded. Exim now tests for this case too.
6654 7. The values of check_spool_space and check_log_space are now held internally
6655 as a number of kilobytes instead of an absolute number of bytes. If a
6656 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6657 kilobyte. This means that much larger values can be stored.
6659 8. Exim monitor: an attempt to get the action menu when not actually pointing
6660 at a message produces an empty menu entitled "No message selected". This
6661 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6662 no entries in it ("Shell widget menu has zero width and/or height"). So I
6663 have added a single, blank menu entry in this case.
6665 9. Added ${quote_local_part.
6667 10. MIME decoding is now applied to the contents of Subject: header lines when
6670 11. Now that a reference to $sender_host_address automatically causes a reverse
6671 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6672 host lookup before query-style lookups in lists that might use this
6673 variable. This has therefore been abolished, and the "net-" prefix is no
6674 longer necessary for query-style lookups.
6676 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6677 have been a typo for LFLAGS, so it has been changed.
6679 13. The install script calls Exim with "-C /dev/null" in order to find the
6680 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6681 to be output. However, since Exim outputs its version number before the
6682 error, it didn't break the script. It just looked ugly. I fixed this by
6683 always allowing "-C /dev/null" if the caller is root.
6685 14. Ignore overlarge ACL variable number when reading spool file - insurance
6686 against a later release with more variables having written the file.
6688 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6689 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6690 this, as well as the form without the "IPv6" on the front.
6692 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6693 OS/Makefile-Darwin file.
6695 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6696 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6698 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6699 variable. However, some OS use ints for inodes. I've added cast to long int
6700 to get rid of the compiler warning.
6702 19. I had forgotten to lock out "/../" in configuration file names when
6703 ALT_CONFIG_PREFIX was set.
6705 20. Routers used for verification do not need to specify transports. However,
6706 if such a router generated a host list, and callout was configured, Exim
6707 crashed, because it could not find a port number from the (non-existent)
6708 transport. It now assumes port 25 in this circumstance.
6710 21. Added the -t option to exigrep.
6712 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6713 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6714 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6715 them is set, LOOKUP_LSEARCH is forced.
6717 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6719 24. Added sender and host information to the "rejected by local_scan()" log
6720 line; previously there was no indication of these.
6722 25. Added .include_if_exists.
6724 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6725 Linux. It turns out that not all file systems support this. Apparently some
6726 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6727 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6728 not supported on the file descriptor, is now ignored when Exim is trying to
6729 sync a directory. This applies only to Linux.
6731 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6733 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6734 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6737 29. Added support for maildirsize files from supplied patch (modified a bit).
6739 30. The use of :fail: followed by an empty string could lead Exim to respond to
6740 sender verification failures with (e.g.):
6742 550 Verification failed for <xxx>
6743 550 Sender verify failed
6745 where the first response line was missing the '-' that indicates it is not
6746 the final line of the response.
6748 31. The loop for finding the name of the user that called Exim had a hardwired
6749 limit of 10; it now uses the value of finduser_retries, which is used for
6750 all other user lookups.
6752 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6755 33. Exim was neglecting to zero errno before one call of strtol() when
6756 expanding a string and expecting an integer value. On some systems this
6757 resulted in spurious "integer overflow" errors. Also, it was casting the
6758 result into an int without checking.
6760 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6761 did not work. The code looks as if it has *never* worked, though it appears
6762 to have been documented since at least release 1.62. I have made it work.
6764 35. The "timeout_DNS" error in retry rules, also documented since at least
6765 1.62, also never worked. As it isn't clear exactly what this means, and
6766 clearly it isn't a major issue, I have abolished the feature by treating it
6767 as "timeout", and writing a warning to the main and panic logs.
6769 36. The display of retry rules for -brt wasn't always showing the error code
6772 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6773 timeout_connect_A, timeout_connect_MX.
6775 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6776 to the empty sender.
6778 39. The daemon was not analysing the content of -oX till after it had closed
6779 stderr and disconnected from the controlling terminal. This meant that any
6780 syntax errors were only noted on the panic log, and the return code from
6781 the command was 0. By re-arranging the code a little, I've made the
6782 decoding happen first, so such errors now appear on stderr, and the return
6783 code is 1. However, the actual setting up of the sockets still happens in
6784 the disconnected process, so errors there are still only recorded on the
6787 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6788 connections (as happens on some IP stacks) was logged at start up time as
6789 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6790 it from "IPv6 and IPv4", which means that two separate sockets are being
6793 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6794 says whether AF_INET or AF_INET6 was passed as an argument.
6796 42. Exiwhat output was messed up when time zones were included in log
6799 43. Exiwhat now gives more information about the daemon's listening ports,
6800 and whether -tls-on-connect was used.
6802 44. The "port" option of the smtp transport is now expanded.
6804 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6805 silently ignored. Now an error message is written to the main and panic
6808 46. There's a new ACL modifier called "logwrite" which writes to a log file
6809 as soon as it is encountered.
6811 47. Added $local_user_uid and $local_user_gid at routing time.
6813 48. Exim crashed when trying to verify a sender address that was being
6816 49. Exim was recognizing only a space character after ".include". It now also
6817 recognizes a tab character.
6819 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6820 extracting the relevant information from the specification. The man page no
6821 longer contains scrambled data for the -d option, and I've added a section
6822 at the front about calling Exim under different names.
6824 51. Added "extra_headers" argument to the "mail" command in filter files.
6826 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6829 53. Installed eximstats 1.29.
6831 54. Added transport_filter_timeout as a generic transport option.
6833 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6834 Cc: header lines. This was required by RFC 822, but it not required by RFC
6837 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6838 incoming messages that did not have them. Now it does so only if the
6839 message originates locally, that is, if there is no associated remote host
6840 address. When Resent- header lines are present, this applies to the Resent-
6841 lines rather than the non-Resent- lines.
6843 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6844 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6846 58. Messages for configuration errors now include the name of the main
6847 configuration file - useful now that there may be more than one file in a
6848 list (.included file names were always shown).
6850 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6851 for those rare installations that do not start the daemon as root or run it
6852 setuid root. I've cut out the call to initgroups() if the daemon is not
6855 60. The Exim user and group can now be bound into the binary as text strings
6856 that are looked up at the start of Exim's processing.
6858 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6860 62. Added $mailstore_basename variable.
6862 63. Installed patch to sieve.c from Michael Haardt.
6864 64. When Exim failed to open the panic log after failing to open the main log,
6865 the original message it was trying to log was written to stderr and debug
6866 output, but if they were not available (the usual case in production), it
6867 was lost. Now it is written to syslog before the two lines that record the
6868 failures to open the logs.
6870 65. Users' Exim filters run in subprocesses under the user's uid. It is
6871 possible for a "deliver" command or an alias in a "personal" command to
6872 provoke an address rewrite. If logging of address rewriting is configured,
6873 this fails because the process is not running as root or exim. There may be
6874 a better way of dealing with this, but for the moment (because 4.30 needs
6875 to be released), I have disabled address rewrite logging when running a
6876 filter in a non-root, non-exim process.
6882 1. The buildconfig auxiliary program wasn't quoting the value set for
6883 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6884 not defined. This bug was masked in 4.22 by the effect that was fixed in
6887 2. Some messages that were rejected after a message id was allocated were
6888 shown as "incomplete" by exigrep. It no longer does this for messages that
6889 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6891 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6892 have allow_domain_literals set, the ID did not get logged in the <= line.
6893 Domain literals are now always recognized in Message-ID: header lines.
6895 4. The first argument for a ${extract expansion item is the key name or field
6896 number. Leading and trailing spaces in this item were not being ignored,
6897 causing some misleading effects.
6899 5. When deliver_drop_privilege was set, single queue runner processes started
6900 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6901 same command in the process it spins off) were not dropping privilege.
6903 6. When the daemon running as "exim" started a queue runner, it always
6904 re-executed Exim in the spun-off process. This is a waste of effort when
6905 deliver_drop_privilege is set. The new process now just calls the
6906 queue-runner function directly.
6912 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6915 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6916 ignored. Though the use of -oP was forcing the writing of a pid file, it
6917 was always written to the default place.
6919 3. If the message "no IP address found for host xxxx" is generated during
6920 incoming verification, it is now followed by identification of the incoming
6921 connection (so you can more easily find what provoked it).
6923 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6925 5. Added some features to "harden" Exim a bit more against certain attacks:
6927 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6928 be put in Local/Makefile. This is like the never_users runtime option,
6929 but it cannot be overridden. The default setting is "root".
6931 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6932 prefix string with which any file named in a -C command line option
6935 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6936 is retained for -C and -D only if the caller of Exim is root. Without
6937 it, the exim user may also use -C and -D and retain privilege.
6939 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6940 command line option is disabled.
6942 6. Macro names set by the -D option must start with an upper case letter, just
6943 like macro names defined in the configuration file.
6945 7. Added "dereference=" facility to LDAP.
6947 8. Two instances of the typo "uknown" in the source files are fixed.
6949 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6950 the Configure-Makefile script screwed up while processing it.
6952 10. Incorporated PCRE 4.4.
6954 11. The SMTP synchronization check was not operating right at the start of an
6955 SMTP session. For example, it could not catch a HELO sent before the client
6956 waited for the greeting. There is now a check for outstanding input at the
6957 point when the greeting is written. Because of the duplex, asynchronous
6958 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6959 way, but not yet received, when the check is performed.
6961 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6962 on TCP/IP sockets, because this apparently causes some broken clients to
6965 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6966 unchanged) from the Cygwin maintainer.
6968 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6969 was supported instead of testing for mbx. Effectively a typo.
6971 15. The spa authenticator server code was not checking that the input it
6972 received was valid base64.
6974 16. The debug output line for the "set" modifier in ACLs was not showing the
6975 name of the variable that was being set.
6977 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6979 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6980 lookup on its own. Something else was needed to trigger the lookup. For
6981 example, a match in host_lookup or the need for a host name in a host list.
6982 Now, if $sender_host_name is referenced and the host name has not yet been
6983 looked up, a lookup is performed. If the lookup fails, the variable remains
6984 empty, and $host_lookup_failed is set to "1".
6986 19. Added "eqi" as a case-independent comparison operator.
6988 20. The saslauthd authentication condition could segfault if neither service
6989 nor realm was specified.
6991 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6992 error message that was logged was misleading, and incoming SMTP
6993 connections were dropped. The message is now more accurate, and temporary
6994 errors are given to SMTP connections.
6996 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6997 (except RSET) with a 421 error, until QUIT is received. However, it was
6998 failing to send a response to QUIT.
7000 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7001 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7002 ACL is now run after the helo_try_verify_hosts code.
7004 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7005 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7006 case-independent, so other case variants are also recognized). Apparently
7007 some systems use these upper case variants.
7009 25. If more than two messages were waiting for the same host, and a transport
7010 filter was specified for the transport, Exim sent two messages over the
7011 same TCP/IP connection, and then failed with "socket operation on non-
7012 socket" when it tried to send the third.
7014 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7016 27. The extern definition of crypt16() in expand.c was not being excluded when
7017 the OS had its own crypt16() function.
7019 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7020 as a preferred synonym for return_size_limit, both as an option and as an
7023 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7025 30. Changed the default configuration ACL to relax the local part checking rule
7026 for addresses that are not in any local domains. For these addresses,
7027 slashes and pipe symbols are allowed within local parts, but the sequence
7028 /../ is explicitly forbidden.
7030 31. SPA server authentication was not clearing the challenge buffer before
7033 32. log_message in a "warn" ACL statement was writing to the reject log as
7034 well as to the main log, which contradicts the documentation and doesn't
7035 seem right (because no rejection is happening). So I have stopped it.
7037 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7038 However, I am unable to do any testing of this.
7040 34. Fixed an infelicity in the appendfile transport. When checking directories
7041 for a mailbox, to see if any needed to be created, it was accidentally
7042 using path names with one or more superfluous leading slashes; tracing
7043 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7045 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7046 contents are added to the log line that is written for every discarded
7047 recipient. (Previously a log_message setting was ignored.)
7049 36. The ${quote: operator now quotes the string if it is empty.
7051 37. The install script runs exim in order to find its version number. If for
7052 some reason other than non-existence or emptiness, which it checks, it
7053 could not run './exim', it was installing it with an empty version number,
7054 i.e. as "exim-". This error state is now caught, and the installation is
7057 38. An argument was missing from the function that creates an error message
7058 when Exim fails to connect to the socket for saslauthd authentication.
7059 This could cause Exim to crash, or give a corrupted message.
7061 39. Added isip, isip4, and isip6 to ${if conditions.
7063 40. The ACL variables $acl_xx are now saved with the message, and can be
7064 accessed later in routers, transports, and filters.
7066 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7067 strings in the file are not string-expanded.
7069 42. If a MAIL command specified a SIZE value that was too large to fit into an
7070 int variable, the check against message_size_limit failed. Such values are
7071 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7072 day this will have to be increased, but I don't think I want to be around
7073 when emails are that large.
7080 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7081 iconv() is not standard in FreeBSD.
7083 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7084 IPv6 enabled. The observed symptom was a segmentation fault on return from
7085 the function os_common_find_running_interfaces() in src/os.c.
7087 3. In the check_special_case() function in daemon.c I had used "errno" as an
7088 argument name, which causes warnings on some systems. This was basically a
7089 typo, since it was named "eno" in the comments!
7091 4. The code that waits for the clock to tick (at a resolution of some fraction
7092 of a second) so as to ensure message-id uniqueness was always waiting for
7093 at least one whole tick, when it could have waited for less. [This is
7094 almost certainly not relevant at current processor speeds, where it is
7095 unlikely to ever wait at all. But we try to future-proof.]
7097 5. The function that sleeps for a time interval that includes fractions of a
7098 second contained a race. It did not block SIGALRM between setting the
7099 timer, and suspending (a couple of lines later). If the interval was short
7100 and the sigsuspend() was delayed until after it had expired, the suspension
7101 never ended. On busy systems this could lead to processes getting stuck for
7104 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7105 process, before it forks any delivery processes. The open lookup caching
7106 mechanism meant that the open file or database connection was passed into
7107 the delivery process. The problem was that delivery processes always tidy
7108 up cached lookup data. This could cause a problem for the next delivery
7109 process started by the queue runner, because the external queue runner
7110 process does not know about the closure. So the next delivery process
7111 still has data in the lookup cache. In the case of a file lookup, there was
7112 no problem because closing a file descriptor in a subprocess doesn't affect
7113 the parent. However, if the lookup was caching a connection to a database,
7114 the connection was closed, and the second delivery process was likely to
7115 see errors such as "PGSQL: query failed: server closed the connection
7116 unexpectedly". The problem has been fixed by closing all cached lookups
7117 in a queue runner before running a delivery process.
7119 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7120 seem to have the "const" qualifier which it has on other OS. I've
7123 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7124 *of the same type* (client or server) with the same public name that an
7125 error should be diagnosed.
7127 9. When Exim looked up a host name for an IP address, but failed to find the
7128 original IP address when looking up the host name (a safety check), it
7129 output the message "<ip address> does not match any IP for NULL", which was
7130 confusing, to say the least. The bug was that the host name should have
7131 appeared instead of "NULL".
7133 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7134 user that is built into the binary, and the -C or -D options is used, root
7135 privilege is dropped before the configuration file is read. In addition,
7136 logging is switched to stderr instead of the normal log files. If the
7137 configuration then re-defines the Exim user, the unprivileged environment
7138 is probably not what is expected, so Exim logs a panic warning message (but
7141 However, if deliver_drop_privilege is set, the unprivileged state may well
7142 be exactly what is intended, so the warning has been cut out in that case,
7143 and Exim is allowed to try to write to its normal log files.
7149 1. smtp_return_error_details was not giving details for temporary sender
7150 or receiver verification errors.
7152 2. Diagnose a configuration error if two authenticators have the same public
7155 3. Exim used not to create the message log file for a message until the first
7156 delivery attempt. This could be confusing when incoming messages were held
7157 for policy or load reasons. The message log file is now created at the time
7158 the message is received, and an initial "Received" line is written to it.
7160 4. The automatically generated man page for command line options had a minor
7161 bug that caused no ill effects; however, a more serious problem was that
7162 the procedure for building the man page automatically didn't always
7163 operate. Consequently, release 4.20 contains an out-of-date version. This
7164 shouldn't happen again.
7166 5. When building Exim with embedded Perl support, the script that builds the
7167 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7168 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7170 6. The freeze_tell option was not being used for messages that were frozen on
7171 arrival, either by an ACL or by local_scan().
7173 7. Added the smtp_incomplete_transaction log selector.
7175 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7176 was accepting AUTH without a new EHLO.
7178 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7179 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7182 10. Logging of TCP/IP connections (when configured) now happens in the main
7183 daemon process instead of the child process, so that the TCP/IP connection
7184 count is more accurate (but it can never be perfect).
7186 11. The use of "drop" in a nested ACL was not being handled correctly in the
7187 outer ACL. Now, if condition failure induced by the nested "drop" causes
7188 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7189 or "require"), the connection is dropped.
7191 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7192 that yield "discard" can now be used with an "accept" or a "discard" verb,
7193 but an error is generated for any others (because I can't see a useful way
7194 to define what should happen).
7196 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7197 are now processed in the same way as lines in the Exim configuration file.
7198 In particular, continuation lines are supported.
7200 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7202 15. Added -ti meaning -t -i.
7204 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7205 domains, and warn by logging if others are found.
7207 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7208 returned from GIFCONF when Exim is trying to find the list of interfaces on
7209 a host. The code in os.c has been modified to copy each ifreq to an aligned
7210 structure in all cases.
7212 Also, in some cases, the returned ifreq's were being copied to a 'struct
7213 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7214 means the last couple of bytes of an IPv6 address could be chopped if the
7215 ifreq contained only a normal sockaddr (14 bytes storage).
7217 18. Named domain lists were not supported in the hosts_treat_as_local option.
7218 An entry such as +xxxx was not recognized, and was treated as a literal
7221 19. Ensure that header lines added by a DATA ACL are included in the reject log
7222 if the ACL subsequently rejects the message.
7224 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7225 MD5 (which is deprecated).
7227 21. When testing a filter file using -bf, Exim was writing a message when it
7228 took the sender from a "From " line in the message, but it was not doing so
7229 when it took $return_path from a Return-Path: header line. It now does.
7231 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7232 with a valid header line field name (a series of printing characters
7233 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7235 23. Changed "disc" in the source to "disk" to conform to the documentation and
7236 the book and for uniformity.
7238 24. Ignore Sendmail's -Ooption=value command line item.
7240 25. When execve() failed while trying to run a command in a pipe transport,
7241 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7242 could be confused with a return value of 69 from the command itself. This
7243 has been changed to 127, the value the shell returns if it is asked to run
7244 a non-existent command. The wording for the related log line suggests a
7245 non-existent command as the problem.
7247 26. If received_header_text expands to an empty string, do not add a Received:
7248 header line to the message. (Well, it adds a token one on the spool, but
7249 marks it "old" so that it doesn't get used or transmitted.)
7251 27. Installed eximstats 1.28 (addition of -nt option).
7253 28. There was no check for failure on the call to getsockname() in the daemon
7254 code. This can fail if there is a shortage of resources on the system, with
7255 ENOMEM, for example. A temporary error is now given on failure.
7257 29. Contrary to the C standard, it seems that in some environments, the
7258 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7259 program. Exim now does this explicitly; it affects the formatting of
7260 timestamps using strftime().
7262 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7263 complaints. I've now initialized all the variables, to avoid this.
7265 32. Header lines added by a system filter were not being "seen" during
7266 transport-time rewrites.
7268 33. The info_callback() function passed to OpenSSL is set up with type void
7269 (*)(SSL *, int, int), as described somewhere. However, when calling the
7270 function (actually a macro) that sets it up, the type void(*)() is
7271 expected. I've put in a cast to prevent warnings from picky compilers.
7273 34. If a DNS black list lookup found a CNAME record, but there were no A
7274 records associated with the domain it pointed at, Exim crashed.
7276 35. If a DNS black list lookup returned more than one A record, Exim ignored
7277 all but the first. It now scans all returned addresses if a particular IP
7278 value is being sought. In this situation, the contents of the
7279 $dnslist_value variable are a list of all the addresses, separated by a
7282 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7283 to accept a host name and all its aliases if the forward lookup for any of
7284 them yielded the IP address of the incoming connection. Now it accepts only
7285 those names whose forward lookup yields the correct IP address. Any other
7286 names are discarded. This closes a loophole whereby a rogue DNS
7287 administrator could create reverse DNS records to break through a
7288 wildcarded host restriction in an ACL.
7290 37. If a user filter or a system filter that ran in a subprocess used any of
7291 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7292 the wrong values were passed to the pipe command ($thisaddress had the
7293 value of $0, $0 had the value of $1, etc). This bug was introduced by
7294 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7296 38. Improved the line breaking for long SMTP error messages from ACLs.
7297 Previously, if there was no break point between 40 and 75 characters, Exim
7298 left the rest of the message alone. Two changes have been made: (a) I've
7299 reduced the minimum length to 35 characters; (b) if it can't find a break
7300 point between 35 and 75 characters, it looks ahead and uses the first one
7301 that it finds. This may give the occasional overlong line, but at least the
7302 remaining text gets split now.
7304 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7305 file descriptors might be low, and that setting 1000 would always raise it.
7306 It turns out that in some environments, the limit is already over 1000 and
7307 that lowering it causes trouble. So now Exim takes care not to decrease it.
7309 40. When delivering a message, the value of $return_path is set to $sender_
7310 address at the start of routing (routers may change the value). By an
7311 oversight, this default was not being set up when an address was tested by
7312 -bt or -bv, which affected the outcome if any router or filter referred to
7315 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7316 log only when "message" or "log_message" are set. However, if one of the
7317 conditions was an address verification, or a call to a nested ACL, the
7318 messages generated by the underlying test were being passed through. This
7319 no longer happens. The underlying message is available in $acl_verify_
7320 message for both "message" and "log_message" expansions, so it can be
7321 passed through if needed.
7323 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7324 new expansion $bh_ to give the encoded byte string without charset
7325 translation. Translation happens only if iconv() is available; HAVE_ICONV
7326 indicates this at build time. HEADERS_CHARSET gives the charset to
7327 translate to; headers_charset can change it in the configuration, and
7328 "headers charset" can change it in an individual filter file.
7330 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7331 that creates RFC 2047 encoded "words" labels them as that charset instead
7332 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7333 expansion operator; (ii) when Exim creates a From: line for a local
7334 message; (iii) when a header line is rewritten to include a "phrase" part.
7336 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7337 buggy, causing it to skip the first lines of messages whose message ID
7338 ended in 'D'. This would not have bitten before Exim release 4.14, because
7339 message IDs were unlikely to end in 'D' before then. The effect was to have
7340 incorrect size information for certain domains.
7342 45. #include "config.h" was missing at the start of the crypt16.c module. This
7343 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7346 46. If there was a timeout during a "random" callout check, Exim treated it as
7347 a failure of the random address, and carried on sending RSET and the real
7348 address. If the delay was just some slowness somewhere, the response to the
7349 original RCPT would be taken as a response to RSET and so on, causing
7350 mayhem of various kinds.
7352 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7353 when I implemented it. It didn't allow for the fact that some option values
7354 may legitimately be negative (e.g. size_addition), and it didn't even do
7355 the right test for positive values.
7357 48. Domain names in DNS records are case-independent. Exim always looks them up
7358 in lower case. Some resolvers return domain names in exactly the case they
7359 appear in the zone file, that is, they may contain uppercase letters. Not
7360 all resolvers do this - some return always lower case. Exim was treating a
7361 change of case by a resolver as a change of domain, similar to a widening
7362 of a domain abbreviation. This triggered its re-routing code and so it was
7363 trying to route what was effectively the same domain again. This normally
7364 caused routing to fail (because the router wouldn't handle the domain
7365 twice). Now Exim checks for this case specially, and just changes the
7366 casing of the domain that it ultimately uses when it transmits the message
7369 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7372 50. If a filter generated a file delivery with a non-absolute name (possible if
7373 no home directory exists for the router), the forbid_file option was not
7376 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7377 the existing equality matching.
7379 52. Exim was using ints instead of ino_t variables in some places where it was
7380 dealing with inode numbers.
7382 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7383 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7384 called TMPDIR, and if it finds it is different, it changes its value.
7386 54. The smtp_printf() function is now made available to local_scan() so
7387 additional output lines can be written before returning. There is also an
7388 smtp_fflush() function to enable the detection of a dropped connection.
7389 The variables smtp_input and smtp_batched_input are exported to
7392 55. Changed the default runtime configuration: the message "Unknown user"
7393 has been removed from the ACL, and instead placed on the localuser router,
7394 using the cannot_route_message feature. This means that any verification
7395 failures that generate their own messages won't get overridden. Similarly,
7396 the "Unrouteable address" message that was in the ACL for unverifiable
7397 relay addresses has also been removed.
7399 56. Added hosts_avoid_esmtp to the smtp transport.
7401 57. The exicyclog script was not checking for the esoteric option
7402 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7403 will work only if exicyclog is run under the appropriate euid.
7405 58. Following a discussion on the list, the rules by which Exim recognises line
7406 endings on incoming messages have been changed. The -dropcr and drop_cr
7407 options are now no-ops, retained only for backwards compatibility. The
7408 following line terminators are recognized: LF CRLF CR. However, special
7409 processing applies to CR:
7411 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7412 nor a local message in the state where . is a terminator.
7414 (ii) If a bare CR is encountered in a header line, an extra space is added
7415 after the line terminator so as not to end the header. The reasoning
7416 behind this is that bare CRs in header lines are most likely either
7417 to be mistakes, or people trying to play silly games.
7419 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7420 was being incorrectly given as 18 bytes larger than it should have been.
7421 This is a VOB (very old bug).
7423 60. This may never have affected anything current, but just in case it has:
7424 When the local host is found other than at the start of a list of hosts,
7425 the local host, those with the same MX, and any that follow, are discarded.
7426 When the list in question was part of a longer list of hosts, the following
7427 hosts (not currently being processed) were also being discarded. This no
7428 longer happens. I'm not sure if this situation could ever has previously
7431 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7434 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7435 header. This is something that is recommended in a new Internet Draft, and
7436 is something that is documented as being done by Sendmail. There are two
7437 possible values. For messages generated by the autoreply transport, Exim
7440 Auto-Submitted: auto-replied
7442 whereas for all other generated messages (e.g. bounces) it adds
7444 Auto-Submitted: auto-generated
7446 63. The "personal" condition in filters now includes a test for the
7447 Auto-Submitted: header. If it contains the string "auto-" the message it
7448 not considered personal.
7450 64. Added rcpt_include_affixes as a generic transport option.
7452 65. Added queue_only_override (default true).
7454 66. Added the syslog_duplication option.
7456 67. If what should have been the first header line of a message consisted of
7457 a space followed by a colon, Exim was mis-interpreting it as a header line.
7458 It isn't of course - it is syntactically invalid and should therefore be
7459 treated as the start of the message body. The misbehaviour could have
7460 caused a number of strange effects, including loss of data in subsequent
7461 header lines, and spool format errors.
7463 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7464 client host had authenticated. This control can now be exercised by an ACL
7465 for more flexibility.
7467 69. By default, callouts do not happen when testing with -bh. There is now a
7468 variant, -bhc, which does actually run the callout code, including
7469 consulting and updating the callout cache.
7471 70. Added support for saslauthd authentication, courtesy of Alexander
7474 71. If statvfs() failed on the spool or log directories while checking their
7475 size for availability, Exim confusingly gave the error "space shortage".
7476 Furthermore, in debugging mode it crashed with a floating point exception.
7477 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7478 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7479 serious problem, Exim now writes to the main and panic logs when this
7480 happens, with details of the failure. It then refuses to accept the
7481 incoming message, giving the message "spool directory problem" or "log
7482 directory problem" with a 421 code for SMTP messages.
7484 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7485 0, 1, and 2 exist, because some OS complain for execs without them (see
7486 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7487 descriptors. However, the code omitted to check that the open succeeded,
7488 causing mysterious errors if for some reason the permissions on /dev/null
7489 got screwed. Now Exim writes a message to the main and panic logs, and
7490 bombs out if it can't open /dev/null.
7492 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7493 interact so that it is all more flexible. It is supposed to remain
7494 backwards compatible. Also added extra_local_interfaces.
7496 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7497 to bomb out with an assertion failure - to the client this appears as a
7498 connection drop. This problem occurs in the part of the code that was taken
7499 from the Samba project. Fortunately, the assertion is in a very simple
7500 function, so I have fixed this by reproducing the function inline in the
7501 one place where it is called, and arranging for authentication to fail
7502 instead of killing the process with assert().
7504 75. The SPA client code was not working when the server requested OEM rather
7505 than Unicode encoding.
7507 76. Added code to make require_files with a specific uid setting more usable in
7508 the case where statting the file as root fails - usually a non-root-mounted
7509 NFS file system. When this happens and the failure is EACCES, Exim now
7510 forks a subprocess and does the per-uid checking as the relevant uid.
7512 77. Added process_log_path.
7514 78. If log_file_path was not explicitly set, a setting of check_log_space or
7515 check_log_inodes was ignored.
7517 79. If a space check for the spool or log partitions fails, the incident is now
7518 logged. Of course, in the latter case the data may get lost...
7520 80. Added the %p formatting code to string_format() so that it can be used to
7521 print addresses in debug_print(). Adjusted all the address printing in the
7522 debugging in store.c to use %p rather than %d.
7524 81. There was a concern that a line of code in smtp_in.c could overflow a
7525 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7526 initially expressed, the concern was not well-founded, because trailing
7527 spaces are removed early. However, if the trailing spaces were followed by
7528 a NULL, they did not get removed, so the overflow was possible. Two fixes
7531 (a) I re-wrote the offending code in a cleaner fashion.
7532 (b) If an incoming SMTP command contains a NULL character, it is rejected
7535 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7536 runs initgroups(), so that if the Exim user is in any additional groups,
7537 they will be used during message reception.
7543 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob