1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
9 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
10 related applications. Previously an "H" was used where available info
11 says that "M" should be, so change to match.
13 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
14 as arguments, so an implementation trying to copy these into a local
15 buffer was taking a taint-enforcement trap. Fix by using dynamically
16 created buffers. Similar fix for radius expansion condition.
18 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
19 reasonable, eg. to count headers. Fix by using dynamically created
20 buffers rather than a local. Do similar fixes for ACL actions "dcc",
21 "log_reject_target", "malware" and "spam"; the arguments are expanded
22 so could be handling tainted values.
24 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
25 broken the (no-op) support for this sendmail command. Restore it
26 to doing nothing, silently, and returning good status.
28 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
29 record path was given (or the default used) without a leading directory
30 path, an error occurred on trying to open it. Use the transport's working
33 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
34 transport. Previously it was the name on the DNS A-record; use instead
35 the head of the CNAME chain leading there (if there is one). This seems
36 to align better with RFC 6125.
38 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
39 smtp_accept_max_per_host allocated resources which were not released
40 when the limit was exceeded. This eventually crashed the daemon. Fix
41 by adding a relase action in that path.
43 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
44 expanded; previously using tainted values was rejected. Fix by using
45 dynamically-created buffers.
47 JH/09 Relax restrictions on ACL verify condition needing access to message
48 headers. Previously they were only permitted in data and non-smtp ACLs;
49 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
50 not_blind, header_sender and header_names_ascii verification.
52 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
53 Previously a macro used one argument twice; when called with the
54 argument as an expression having side-effects, incorrect operation
55 resulted. Use an inlineable function.
57 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
58 held open for a verify callout. Previously this wan not accounted for
59 and a corrupt onward SMTP conversation resulted.
61 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
62 Previously they were embedded naively in the constructed address; when
63 needed, strip the quoting and quote the entire local_part.
64 Also make the inbound_srs expansion condition handle quoting.
66 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
67 excluded, not matching the documentation.
69 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
70 was given for the sqlite_dbfile a trap resulted.
72 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
73 "name" argument resulted in a trap. There is no reason to disallow such;
74 this was a coding error.
76 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
77 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
78 spent suspended, ignoring the Posix definition. Previously we assumed
79 it did and a constant offset from real time could be used as a correction.
80 Change to using the same clock source for the start-of-message and the
81 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
82 exists, just to get a clock slightly more aligned to reality.
84 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
85 RFC says it is optional some validators care. The missing char was not
86 intended but triggered by a line-wrap alignement. Discovery and fix by
87 Guillaume Outters, hacked on by JH.
89 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
90 name being quoted was tainted a trap would be taken. Fix by using
91 dynamicaly created buffers. The routine could have been called by a
92 rewrite with the "h" flag, by using the "-F" command-line option, or
93 by using a "name=" option on a control=submission ACL modifier.
95 JH/19 SPF: change the Authentication-Results expansion component to give
96 smtp.helo when the sender domain is empty. Previously it gave
103 JH/01 Avoid costly startup code when not strictly needed. This reduces time
104 for some exim process initialisations. It does mean that the logging
105 of TLS configuration problems is only done for the daemon startup.
107 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
109 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
110 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
113 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
114 DKIM signing is being done. Previously a transport_filter would always
115 disable CHUNKING, falling back to traditional DATA.
117 JH/05 Regard command-line receipients as tainted.
119 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
121 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
122 PAM library frees one of the arguments given to it, despite the
123 documentation. Therefore a plain malloc must be used.
125 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
126 on-stack buffers were used, resulting in a taint trap when DSN information
127 copied from a received message was written into the buffer.
129 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
130 the ordering of its ARC headers. This caused a crash.
132 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
133 a new record was being constructed with information from the peer, a trap
136 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
137 installation would get error messages from DMARC verify, when it hit the
138 nonexistent file indicated by the default. Distros wanting DMARC enabled
139 should both provide the file and set the option.
140 Also enforce no DMARC verification for command-line sourced messages.
142 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
143 could, depending on the platform, hang at the STARTTLS response.
145 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
146 message on a connection. Previously if one message had ARC headers and
147 the following one did not, a crash could result when adding an
148 Authentication-Results: header.
150 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
151 Exim main code and Exim-related utities. The introduction of taint
152 tracking also did many adjustments to string handling. Since then, eximon
153 frequently terminated with an assert failure.
155 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
156 check for 452 responses. This slightly helps the inefficieny of doing
157 a large alias-expansion into a recipient-limited target. The max_rcpt
158 transport option still applies (and at the current default, will override
159 the new feature). The check is done for either cause of synch, and forces
160 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
161 connection. The new facility is not tunable at this time.
163 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
164 library live data was being used, so the results became garbage. Make
165 copies while it is still usable.
167 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
168 on delivery deferred (==) and failed (**) lines (if a delivery was
169 attemtped). Previously it was only on completion (=>) lines.
171 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
172 for the expansion of the server_scram_iter and server_scram_salt options.
174 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
175 are now specifically given a NO_DATA response without hitting the system
176 resolver. The library goes on to do the now-standard TXT lookup.
177 Use of dnsdb lookups is not affected.
179 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
180 only retrieve the errormessage once. Previously two calls to dlerror()
181 were used, and the second one (for mainlog/paniclog) retrieved null
184 JH/20 Taint checking: disallow use of tainted data for
185 - the appendfile transport file and directory options
186 - the pipe transport command
187 - the autoreply transport file, log and once options
188 - file names used by the redirect router (including filter files)
190 - paths used by single-key lookups
191 Previously this was permitted.
193 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
194 adjusted the size of a major service buffer; this failed because the
195 buffer was in use at the time. Change to a compile-time increase in the
196 buffer size, when this authenticator is compiled into exim.
198 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
199 previous fast-mode was untenable in the face of glibs using mmap to
200 support larger malloc requests.
202 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
203 New values supported, if defined on system where compiled:
204 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
205 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
207 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
208 running a limited number of proceses in parallel, a benefit is gained. The
209 amount varies with the platform hardware and load. The use of the option
210 queue_run_in_order means we cannot do this, as ordering becomes
213 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
214 had introduced a string-copy (for ensuring NUL-termination) which was not
215 appropriate for that case, which can include embedded NUL bytes in the
216 block of data. Investigation showed the copy to actually be needless, the
217 data being length-specified.
219 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
220 done during a receiving connection, and both used TLS, global info was
221 used rather than per-connection info for tracking the state of data
222 queued for transmission. This could result in a connection hang.
224 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
225 Previously, when delivering serveral messages down a single connection
226 only the first would provide a SIZE. This was due to the size information
227 not being properly tracked.
229 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
230 TAI (at 37 seconds currently), pretend to be in UTC for time-related
231 expansion and logging. Previously, spurious values such as a future
232 minute could be seen.
234 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
235 it could crash from a null-deref. This could also affect the
236 ${addresses: } operator and ${readsock } item.
238 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
239 message following a mime one, the variable was not reset.
241 JH/30 When an pipelined-connect fails at the first response, assume incorrect
242 cached capability (perhaps the peer reneged?) and immediately retry in
245 JH/31 Fix spurious detection of timeout while writing to transport filter.
247 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
248 an attempt to copy the string was made before checking it.
250 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
251 taint of the lookup key was maintained; we now regard the presence in the
252 filesystem as sufficient validation.
254 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
255 argument is supplied.
257 JH/35 The dsearch lookup now requires that the directory is an absolute path.
258 Previously this was not checked, and nonempty relative paths made an
259 access under Exim's current working directory.
261 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
262 Previously no event was raised.
264 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
265 parameter supplied by the sender MAIL FROM command. Previously it was
266 ignored, and only the check_spool_space option value for the required
269 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
270 the size of the signing public-key. Previously it was instead giving
271 the size of the signature hash.
273 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
274 the default. See the (new) dkim_verify_min_keysizes option.
276 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
277 and an ACL use a lookup for checking either the local_part or domain,
278 stale data could be accessed. Ensure that variable references are
279 dropped between messages.
281 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
282 by the client was not checked as pointing within response data before
283 being used. A malicious client could thus cause an out-of-bounds read and
284 possibly gain authentication. Fix by adding the check.
286 JH/42 Internationalisation: change the default for downconversion in the smtp
287 transport to be "if needed". Previously it was "as previously set" for
288 the message, which usually meant "if needed" for message-submission but
289 "no" for everything else. However, MTAs have been seen using SMTPUTF8
290 even when the envelope addresses did not need it, resulting in forwarding
291 failures to non-supporting MTAs. A downconvert in such cases will be
292 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
293 transport. The change does mean that addresses needing conversion will
294 be converted when previously a delivery failure would occur.
296 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
297 response was received it would be used unchecked in a fail-DSN, violating
298 standards on line-length limits. Truncate if needed.
300 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
301 doesn't work. (Additionally add a new main config option to configure the
308 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
309 side, to decode a TLS 1.3 packet capture.
311 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
312 Previously the default library behaviour applied, sending two, each in
315 JH/03 Debug output for ACL now gives the config file name and line number for
318 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
320 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
322 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
323 buffer overrun for (non-chunking) other transports.
325 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
326 TLS1.3, means that a server rejecting a client certificate is not visible
327 to the client until the first read of encrypted data (typically the
328 response to EHLO). Add detection for that case and treat it as a failed
329 TLS connection attempt, so that the normal retry-in-clear can work (if
330 suitably configured).
332 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
333 and/or domain. Found and fixed by Jason Betts.
335 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
336 configuration). If a CNAME target was not a wellformed name pattern, a
339 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
340 the OS reports them interleaved with other addresses.
342 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
343 used both for input and for a verify callout, both encrypted, SMTP
344 responses being sent by the server could be lost. This resulted in
345 dropped connections and sometimes bounces generated by a peer sending
348 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
349 string. Previously it was possible to cause undefined behaviour in a
350 library routine (usually a crash). Found by "zerons".
352 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
355 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
356 API was removed, so update to use the newer ones.
358 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
359 any timeout set, is taking a long time. Previously we would hang on to a
360 rotated logfile "forever" if the input was arriving with long gaps
361 (a previous attempt to fix addressed lack, for a long time, of initial
364 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
365 shared (NFS) environment. The length of the tempfile name is now
366 4 + 16 ("hdr.$message_exim_id") which might break on file
367 systems which restrict the file name length to lower values.
370 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
371 shared (NFS) environment.
373 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
374 did for all versions <4.90). Notably -M, -m, --invert, -I may be
377 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
378 on some platforms for bit 31.
380 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
381 to changes apparently associated with TLS1.3 handling some of the APIs
382 previously used were either nonfunctional or inappropriate. Strings
383 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
384 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
385 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
386 This affects log line X= elements, the $tls_{in,out}_cipher variables,
387 and the use of specific cipher names in the encrypted= ACL condition.
389 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
391 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
392 verification result was not updated unless hosts_require_ocsp applied.
394 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
395 queue_list_requires_admin set to false, non-admin users were denied the
398 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
399 directory-of-certs mode. Previously they were advertised despite the
402 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
403 A single TCP connection by a client will now hold a TLS connection open
404 for multiple message deliveries, by default. Previously the default was to
407 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
408 default. If built with the facility, DANE will be used. The facility
409 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
411 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
412 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
413 must be defined and you must still, unless you define DISABLE_TLS, manage
414 the the include-dir and library-file requirements that go with that
415 choice. Non-TLS builds are still supported.
417 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
420 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
421 default. If the platform supports and has the facility enabled, it will
422 be requested on all coneections.
424 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
425 controlled by the build-time option SUPPORT_PIPE_CONNECT.
427 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
429 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
430 success-DSN messages. Previously the From: header was always the default
431 one for these; the option was ignored.
433 JH/28 Fix the timeout on smtp response to apply to the whole response.
434 Previously it was reset for every read, so a teergrubing peer sending
435 single bytes within the time limit could extend the connection for a
436 long time. Credit to Qualsys Security Advisory Team for the discovery.
438 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
439 delivery address, which leaked information of the results of local
440 forwarding. Change to the original envelope recipient address, per
443 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
444 requested. Previously not bounce was generated and a log entry of
445 error ignored was made.
447 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
449 JH/32 Introduce a general tainting mechanism for values read from the input
450 channel, and values derived from them. Refuse to expand any tainted
451 values, to catch one form of exploit.
453 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
454 was unused and the unexpanded text used for the test. Found and
455 fixed by Ruben Jenster.
457 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
458 an attempt to use a TLS library read routine dereffed a nul pointer,
461 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
462 them in our smtp response.
464 JH/36 Have the general router option retry_use_local_part default to true when
465 any of the restrictive preconditions are set (to anything). Previously it
466 was only for check_local user. The change removes one item of manual
467 configuration which is required for proper retries when a remote router
468 handles a subset of addresses for a domain.
470 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
471 link count into consideration.
473 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
474 caused the extension of big_buffer, the following lines were ignored.
476 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
477 accordance with RFC 2308. Previously there was no expiry, so a longlived
478 receive process (eg. due to ACL delays) versus a short SOA value could
481 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
483 JH/39 Promote DMARC support to mainline.
485 JH/40 Bug 2452: Add a References: header to DSNs.
487 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
488 parameters. The relevant library call is documented as "Deprecated: This
489 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
490 3.6.0, DH parameters are negotiated following RFC7919."
492 HS/06 Change the default of dnssec_request_domains to "*"
494 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
495 carried on and emitted a BDAT command, even when PIPELINING was not
498 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
499 buffer was used for the filename, resulting in a trap when tainted
500 arguments (eg. $domain) were used.
502 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
503 recommended to avoid a possible server-load attack. The feature can be
504 re-enabled via the openssl_options main cofiguration option.
506 JH/45 local_scan API: documented the current smtp_printf() call. This changed
507 for version 4.90 - adding a "more data" boolean to the arguments.
508 Bumped the ABI version number also, this having been missed previously;
509 release versions 4.90 to 4.92.3 inclusive were effectively broken in
510 respect of usage of smtp_printf() by either local_scan code or libraries
511 accessed via the ${dlfunc } expansion item. Both will need coding
512 adjustment for any calls to smtp_printf() to match the new function
513 signature; a FALSE value for the new argument is always safe.
515 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
516 the file-offset (which the Linux syscall does, and exim expects); this
517 resulted in an indefinite loop.
519 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
520 to do ARC verification. The Authentication-Results: header line added
521 by the configuration then had no ARC item.
527 JH/01 Remove code calling the customisable local_scan function, unless a new
528 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
530 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
531 non-signal-safe functions being used.
533 JH/03 Bug 2269: When presented with a received message having a stupidly large
534 number of DKIM-Signature headers, disable DKIM verification to avoid
535 a resource-consumption attack. The limit is set at twenty.
537 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
538 report of oldest_pass in ${authres } in consequence, and separate out
539 some descriptions of reasons for verification fail.
541 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
542 files in the spool were present and unlocked. A queue-runner could spot
543 them, resulting in a duplicate delivery. Fix that by doing the unlock
544 after the unlink. Investigation by Tim Stewart. Take the opportunity to
545 add more error-checking on spoolfile handling while that code is being
548 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
549 No known attacks, no CVE, this is defensive hardening.
551 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
552 a queue-runner could start a delivery while other operations were ongoing.
553 Cutthrough delivery was a common victim, resulting in duplicate delivery.
554 Found and investigated by Tim Stewart. Fix by using the open message data
555 file handle rather than opening another, and not locally closing it (which
556 releases a lock) for that case, while creating the temporary .eml format
557 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
559 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
560 $sender_verify_failure/$recipient_verify_failure to "random".
562 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
565 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
566 Previously this would segfault.
568 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
571 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
572 like zero, since the resolver should be doing this for us, But we need one
573 as a CNAME but no MX presence gets the CNAME returned; we need to check
574 that doesn't point to an MX to declare it "no MX returned" rather than
575 "error, loop". A new main option is added so the older capability of
576 following some limited number of chain links is maintained.
578 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
580 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
581 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
582 now-deprecated earlier definitions used only specified the range up to TLS
583 1.2 (in the older-version library docs).
585 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
587 JH/15 Rework TLS client-side context management. Stop using a global, and
588 explicitly pass a context around. This enables future use of TLS for
589 connections to service-daemons (eg. malware scanning) while a client smtp
590 connection is using TLS; with cutthrough connections this is quite likely.
592 JH/16 Fix ARC verification to do AS checks in reverse order.
594 JH/17 Support a "tls" option on the ${readsocket } expansion item.
596 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
597 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
598 Previously the "utf8" would be re-prepended for every additional message.
600 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
601 Previously thery were accepted, resulting in issues when attempting to
602 forward messages to a non-supporting MTA.
604 PP/02 Let -n work with printing macros too, not just options.
606 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
607 one parent address was copied, and bogus data was used at delivery-logging
608 time. Either a crash (after delivery) or bogus log data could result.
609 Discovery and analysis by Tim Stewart.
611 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
612 Previously if the string ended mid-character, we did not insert the
613 promised '?' replacement.
615 PP/04 Documentation: current string operators work on bytes, not codepoints.
617 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
618 should pack well giving a smaller memory footprint so better caching and
619 therefore performance. Group the declarations where this can't be done so
620 that the byte-sized flag variables are not interspersed among pointer
621 variables, giving a better chance of good packing by the compiler.
623 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
624 non-null, to avoid issues with sites running BATV. Previously reports were
625 sent with an empty envelope sender so looked like bounces.
627 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
628 The ignore_error flag wasn't being returned from the filter subprocess so
629 was not set for later routers. Investigation and fix by Matthias Kurz.
631 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
632 and a msg:complete for the whole, when a message is manually removed using
633 -Mrm. Developement by Matthias Kurz, hacked on by JH.
635 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
636 a "Gnu special" function, asprintf() in the DB utility binary builds; I
637 hope that is portable enough.
639 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
640 requiring a known-CA anchor certificate; make it now rely entirely on the
641 TLSA as an anchor. Checking the name on the leaf cert against the name
642 on the A-record for the host is still done for TA (but not for EE mode).
644 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
645 included in delivery lines for non-proxied connections, when compiled with
646 SUPPORT_SOCKS and running with proxy logging enabled.
648 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
649 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
650 move the existing event to fire before the normal logging of message
651 failure so that custom logging is bracketed by normal logging.
653 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
654 msg:fail:internal event. Developement by Matthias Kurz.
656 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
657 far too small for todays use of crypto signatures stored there. Go all
658 the way to the max DNS message size of 64kB, even though this might be
659 overmuch for IOT constrained device use.
661 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
662 copy a string over itself. The library routine is documented as not
663 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
665 JH/32 For main options check_spool_space and check_inode_space, where the
666 platform supports 64b integers, support more than the previous 2^31 kB
667 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
668 the previous G, M, k.
670 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
671 $authenticated_fail_id variable on authentication failure. Previously
674 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
675 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
676 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
677 GNUTLS_SEC_PARAM_MEDIUM.
679 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
680 side. Previously we would continue as if no SNI had been received.
682 JH/36 Harden the handling of string-lists. When a list consisted of a sole
683 "<" character, which should be a list-separator specification, we walked
684 off past the nul-terimation.
686 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
687 causes) even when the retry time is not yet met. Previously they were
688 not, meaning that when (say) an account was over-quota and temp-rejecting,
689 and multiple senders' messages were queued, only one sender would get
690 notified on each configured delay_warning cycle.
692 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
694 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
695 cipher-suites, an error can be left on the stack even for a succeeding
696 accept; this results in impossible error messages when a later operation
699 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
700 return error codes indicating retry. Under TLS1.3 this becomes required.
702 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
703 it only wrote the new authenticators, resulting in a lack of tracking of
704 peer changes of ESMTP extensions until the next cache flush.
706 JH/41 Fix the loop reading a message header line to check for integer overflow,
707 and more-often against header_maxsize. Previously a crafted message could
708 induce a crash of the recive process; now the message is cleanly rejected.
710 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
711 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
717 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
718 When redis_servers is set to a list of > 1 element, and the Redis servers
719 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
720 case of MOVED into a DEFER case instead, thus moving the query onto the
721 next server in the list. For a cluster of N elements, all N servers must
722 be defined in redis_servers.
724 GF/02 Catch and remove uninitialized value warning in exiqsumm
725 Check for existence of @ARGV before looking at $ARGV[0]
727 JH/01 Replace the store_release() internal interface with store_newblock(),
728 which internalises the check required to safely use the old one, plus
729 the allocate and data copy operations duplicated in both (!) of the
730 extant use locations.
732 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
733 modifier. This matches the restriction on the commandline.
735 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
736 Previously only the last row was returned.
738 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
739 we assumed that tags in the header were well-formed, and parsed the
740 element content after inspecting only the first char of the tag.
741 Assumptions at that stage could crash the receive process on malformed
744 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
745 While running the DKIM ACL we operate on the Permanent memory pool so that
746 variables created with "set" persist to the DATA ACL. Also (at any time)
747 DNS lookups that fail create cache records using the Permanent pool. But
748 expansions release any allocations made on the current pool - so a dnsdb
749 lookup expansion done in the DKIM ACL releases the memory used for the
750 DNS negative-cache, and bad things result. Solution is to switch to the
751 Main pool for expansions.
752 While we're in that code, add checks on the DNS cache during store_reset,
753 active in the testsuite.
754 Problem spotted, and debugging aided, by Wolfgang Breyha.
756 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
757 When none of the hosts presented to a transport match an already-open
758 connection, close it and proceed with the list. Previously we would
759 queue the message. Spotted by Lena with Yahoo, probably involving
762 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
763 Previously a spurious "250 OK id=" response was appended to the proper
766 JH/08 The "support for" informational output now, which built with Content
767 Scanning support, has a line for the malware scanner interfaces compiled
768 in. Interface can be individually included or not at build time.
770 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
771 by the template makefile "src/EDITME". The "STREAM" support for an older
772 ClamAV interface method is removed.
774 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
775 rows affected is given instead).
777 JH/11 The runtime Berkeley DB library version is now additionally output by
778 "exim -d -bV". Previously only the compile-time version was shown.
780 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
781 SMTP connection. Previously, when one had more receipients than the
782 first, an abortive onward connection was made. Move to full support for
783 multiple onward connections in sequence, handling cutthrough connection
784 for all multi-message initiating connections.
786 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
787 routers. Previously, a multi-recipient message would fail to match the
788 onward-connection opened for the first recipient, and cause its closure.
790 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
791 a timeout on read on a GnuTLS initiating connection, resulting in the
792 initiating connection being dropped. This mattered most when the callout
793 was marked defer_ok. Fix to keep the two timeout-detection methods
796 JH/15 Relax results from ACL control request to enable cutthrough, in
797 unsupported situations, from error to silently (except under debug)
798 ignoring. This covers use with PRDR, frozen messages, queue-only and
801 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
803 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
804 metadata, resulting in a crash in free().
806 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
807 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
808 Broken also in d185889f4, with init system revamp.
810 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
811 Previously we abruptly closed the connection after reading a malware-
812 found indication; now we go on to read the "scan ok" response line,
815 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
816 ACL. Previously, a crash would result.
818 JH/19 Speed up macro lookups during configuration file read, by skipping non-
819 macro text after a replacement (previously it was only once per line) and
820 by skipping builtin macros when searching for an uppercase lead character.
822 JH/20 DANE support moved from Experimental to mainline. The Makefile control
823 for the build is renamed.
825 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
826 was allocated for every new TLS startup, meaning one per message. Fix
827 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
829 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
830 reported the original. Fix to report (as far as possible) the ACL
831 result replacing the original.
833 JH/23 Fix memory leak during multi-message connections using STARTTLS under
834 OpenSSL. Certificate information is loaded for every new TLS startup,
835 and the resources needed to be freed.
837 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
839 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
842 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
843 DATA response info to the (existing) per-recipient response info for
844 the "C=" log element. It can have useful tracking info from the
845 destination system. Patch from Simon Arlott.
847 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
848 length value. Previously this would segfault.
850 HS/02 Support Avast multiline protoocol, this allows passing flags to
851 newer versions of the scanner.
853 JH/28 Ensure that variables possibly set during message acceptance are marked
854 dead before release of memory in the daemon loop. This stops complaints
855 about them when the debug_store option is enabled. Discovered specifically
856 for sender_rate_period, but applies to a whole set of variables.
857 Do the same for the queue-runner and queue-list loops, for variables set
858 from spool message files. Do the same for the SMTP per-message loop, for
859 certain variables indirectly set in ACL operations.
861 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
862 as a multi-recipient message from a mailinglist manager). The coding had
863 an arbitrary cutoff number of characters while checking for more input;
864 enforced by writing a NUL into the buffer. This corrupted long / fast
865 input. The problem was exposed more widely when more pipelineing of SMTP
866 responses was introduced, and one Exim system was feeding another.
867 The symptom is log complaints of SMTP syntax error (NUL chars) on the
868 receiving system, and refused recipients seen by the sending system
869 (propating to people being dropped from mailing lists).
870 Discovered and pinpointed by David Carter.
872 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
873 replaced by the ${authresults } expansion.
875 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
877 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
878 allows proper process termination in container environments.
880 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
881 Previously the "final dot" had a newline after it; ensure it is CR,LF.
883 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
884 and "err_perm", deprecated since 4.83 when the RFC-defined words
885 "temperror" and "permerror" were introduced.
887 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
888 transport-filters or DKIM-signing. The restriction was lost in the
889 consolidation of verify-callout and delivery SMTP handling.
890 Extend the restriction to also cover ARC-signing.
892 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
893 in defer=pass mode supply a 450 to the initiator. Previously the message
896 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
897 tls_require_ciphers is used as before.
899 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
900 "pass_unscanned". Only tmpfails from the scanner are written to
901 the paniclog, as they may require admin intervention (permission
902 denied, license issues). Other scanner errors (like decompression
903 bombs) do not cause a paniclog entry.
905 JH/36 Fix reinitialisation of DKIM logging variable between messages.
906 Previously it was possible to log spurious information in receive log
909 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
910 triggered odd behaviour from Outlook Express clients.
912 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
915 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
916 since the IETF WG has not yet settled on that versus the original
917 "bare" representation.
919 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
920 Previously the millisecond value corrupted the output.
921 Fix also for syslog_pid=no and log_selector +pid, for which the pid
922 corrupted the output.
928 JH/01 Rework error string handling in TLS interface so that the caller in
929 more cases is responsible for logging. This permits library-sourced
930 string to be attached to addresses during delivery, and collapses
931 pairs of long lines into single ones.
933 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
934 during configuration. Wildcards are allowed and expanded.
936 JH/02 Rework error string handling in DKIM to pass more info back to callers.
937 This permits better logging.
939 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
940 do not close it down and have the child transport start it up again on
941 the passed-on TCP connection. Instead, proxy the child (and any
942 subsequent ones) for TLS via a unix-domain socket channel. Logging is
943 affected: the continued delivery log lines do not have any DNSSEC, TLS
944 Certificate or OCSP information. TLS cipher information is still logged.
946 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
947 identical IP addresses on different listening ports. Will also affect
950 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
951 add noisy ifdef guards to special-case this sillyness.
952 Patch from Bernd Kuhls.
954 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
955 than 255 are no longer allowed.
957 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
958 Disable the session-cache too, which might reduce our load. Since we
959 currrectly use a new context for every connection, both as server and
960 client, there is no benefit for these.
961 GnuTLS appears to not support tickets server-side by default (we don't
962 call gnutls_session_ticket_enable_server()) but client side is enabled
963 by default on recent versions (3.1.3 +) unless the PFS priority string
966 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
967 <https://reproducible-builds.org/specs/source-date-epoch/>.
969 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
970 the check for any unsuccessful recipients did not notice the limit, and
971 erroneously found still-pending ones.
973 JH/08 Pipeline CHUNKING command and data together, on kernels that support
974 MSG_MORE. Only in-clear (not on TLS connections).
976 JH/09 Avoid using a temporary file during transport using dkim. Unless a
977 transport-filter is involved we can buffer the headers in memory for
978 creating the signature, and read the spool data file once for the
979 signature and again for transmission.
981 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
982 4.77 as the kernel support then wasn't solid, having issues in 64bit
983 mode. Now, it's been long enough. Add support for FreeBSD also.
985 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
986 case where the routing stage had gathered several addresses to send to
987 a host before calling the transport for the first, we previously failed
988 to close down TLS in the old transport process before passing the TCP
989 connection to the new process. The new one sent a STARTTLS command
990 which naturally failed, giving a failed delivery and bloating the retry
991 database. Investigation and fix prototype from Wolfgang Breyha.
993 JH/12 Fix check on SMTP command input synchronisation. Previously there were
994 false-negatives in the check that the sender had not preempted a response
995 or prompt from Exim (running as a server), due to that code's lack of
996 awareness of the SMTP input buffering.
998 PP/04 Add commandline_checks_require_admin option.
999 Exim drops privileges sanely, various checks such as -be aren't a
1000 security problem, as long as you trust local users with access to their
1001 own account. When invoked by services which pass untrusted data to
1002 Exim, this might be an issue. Set this option in main configuration
1003 AND make fixes to the calling application, such as using `--` to stop
1006 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1007 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1008 responses to those, into a single TLS record each way (this usually means
1009 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1012 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1013 affects you only if you're dancing at the edge of the param size limits.
1014 If you are, and this message makes sense to you, then: raise the
1015 configured limit or use OpenSSL 1.1. Nothing we can do for older
1018 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1019 cmdline element to get the documented default one. Previously it was
1022 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1023 are unsuable for non-message-specific reasons - eg. connection timeout,
1024 banner-time rejection.
1026 JH/16 Fix logging of delivery remote port, when specified by router, under
1029 PP/06 Repair manualroute's ability to take options in any order, even if one
1030 is the name of a transport.
1033 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1035 JH/17 Change the list-building routines interface to use the expanding-string
1036 triplet model, for better allocation and copying behaviour.
1038 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1039 Previously it was constructed the first time a possibly-matching string
1040 was met in the configuration file input during startup; now it is done
1043 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1044 compatible one, to avoid the (poorly documented) possibility of a config
1045 file in the working directory redirecting the DB files, possibly correpting
1046 some existing file. CVE-2017-10140 assigned for BDB.
1048 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1049 cache-hot. Previously, although the result was properly cached, the
1050 initial verify call returned a defer.
1052 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1053 the main verify for receipient in uncached-mode.
1055 JH/22 Retire historical build files to an "unsupported" subdir. These are
1056 defined as "ones for which we have no current evidence of testing".
1058 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1059 if present. Previously it was ignored.
1061 JH/24 Start using specified-initialisers in C structure init coding. This is
1062 a C99 feature (it's 2017, so now considered safe).
1064 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1065 if was a fixed-sized field and bitmask ops via macros; it is now more
1068 PP/07 GitHub PR 56: Apply MariaDB build fix.
1069 Patch provided by Jaroslav Škarvada.
1071 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1072 during Coverity cleanups [4.87 JH/47]
1073 Diagnosis and fix provided by Michael Fischer v. Mollard.
1075 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1076 the right size to place the terminating semicolon on its own folded
1077 line, the header hash was calculated to an incorrect value thanks to
1078 the (relaxed) space the fold became.
1080 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1081 and confused the parent.
1083 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1084 which could crash as a result. This could lead to undeliverable messages.
1086 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1089 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1090 matters on fast-turnover and PID-randomising systems, which were getting
1091 out-of-order delivery.
1093 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1094 a possibly-overlapping copy. The symptom was that "Remote host closed
1095 connection in response to HELO" was logged instead of the actual 4xx
1098 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1099 Previously only that bufferd was discarded, resulting in SYMTP command
1102 JH/32 DKIM: when a message has multiple signatures matching an identity given
1103 in dkim_verify_signers, run the dkim acl once for each. Previously only
1104 one run was done. Bug 2189.
1106 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1107 "panic the current process" to "deliberately defer". The panic log is
1108 still written with the problem list name; the mail and reject logs now
1109 get a temp-reject line for the message that was being handled, saying
1110 something like "domains check lookup or other defer". The SMTP 451
1111 message is still "Temporary local problem".
1113 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1114 A crafted sequence of BDAT commands could result in in-use memory beeing
1115 freed. CVE-2017-16943.
1117 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1118 from SMTP input. Previously it was always done; now only done for DATA
1119 and not BDAT commands. CVE-2017-16944.
1121 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1122 to the message (such as an overlong header line). Previously this was
1123 not done and we did not exit BDAT mode. Followon from the previous item
1124 though a different problem.
1130 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1131 than -2003 did; needs libidn2 in addition to libidn.
1133 JH/02 The path option on a pipe transport is now expanded before use.
1135 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1136 Patch provided by "Björn", documentation fix added too.
1138 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1139 missing a wire-to-host endian conversion.
1141 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1142 close after a BDAT command line could be taken as a following command,
1143 giving a synch failure. Fix by only checking for synch immediately
1144 before acknowledging the chunk.
1146 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1147 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1148 macro. Patches provided by Josh Soref.
1150 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1151 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1152 by IANA as a service extension. However, John Klensin suggests that we
1155 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1156 the dkim code may be unix-mode line endings rather than smtp wire-format
1157 CRLF, so prepend a CR to any bare LF.
1159 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1160 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1162 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1163 the first were themselves being wrongly included in the feed into dkim
1164 processing; with most chunk sizes in use this resulted in an incorrect
1165 body hash calculated value.
1167 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1168 DKIM signature block, for verification. Although advised against by
1169 standards it is specifically not ruled illegal.
1171 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1173 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1174 missing a body hash (the bh= tag).
1176 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1177 It seems that HAProxy sends the Proxy Protocol information in clear and
1178 only then does a TLS startup, so do the same.
1180 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1181 TCP connections (such as for Spamd) unless the daemon successfully set
1182 Fast Open mode on its listening sockets. This fixes breakage seen on
1183 too-old kernels or those not configured for Fast Open, at the cost of
1184 requiring both directions being enabled for TFO, and TFO never being used
1185 by non-daemon-related Exim processes.
1187 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1188 endings, at least on the first header line. Try to canonify any that get
1189 past that check, despite the cost.
1191 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1192 now limited to an arbitrary five deep, while parsing addresses with the
1193 strip_excess_angle_brackets option enabled.
1195 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1196 instead leave the unprompted TLS handshake in socket buffer for the
1197 TLS library to consume.
1199 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1201 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1203 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1204 region in one operation, for speed, and this leaves assigned pointers
1205 dangling. Add checks run only under the testsuite which checks all
1206 variables at a store-reset and panics on a dangling pointer; add code
1207 explicitly nulling out all the variables discovered. Fixes one known
1208 bug: a transport crash, where a dangling pointer for $sending_ip_address
1209 originally assigned in a verify callout, is re-used.
1211 PP/06 Drop '.' from @INC in various Perl scripts.
1213 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1215 PP/08 Reduce a number of compilation warnings under clang; building with
1216 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1217 should be warning-free.
1219 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1221 HS/01 Fix portability problems introduced by PP/08 for platforms where
1222 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1224 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1225 chunk. This allows us to accept broken chunked messages. We need a more
1226 general solution here.
1228 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1229 already-broken messages in the queue.
1231 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1233 JH/19 Fix reference counting bug in routing-generated-address tracking.
1239 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1240 supports it and a size is available (ie. the sending peer gave us one).
1242 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1243 years of being deprecated). The replacements are the ACLs
1244 acl_smtp_mime and acl_not_smtp_mime.
1246 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1247 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1248 or even an in-clear connection were permitted. Now, if the host lookup
1249 was dnssec and dane was requested then the host is only used if the
1250 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1251 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1252 if one fails this test.
1253 This means that a poorly-configured remote DNS will make it incommunicado;
1254 but it protects against a DNS-interception attack on it.
1256 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1257 non-noisy when a race steals the message being considered.
1259 JH/05 If main configuration option tls_certificate is unset, generate a
1260 self-signed certificate for inbound TLS connections.
1262 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1263 in rewrites and routers.
1265 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1266 and logged a warning sing 4.83; now they are a configuration file error.
1268 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1269 (lacking @domain). Apply the same qualification processing as RCPT.
1271 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1273 JH/10 Support ${sha256:} applied to a string (as well as the previous
1276 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1277 a cutthrough deliver is pending, as we always want to make a connection.
1278 This also avoids re-routing the message when later placing the cutthrough
1279 connection after a verify cache hit.
1280 Do not update it with the verify result either.
1282 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1283 when routing results in more than one destination address.
1285 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1286 signing (which inhibits the cutthrough capability). Previously only
1287 the presence of an option was tested; now an expansion evaluating as
1288 empty is permissible (obviously it should depend only on data available
1289 when the cutthrough connection is made).
1291 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1292 the relevant preceding SMTP command did not note the pipelining mode.
1294 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1295 Previously they were not counted.
1297 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1298 as one having no matching records. Previously we deferred the message
1299 that needed the lookup.
1301 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1302 distinguished as "(=".
1304 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1305 for missing MX records. Previously it only worked for missing A records.
1307 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1309 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1310 after the data-go-ahead and data-ack. Patch from Jason Betts.
1312 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1313 even for a "none" policy. Patch from Tony Meyer.
1315 JH/22 Fix continued use of a connection for further deliveries. If a port was
1316 specified by a router, it must also match for the delivery to be
1319 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1320 When one of the recipients of a message was unsuitable for the connection
1321 (has no matching addresses), we lost track of needing to mark it
1322 deferred. As a result mail would be lost.
1324 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1326 JH/25 Decoding ACL controls is now done using a binary search; the source code
1327 takes up less space and should be simpler to maintain. Merge the ACL
1328 condition decode tables also, with similar effect.
1330 JH/26 Fix problem with one_time used on a redirect router which returned the
1331 parent address unchanged. A retry would see the parent address marked as
1332 delivered, so not attempt the (identical) child. As a result mail would
1335 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1336 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1337 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1340 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1341 Main config options check_{log,spool}_{inodes,space} are now
1342 100 inodes, 10MB unless set otherwise in the configuration.
1344 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1345 Previously it only applied to the main-section connection policy
1348 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1350 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1351 by me. Added RFC7919 DH primes as an alternative.
1353 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1354 are not in the system include path.
1356 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1357 GnuTLS, when a session startup failed (eg because the client disconnected)
1358 Exim did stdio operations after fclose. This was exposed by a recent
1359 change which nulled out the file handle after the fclose.
1361 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1362 signed directly by the cert-signing cert, rather than an intermediate
1363 OCSP-signing cert. This is the model used by LetsEncrypt.
1365 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1367 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1368 an incoming connection.
1370 HS/02 Bug 1802: Do not half-close the connection after sending a request
1373 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1374 fallback to "prime256v1".
1376 JH/34 SECURITY: Use proper copy of DATA command in error message.
1377 Could leak key material. Remotely exploitable. CVE-2016-9963.
1383 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1384 and 3.4.4 - once the server is enabled to respond to an OCSP request
1385 it does even when not requested, resulting in a stapling non-aware
1386 client dropping the TLS connection.
1388 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1389 support variable-length bit vectors. No functional change.
1391 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1392 The I= interface field on outgoing lines is now after the H= remote
1393 host field, same as incoming lines. There is a separate
1394 outgoing_interface log selector which allows you to disable the
1397 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1398 If not running log_selector +smtp_connection the mainlog would be held
1399 open indefinitely after a "too many connections" event, including to a
1400 deleted file after a log rotate. Leave the per net connection logging
1401 leaving it open for efficiency as that will be quickly detected by the
1402 check on the next write.
1404 HS/01 Bug 1671: Fix post transport crash.
1405 Processing the wait-<transport> messages could crash the delivery
1406 process if the message IDs didn't exist for some reason. When
1407 using 'split_spool_directory=yes' the construction of the spool
1408 file name failed already, exposing the same netto behaviour.
1410 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1411 mime_regex ACL conditions.
1413 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1414 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1415 to HELO, local diagnostic string.
1417 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1418 log line to debug. Even when configured with a tls authenticator many
1419 client connections are expected to not authenticate in this way, so
1420 an authenticate fail is not an error.
1422 HS/02 Add the Exim version string to the process info. This way exiwhat
1423 gives some more detail about the running daemon.
1425 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1426 matter for fast-change records such as DNSBLs.
1428 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1429 retry record, even if constant. There may be multiple transports with
1430 different interface settings and the retry behaviour needs to be kept
1433 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1435 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1437 JH/10 Bug 840: fix log_defer_output option of pipe transport
1439 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1440 hosts_randomize. This matters a lot when combined with mua_wrapper.
1442 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1443 ${quote_pgsql:<string>} operator.
1445 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1447 JH/14 Change header file name for freeradius-client. Relevant if compiling
1448 with Radius support; from the Gentoo tree and checked under Fedora.
1450 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1452 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1453 option transport_filter, meaning no filtering.
1455 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1457 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1458 defaults to "*" (all hosts). The variable is now available when not built
1459 with TLS, default unset, mainly to enable keeping the testsuite sane.
1460 If a server certificate is not supplied (via tls_certificate) an error is
1461 logged, and clients will find TLS connections fail on startup. Presumably
1462 they will retry in-clear.
1463 Packagers of Exim are strongly encouraged to create a server certificate
1464 at installation time.
1466 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1467 with the $config_file variable.
1469 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1470 in transport context, after the attempt, and per-recipient. The latter type
1471 is per host attempted. The event data is the error message, and the errno
1472 information encodes the lookup type (A vs. MX) used for the (first) host,
1473 and the trailing two digits of the smtp 4xx response.
1475 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1476 to write to mainlog (or rejectlog, paniclog) in the window between file
1477 creation and permissions/ownership being changed. Particularly affects
1478 installations where exicyclog is run as root, rather than exim user;
1479 result is that the running daemon panics and dies.
1481 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1483 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1484 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1485 "pri" and "weight". Note that the previous implicit priority given by the
1486 list order is no longer honoured.
1488 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1489 for DKIM processing.
1491 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1492 by defining SUPPORT_SOCKS.
1494 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1495 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1496 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1497 variables are renamed to proxy_{local,external}_{address,port}.
1499 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1500 for a build by defining SUPPORT_I18N
1502 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1503 of the query string, and make ${quote_redis:} do that quoting.
1505 JH/29 Move Events support from Experimental to mainline, enabled by default
1506 and removable for a build by defining DISABLE_EVENT.
1508 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1510 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1511 cached by the daemon.
1513 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1514 by defining LOOKUP_REDIS. The libhiredis library is required.
1516 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1517 keys are given for lookup.
1519 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1520 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1521 only supported when built with TLS support. The PolarSSL SHA routines
1522 are still used when the TLS library is too old for convenient support.
1524 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1525 openssl_options), for security. OpenSSL forces this from version 1.1.0
1526 server-side so match that on older versions.
1528 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1529 allocation for $value could be released as the expansion processing
1530 concluded, but leaving the global pointer active for it.
1532 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1533 and to use the domains and local_parts ACL conditions.
1535 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1536 incorrectly not doubled on cutthrough transmission, hence seen as a
1537 body-termination at the receiving system - resulting in truncated mails.
1538 Commonly the sender saw a TCP-level error, and retransmitted the message
1539 via the normal store-and-forward channel. This could result in duplicates
1540 received - but deduplicating mailstores were liable to retain only the
1541 initial truncated version.
1543 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1545 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1547 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1548 we're in there, support oversigning also; bug 1309.
1550 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1552 HS/04 Add support for keep_environment and add_environment options.
1554 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1555 either intentional arithmetic overflow during PRNG, or testing config-
1558 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1559 delivery resulted in actual delivery. Cancel cutthrough before DATA
1562 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1563 rejects a recipient: pass the reject to the originator.
1565 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1566 Many were false-positives and ignorable, but it's worth fixing the
1569 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1570 for the new environment-manipulation done at startup. Move the routines
1571 from being local to tls.c to being global via the os.c file.
1573 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1574 an extract embedded as result-arg for a map, the first arg for extract
1575 is unavailable so we cannot tell if this is a numbered or keyed
1576 extraction. Accept either.
1582 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1585 JH/02 The smtp transport option "multi_domain" is now expanded.
1587 JH/03 The smtp transport now requests PRDR by default, if the server offers
1590 JH/04 Certificate name checking on server certificates, when exim is a client,
1591 is now done by default. The transport option tls_verify_cert_hostnames
1592 can be used to disable this per-host. The build option
1593 EXPERIMENTAL_CERTNAMES is withdrawn.
1595 JH/05 The value of the tls_verify_certificates smtp transport and main options
1596 default to the word "system" to access the system default CA bundle.
1597 For GnuTLS, only version 3.0.20 or later.
1599 JH/06 Verification of the server certificate for a TLS connection is now tried
1600 (but not required) by default. The verification status is now logged by
1601 default, for both outbound TLS and client-certificate supplying inbound
1604 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1607 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1608 Status Notification (bounce) messages are now MIME format per RFC 3464.
1609 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1610 under the control of the dsn_advertise_hosts option, and routers may
1611 have a dsn_lasthop option.
1613 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1614 default, modifiable by a malware= option. The list separator for
1615 the options can now be changed in the usual way. Bug 68.
1617 JH/10 The smtp_receive_timeout main option is now expanded before use.
1619 JH/11 The incoming_interface log option now also enables logging of the
1620 local interface on delivery outgoing connections.
1622 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1623 if the interface and destination host and port all match.
1625 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1628 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1629 Patch from Andrew Lewis.
1631 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1632 now supports optional time-restrictions, weighting, and priority
1633 modifiers per server. Patch originally by <rommer@active.by>.
1635 JH/16 The spamd_address main option now supports a mixed list of local
1636 and remote servers. Remote servers can be IPv6 addresses, and
1637 specify a port-range.
1639 JH/17 Bug 68: The spamd_address main option now supports an optional
1640 timeout value per server.
1642 JH/18 Bug 1581: Router and transport options headers_add/remove can
1643 now have the list separator specified.
1645 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1648 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1651 JH/21 Support for the A6 type of dns record is withdrawn.
1653 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1654 rather than the verbs used.
1656 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1657 from 255 to 1024 chars.
1659 JH/24 Verification callouts now attempt to use TLS by default.
1661 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1662 are generic router options now. The defaults didn't change.
1664 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1665 Original patch from Alexander Shikoff, worked over by JH.
1667 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1668 files if ZCAT_COMMAND is not executable.
1670 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1672 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1674 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1675 Normally benign, it bites when the pair was led to by a CNAME;
1676 modern usage is to not canonicalize the domain to a CNAME target
1677 (and we were inconsistent anyway for A-only vs AAAA+A).
1679 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1681 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1682 when evaluating $sender_host_dnssec.
1684 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1685 $sender_helo_dnssec variable.
1687 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1689 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1691 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1693 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1694 documented as working, but never had. Support all but $spam_report.
1696 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1697 added for tls authenticator.
1699 HS/03 Add perl_taintmode main config option
1705 TL/01 When running the test suite, the README says that variables such as
1706 no_msglog_check are global and can be placed anywhere in a specific
1707 test's script, however it was observed that placement needed to be near
1708 the beginning for it to behave that way. Changed the runtest perl
1709 script to read through the entire script once to detect and set these
1710 variables, reset to the beginning of the script, and then run through
1711 the script parsing/test process like normal.
1713 TL/02 The BSD's have an arc4random API. One of the functions to induce
1714 adding randomness was arc4random_stir(), but it has been removed in
1715 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1716 function when detected.
1718 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1719 cause callback expansion.
1721 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1722 syntax errors in an expansion can be treated as a string instead of
1723 logging or causing an error, due to the internal use of bool_lax
1724 instead of bool when processing it.
1726 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1727 server certificates when making smtp deliveries.
1729 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1731 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1733 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1735 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1736 Merged patch from Sebastian Wiedenroth.
1738 JH/05 Fix results-pipe from transport process. Several recipients, combined
1739 with certificate use, exposed issues where response data items split
1740 over buffer boundaries were not parsed properly. This eventually
1741 resulted in duplicates being sent. This issue only became common enough
1742 to notice due to the introduction of connection certificate information,
1743 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1745 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1746 size buffer was used, resulting in syntax errors when an expansion
1749 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1750 version 3.3.6 or later.
1752 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1753 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1754 both become "event_action", the variables become $event_name, $event_data
1755 and $event_defer_errno. There is a new variable $verify_mode, usable in
1756 routers, transports and related events. The tls:cert event is now also
1757 raised for inbound connections, if the main configuration event_action
1760 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1761 early OCSP support, but no stapling (appears to be less than 1.0.0).
1763 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1764 server certificate names available under the smtp transport option
1765 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1768 JH/10 Time-related extraction expansions from certificates now use the main
1769 option "timezone" setting for output formatting, and are consistent
1770 between OpenSSL and GnuTLS compilations. Bug 1541.
1772 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1773 encoded parameter in the incoming message. Bug 1558.
1775 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1776 include certificate info, eximon was claiming there were spoolfile
1779 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1781 JH/14 Log delivery-related information more consistently, using the sequence
1782 "H=<name> [<ip>]" wherever possible.
1784 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1785 are problematic for Debian distribution, omit them from the release
1788 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1790 JH/16 Fix string representation of time values on 64bit time_t architectures.
1793 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1794 output list separator was used.
1799 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1800 checkers that were complaining about end of non-void function with no
1803 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1804 This was a regression introduced in 4.83 by another bugfix.
1806 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1808 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1809 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1815 TF/01 Correctly close the server side of TLS when forking for delivery.
1817 When a message was received over SMTP with TLS, Exim failed to clear up
1818 the incoming connection properly after forking off the child process to
1819 deliver the message. In some situations the subsequent outgoing
1820 delivery connection happened to have the same fd number as the incoming
1821 connection previously had. Exim would try to use TLS and fail, logging
1822 a "Bad file descriptor" error.
1824 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1825 utilities have not been installed.
1827 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1828 temporary space as the ACL may create new global variables.
1830 TL/01 LDAP support uses per connection or global context settings, depending
1831 upon the detected version of the libraries at build time.
1833 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1834 to extract and use the src ip:port in logging and expansions as if it
1835 were a direct connection from the outside internet. PPv2 support was
1836 updated based on HAProxy spec change in May 2014.
1838 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1840 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1841 Properly escape header and check for NULL return.
1843 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1846 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1848 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1849 characters in header names, implemented as a verify condition.
1850 Contributed by Michael Fischer v. Mollard.
1852 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1853 results permerror and temperror. Previous values are deprecated but
1854 still accepted. In a future release, err_perm and err_temp will be
1855 completely removed, which will be a backward incompatibility if the
1856 ACL tests for either of these two old results. Patch contributed by
1857 user bes-internal on the mailing list.
1859 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1861 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1862 selectors, in both main and reject logs.
1864 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1867 JH/07 Add malware type "sock" for talking to simple daemon.
1869 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1871 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1872 routers/transports under cutthrough routing.
1874 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1875 numbers. Touch up "bool" conditional to keep the same definition.
1877 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1879 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1881 JH/12 Expand items in router/transport headers_add or headers_remove lists
1882 individually rather than the list as a whole. Bug 1452.
1884 Required for reasonable handling of multiple headers_ options when
1885 they may be empty; requires that headers_remove items with embedded
1886 colons must have them doubled (or the list-separator changed).
1888 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1889 view the policy declared in the DMARC record. Currently, $dmarc_status
1890 is a combined value of both the record presence and the result of the
1893 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1895 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1896 dnslookup router and the smtp transport (applying to the forward
1899 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1900 of ldap servers used for a specific lookup. Patch provided by Heiko
1903 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1904 New variable $lookup_dnssec_authenticated for observability.
1906 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1907 Patch submitted by Lars Timman.
1909 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1911 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1912 Requires trusted mode and valid format message id, aborts otherwise.
1913 Patch contributed by Heiko Schlichting.
1915 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1916 certextract with support for various fields. Bug 1358.
1918 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1919 is requested by default, modifiable by smtp transport option
1922 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1923 operate on certificate variables to give certificate fingerprints
1924 Also new ${sha256:cert_variable}.
1926 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1928 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1929 Christian Aistleitner.
1931 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1933 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1934 file. Patch from Wolfgang Breyha.
1936 JH/25 Expand the coverage of the delivery $host and $host_address to
1937 client authenticators run in verify callout. Bug 1476.
1939 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1940 align with daemon_smtp_ports. Bug 72.
1942 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1943 support and error reporting did not work properly.
1945 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1946 and is readable. Patch from Andrew Colin Kissa.
1948 TL/14 Enhance documentation of ${run expansion and how it parses the
1949 commandline after expansion, particularly in the case when an
1950 unquoted variable expansion results in an empty value.
1952 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1954 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1955 by renaming away from it. Addresses GH issue 15, reported by
1958 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1960 TL/15 SECURITY: prevent double expansion in math comparison functions
1961 (can expand unsanitized data). Not remotely exploitable.
1968 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1970 PP/02 Make -n do something, by making it not do something.
1971 When combined with -bP, the name of an option is not output.
1973 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1976 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1977 $sender_host_name and config options to manage this, and basic check
1980 PP/05 DSCP support for outbound connections and control modifier for inbound.
1982 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1983 (Only plugin which currently uses this is kerberos4, which nobody should
1984 be using, but we should make it available and other future plugins might
1985 conceivably use it, even though it would break NAT; stuff *should* be
1986 using channel bindings instead).
1988 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1989 name; added for Sendmail compatibility; requires admin caller.
1990 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1991 just ignore it); requires trusted caller.
1992 Also parse but ignore: -Ac -Am -X<logfile>
1995 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1997 TL/02 Add +smtp_confirmation as a default logging option.
1999 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2000 Patch by Magnus Holmgren from 2007-02-20.
2002 TL/04 Bugzilla 1281 - Spec typo.
2003 Bugzilla 1283 - Spec typo.
2004 Bugzilla 1290 - Spec grammar fixes.
2006 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2008 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2010 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2011 mailing list by by Dmitry Isaikin.
2013 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2015 JH/02 Support "G" suffix to numbers in ${if comparisons.
2017 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2019 NM/01 Bugzilla 1197 - Spec typo
2020 Bugzilla 1196 - Spec examples corrections
2022 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2024 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2025 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2028 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2029 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2031 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2032 "acl {{name}{arg}...}", and optional args on acl condition
2035 JH/05 Permit multiple router/transport headers_add/remove lines.
2037 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2039 JH/07 Avoid using a waiting database for a single-message-only transport.
2040 Performance patch from Paul Fisher. Bugzilla 1262.
2042 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2045 JH/09 Add $headers_added variable, with content from use of ACL modifier
2046 add_header (but not yet added to the message). Bugzilla 199.
2048 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2049 Pulled from Bugzilla 817 by Wolfgang Breyha.
2051 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2053 (nb: this is the same fix as in Exim 4.80.1)
2055 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2058 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2061 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2063 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2064 advertise SMTP AUTH mechanism to us, instead of a generic
2065 protocol violation error. Also, make Exim more robust to bad
2066 data from the Dovecot auth socket.
2068 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2070 When a queue runner is handling a message, Exim first routes the
2071 recipient addresses, during which it prunes them based on the retry
2072 hints database. After that it attempts to deliver the message to
2073 any remaining recipients. It then updates the hints database using
2076 So if a recipient address works intermittently, it can get repeatedly
2077 deferred at routing time. The retry hints record remains fresh so the
2078 address never reaches the final cutoff time.
2080 This is a fairly common occurrence when a user is bumping up against
2081 their storage quota. Exim had some logic in its local delivery code
2082 to deal with this. However it did not apply to per-recipient defers
2083 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2085 This change adds a proper retry rule check during routing so that the
2086 final cutoff time is checked against the message's age. We only do
2087 this check if there is an address retry record and there is not a
2088 domain retry record; this implies that previous attempts to handle
2089 the address had the retry_use_local_parts option turned on. We use
2090 this as an approximation for the destination being like a local
2091 delivery, as in LMTP.
2093 I suspect this new check makes the old local delivery cutoff check
2094 redundant, but I have not verified this so I left the code in place.
2096 TF/02 Correct gecos expansion when From: is a prefix of the username.
2098 Test 0254 submits a message to Exim with the header
2102 When I ran the test suite under the user fanf2, Exim expanded
2103 the header to contain my full name, whereas it should have added
2104 a Resent-Sender: header. It erroneously treats any prefix of the
2105 username as equal to the username.
2107 This change corrects that bug.
2109 GF/01 DCC debug and logging tidyup
2110 Error conditions log to paniclog rather than rejectlog.
2111 Debug lines prefixed by "DCC: " to remove any ambiguity.
2113 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2115 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2116 Bug spotted by Jeremy Harris; was flawed since initial commit.
2117 Would have resulted in OCSP responses post-SNI triggering an Exim
2118 NULL dereference and crash.
2120 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2122 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2123 Bug detection, analysis and fix by Samuel Thibault.
2124 Bugzilla 1331, Debian bug #698092.
2126 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2128 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2129 Server implementation by Todd Lyons, client by JH.
2130 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2131 config variable "prdr_enable" controls whether the server
2132 advertises the facility. If the client requests PRDR a new
2133 acl_data_smtp_prdr ACL is called once for each recipient, after
2134 the body content is received and before the acl_smtp_data ACL.
2135 The client is controlled by both of: a hosts_try_prdr option
2136 on the smtp transport, and the server advertisement.
2137 Default client logging of deliveries and rejections involving
2138 PRDR are flagged with the string "PRDR".
2140 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2141 fclose(). Diagnosis by Todd Lyons.
2143 PP/17 Update configure.default to handle IPv6 localhost better.
2144 Patch by Alain Williams (plus minor tweaks).
2147 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2148 This is now consistent with GnuTLS, and is now documented: the
2149 previous undocumented portable approach to treating the option as
2150 unset was to force an expansion failure. That still works, and
2151 an empty string is now equivalent.
2153 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2154 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2155 not performing validation itself.
2157 PP/20 Added force_command boolean option to pipe transport.
2158 Patch from Nick Koston, of cPanel Inc.
2160 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2163 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2165 PP/21 Fix eximon continuous updating with timestamped log-files.
2166 Broken in a format-string cleanup in 4.80, missed when I repaired the
2167 other false fix of the same issue.
2168 Report and fix from Heiko Schlichting.
2171 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2172 Report from Prashanth Katuri.
2174 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2175 It's SecureTransport, so affects any MacOS clients which use the
2176 system-integrated TLS libraries, including email clients.
2178 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2179 using a MIME ACL for non-SMTP local injection.
2180 Report and assistance in diagnosis by Warren Baker.
2182 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2184 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2186 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2187 last id that failed so it may be referenced in subsequent ACL's.
2189 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2192 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2193 ldap library initialization, allowing self-signed CA's to be
2194 used. Also properly sets require_cert option later in code by
2195 using NULL (global ldap config) instead of ldap handle (per
2196 session). Bug diagnosis and testing by alxgomz.
2198 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2199 the src/util/ subdirectory.
2201 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2202 renamed to Transport Post Delivery Action by Jeremy Harris, as
2205 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2206 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2207 redis_servers = needs to be configured which will be used by the redis
2208 lookup. Patch from Warren Baker, of The Packet Hub.
2210 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2212 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2213 hostname or reverse DNS when processing a host list. Used suggestions
2214 from multiple comments on this bug.
2216 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2218 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2219 Missed a few lines, added it to make the runtest require no keyboard
2222 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2223 contains upper case chars. Make router use caseful_local_part.
2225 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2226 support when GnuTLS has been built with p11-kit.
2232 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2234 This, or similar/improved, will also be change PP/11 of 4.82.
2240 PP/01 Handle short writes when writing local log-files.
2241 In practice, only affects FreeBSD (8 onwards).
2242 Bugzilla 1053, with thanks to Dmitry Isaikin.
2244 NM/01 Bugzilla 949 - Documentation tweak
2246 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2249 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2251 PP/02 Implemented gsasl authenticator.
2253 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2255 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2256 `pkg-config foo` for cflags/libs.
2258 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2259 with rest of GSASL and with heimdal_gssapi.
2261 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2262 `pkg-config foo` for cflags/libs for the TLS implementation.
2264 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2265 properties get this fed in as external SSF. A number of robustness
2266 and debugging improvements to the cyrus_sasl authenticator.
2268 PP/08 cyrus_sasl server now expands the server_realm option.
2270 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2271 Patch by Jeremy Harris.
2273 PP/10 Added dbmjz lookup type.
2275 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2277 PP/12 MAIL args handles TAB as well as SP, for better interop with
2278 non-compliant senders.
2279 Analysis and variant patch by Todd Lyons.
2281 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2282 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2283 Patch from Dirk Mueller <dmueller@suse.com>
2285 PP/13 tls_peerdn now print-escaped for spool files.
2286 Observed some $tls_peerdn in wild which contained \n, which resulted
2287 in spool file corruption.
2289 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2290 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2291 or write after TLS renegotiation, which otherwise led to messages
2294 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2295 as a tracking header (ie: a signed header comes before the signature).
2296 Patch from Wolfgang Breyha.
2298 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2299 comma-sep list; embedded commas doubled.
2301 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2303 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2305 Report and patch from Dmitry Banschikov.
2307 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2308 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2309 failures. We appear to now support TLS1.1+ with Exim.
2311 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2312 lets Exim select keys and certificates based upon TLS SNI from client.
2313 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2314 before an outbound SMTP session. New log_selector, +tls_sni.
2316 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2317 NULL dereference. Report and patch from Alun Jones.
2319 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2320 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2321 Not seeing resolver debug output on NetBSD, but suspect this is a
2322 resolver implementation change.
2324 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2325 Left warnings. Added "eximon gdb" invocation mode.
2327 PP/21 Defaulting "accept_8bitmime" to true, not false.
2329 PP/22 Added -bw for inetd wait mode support.
2331 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2332 locate the relevant includes and libraries. Made this the default.
2334 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2335 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2337 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2338 This may cause build issues on older platforms.
2340 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2341 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2342 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2343 Added SNI support via GnuTLS too.
2344 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2346 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2348 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2349 Applied second patch from Janne, implementing suggestion to default
2350 multiple-strings-in-record handling to match SPF spec.
2352 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2354 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2355 read-only, out of scope).
2356 Patch from Wolfgang Breyha, report from Stuart Northfield.
2358 PP/29 Fix three issues highlighted by clang analyser static analysis.
2359 Only crash-plausible issue would require the Cambridge-specific
2360 iplookup router and a misconfiguration.
2361 Report from Marcin Mirosław.
2363 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2365 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2366 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2367 As part of this, removing so much warning spew let me fix some minor
2368 real issues in debug logging.
2370 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2371 assignment on my part. Fixed.
2373 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2374 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2375 Janne Snabb (who went above and beyond: thank you).
2377 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2378 string otherwise requires a connection and a bunch more work and it's
2379 relatively easy to get wrong. Should also expose TLS library linkage
2382 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2383 64-bit ${eval} (JH/03).
2385 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2386 GNU libc to support some of the 64-bit stuff, should not lead to
2387 conflicts. Defined before os.h is pulled in, so if a given platform
2388 needs to override this, it can.
2390 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2391 protection layer was required, which is not implemented.
2392 Bugzilla 1254, patch from Wolfgang Breyha.
2394 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2395 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2396 tls_dhparam take prime identifiers. Also unbreak combination of
2397 OpenSSL+DH_params+TLSSNI.
2399 PP/39 Disable SSLv2 by default in OpenSSL support.
2405 PP/01 Solaris build fix for Oracle's LDAP libraries.
2406 Bugzilla 1109, patch from Stephen Usher.
2408 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2410 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2413 TF/02 Fix a couple more cases where we did not log the error message
2414 when unlink() failed. See also change 4.74-TF/03.
2416 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2417 lock up or crash if it happened to be inside a call to libc when it
2418 got a SIGUSR1 from exiwhat.
2420 The SIGUSR1 handler appends the current process status to the process
2421 log which is later printed by exiwhat. It used to use the general
2422 purpose logging code to do this, but several functions it calls are
2423 not safe for signals.
2425 The new output code in the SIGUSR1 handler is specific to the process
2426 log, and simple enough that it's easy to inspect for signal safety.
2427 Removing some special cases also simplifies the general logging code.
2428 Removing the spurious timestamps from the process log simplifies
2431 TF/04 Improved ratelimit ACL condition.
2433 The /noupdate option has been deprecated in favour of /readonly which
2434 has clearer semantics. The /leaky, /strict, and /readonly update modes
2435 are mutually exclusive. The update mode is no longer included in the
2436 database key; it just determines when the database is updated. (This
2437 means that when you upgrade Exim will forget old rate measurements.)
2439 Exim now checks that the per_* options are used with an update mode that
2440 makes sense for the current ACL. For example, when Exim is processing a
2441 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2442 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2443 must specify per_mail/readonly. If you omit the update mode it defaults to
2444 /leaky where that makes sense (as before) or /readonly where required.
2446 The /noupdate option is now undocumented but still supported for
2447 backwards compatibility. It is equivalent to /readonly except that in
2448 ACLs where /readonly is required you may specify /leaky/noupdate or
2449 /strict/noupdate which are treated the same as /readonly.
2451 A useful new feature is the /count= option. This is a generalization
2452 of the per_byte option, so that you can measure the throughput of other
2453 aggregate values. For example, the per_byte option is now equivalent
2454 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2456 The per_rcpt option has been generalized using the /count= mechanism
2457 (though it's more complicated than the per_byte equivalence). When it is
2458 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2459 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2460 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2461 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2462 Note that using per_rcpt with a non-readonly update mode in more than
2463 one ACL will cause the recipients to be double-counted. (The per_mail
2464 and per_byte options don't have this problem.)
2466 The handling of very low rates has changed slightly. If the computed rate
2467 is less than the event's count (usually one) then this event is the first
2468 after a long gap. In this case the rate is set to the same as this event's
2469 count, so that the first message of a spam run is counted properly.
2471 The major new feature is a mechanism for counting the rate of unique
2472 events. The new per_addr option counts the number of different
2473 recipients that someone has sent messages to in the last time period. It
2474 behaves like per_rcpt if all the recipient addresses are different, but
2475 duplicate recipient addresses do not increase the measured rate. Like
2476 the /count= option this is a general mechanism, so the per_addr option
2477 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2478 example, measure the rate that a client uses different sender addresses
2479 with the options per_mail/unique=$sender_address. There are further
2480 details in the main documentation.
2482 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2484 TF/06 Removed a few PCRE remnants.
2486 TF/07 Automatically extract Exim's version number from tags in the git
2487 repository when doing development or release builds.
2489 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2490 Bugzilla 879. Patch from Paul Fisher.
2492 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2493 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2496 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2498 PP/05 Variable $av_failed, true if the AV scanner deferred.
2499 Bugzilla 1078. Patch from John Horne.
2501 PP/06 Stop make process more reliably on build failure.
2502 Bugzilla 1087. Patch from Heiko Schlittermann.
2504 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2505 Bugzilla 1089. Patch from Heiko Schlittermann.
2507 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2508 Bugzilla 1131. Patch from Holger Weiß.
2510 PP/09 Handle IPv6 addresses with SPF.
2511 Bugzilla 860. Patch from Wolfgang Breyha.
2513 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2515 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2518 PP/11 match_* no longer expand right-hand-side by default.
2519 New compile-time build option, EXPAND_LISTMATCH_RHS.
2520 New expansion conditions, "inlist", "inlisti".
2522 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2524 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2526 PP/14 fix log_write() format string regression from TF/03.
2527 Bugzilla 1152. Patch from Dmitry Isaikin.
2533 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2535 PP/02 Harmonised TLS library version reporting; only show if debugging.
2536 Layout now matches that introduced for other libraries in 4.74 PP/03.
2538 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2540 PP/04 New "dns_use_edns0" global option.
2542 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2545 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2546 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2548 TK/01 Updated PolarSSL code to 0.14.2.
2549 Bugzilla 1097. Patch from Andreas Metzler.
2551 PP/07 Catch divide-by-zero in ${eval:...}.
2552 Fixes bugzilla 1102.
2554 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2557 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2558 format-string attack -- SECURITY: remote arbitrary code execution.
2560 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2561 time unintentionally subject to list matching rules, letting the header
2562 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2563 arbitrary string expansion). This allowed for information disclosure.
2565 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2566 INT_MIN/-1 -- value coerced to INT_MAX.
2572 NM/01 Workaround for PCRE version dependency in version reporting
2575 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2576 This fixes portability to compilers other than gcc, notably
2577 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2579 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2580 makefiles for portability to HP-UX and POSIX correctness.
2582 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2583 Also via indented variable definition in the Makefile.
2584 (Debugging by Oliver Heesakkers).
2586 PP/02 Restore caching of spamd results with expanded spamd_address.
2587 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2589 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2590 Improves build reliability. Fix from: Frank Elsner
2592 NM/02 Fix wide character breakage in the rfc2047 coding
2593 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2595 NM/03 Allow underscore in dnslist lookups
2596 Fixes bug 1026. Patch from Graeme Fowler
2598 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2599 Code patches from Adam Ciarcinski of NetBSD.
2601 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2604 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2605 is logged, to avoid truncation. Patch from John Horne.
2607 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2608 Patch from Jakob Hirsch.
2610 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2611 SQL string expansion failure details.
2612 Patch from Andrey Oktyabrski.
2614 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2615 Patch from Simon Arlott.
2617 PP/09 New lookups functionality failed to compile on old gcc which rejects
2618 extern declarations in function scope.
2619 Patch from Oliver Fleischmann
2621 PP/10 Use sig_atomic_t for flags set from signal handlers.
2622 Check getgroups() return and improve debugging.
2623 Fixed developed for diagnosis in bug 927 (which turned out to be
2626 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2627 Patch from Mark Zealey.
2629 PP/12 Bugzilla 1056: Improved spamd server selection.
2630 Patch from Mark Zealey.
2632 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2633 Based on patch from Heiko Schlittermann.
2635 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2636 Patch from Uwe Doering, sign-off by Michael Haardt.
2638 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2639 variable declaration deep within a block. Bug and patch from
2642 PP/15 lookups-Makefile IRIX compatibility coercion.
2644 PP/16 Make DISABLE_DKIM build knob functional.
2646 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2647 Patch by Simon Arlott
2649 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2650 variable argument macros. Our copy now differs from upstream.
2656 TF/01 Failure to get a lock on a hints database can have serious
2657 consequences so log it to the panic log.
2659 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2660 controlled using the smtp_confirmation log selector.
2662 TF/03 Include the error message when we fail to unlink a spool file.
2664 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2665 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2666 for maintaining out-of-tree patches for some time.
2668 PP/01 Bugzilla 139: Documentation and portability issues.
2669 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2670 Handle per-OS dynamic-module compilation flags.
2672 PP/02 Let /dev/null have normal permissions.
2673 The 4.73 fixes were a little too stringent and complained about the
2674 permissions on /dev/null. Exempt it from some checks.
2675 Reported by Andreas M. Kirchwitz.
2677 PP/03 Report version information for many libraries, including
2678 Exim version information for dynamically loaded libraries. Created
2679 version.h, now support a version extension string for distributors
2680 who patch heavily. Dynamic module ABI change.
2682 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2683 privilege escalation vulnerability whereby the Exim run-time user
2684 can cause root to append content of the attacker's choosing to
2687 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2690 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2691 If dropping privileges for untrusted macros, we disabled normal logging
2692 on the basis that it would fail; for the Exim run-time user, this is not
2693 the case, and it resulted in successful deliveries going unlogged.
2694 Fixed. Reported by Andreas Metzler.
2700 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2701 only prepend for the Resent-* case. Fixes regression introduced in
2702 Exim 4.70 by NM/22 for Bugzilla 607.
2704 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2705 increasing numbers of administrators be bitten by this.
2707 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2709 PP/03 Bugzilla 994: added openssl_options main configuration option.
2711 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2713 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2715 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2717 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2718 without a peer certificate, leading to a segfault because of an
2719 assumption that peers always have certificates. Be a little more
2720 paranoid. Problem reported by Martin Tscholak.
2722 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2723 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2724 NB: ClamAV planning to remove STREAM in "middle of 2010".
2725 CL also introduces -bmalware, various -d+acl logging additions and
2726 more caution in buffer sizes.
2728 PP/09 Implemented reverse_ip expansion operator.
2730 PP/10 Bugzilla 937: provide a "debug" ACL control.
2732 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2734 PP/12 Bugzilla 973: Implement --version.
2736 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2738 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2740 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2742 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2743 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2744 ignore trailing whitespace.
2746 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2748 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2751 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2752 Notification from Dr Andrew Aitchison.
2754 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2755 ExtendedDetectionInfo response format.
2756 Notification from John Horne.
2758 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2761 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2762 XSL and documented dependency on system catalogs, with examples of how
2765 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2768 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2769 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2770 configuration file which is writeable by the Exim user or group.
2772 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2773 of configuration files to cover files specified with the -C option if
2774 they are going to be used with root privileges, not just the default
2777 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2778 option (effectively making it always true).
2780 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2781 files to be used while preserving root privileges.
2783 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2784 that rogue child processes cannot use them.
2786 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2787 run-time user, instead of root.
2789 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2790 Exim run-time user without dropping privileges.
2792 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2793 result string, instead of calling string_vformat() twice with the same
2796 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2797 for other users. Others should always drop root privileges if they use
2798 -C on the command line, even for a whitelisted configure file.
2800 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2802 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2808 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2809 $data_path, and $header_path variables; fixed documentation bugs and
2812 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2813 exipick to access non-standard spools, including the "frozen" queue
2816 NM/01 Bugzilla 965: Support mysql stored procedures.
2817 Patch from Alain Williams
2819 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2821 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2822 Patch from Andreas Metzler
2824 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2825 Patch from Kirill Miazine
2827 NM/05 Bugzilla 671: Added umask to procmail example.
2829 JJ/03 installed exipick 20100323.0, fixing doc bug
2831 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2832 directory. Notification and patch from Dan Rosenberg.
2834 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2836 TK/02 Improve log output when DKIM signing operation fails.
2838 MH/01 Treat the transport option dkim_domain as a colon separated
2839 list, not as a single string, and sign the message with each element,
2840 omitting multiple occurences of the same signer.
2842 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2843 Bugzilla 985, 986. Patch by Simon Arlott
2845 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2846 Patch by Simon Arlott
2848 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2849 MBX locking. Notification from Dan Rosenberg.
2855 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2857 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2859 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2861 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2863 NM/04 Bugzilla 915: Flag broken perl installation during build.
2869 TK/01 Added patch by Johannes Berg that expands the main option
2870 "spamd_address" if it starts with a dollar sign.
2872 TK/02 Write list of recipients to X-Envelope-Sender header when building
2873 the mbox-format spool file for content scanning (suggested by Jakob
2876 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2877 (http://www.dcc-servers.net/) support via dccifd. Activated by
2878 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2880 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2881 by Mark Daniel Reidel <mr@df.eu>.
2883 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2884 When building exim an external PCRE library is now needed -
2885 PCRE is a system library on the majority of modern systems.
2886 See entry on PCRE_LIBS in EDITME file.
2888 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2889 conversation. Added nologin parameter to request.
2890 Patch contributed by Kirill Miazine.
2892 TF/01 Do not log submission mode rewrites if they do not change the address.
2894 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2896 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2897 log files in place. Contributed by Roberto Lima.
2899 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2901 TF/03 Bugzilla 615: When checking the local_parts router precondition
2902 after a local_part_suffix or local_part_prefix option, Exim now
2903 does not use the address's named list lookup cache, since this
2904 contains cached lookups for the whole local part.
2906 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2907 Robert Millan. Documentation is in experimental-spec.txt.
2909 TF/04 Bugzilla 668: Fix parallel build (make -j).
2911 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2913 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2914 Patch provided by Jan Srzednicki.
2916 TF/05 Leading white space used to be stripped from $spam_report which
2917 wrecked the formatting. Now it is preserved.
2919 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2920 that they are available at delivery time.
2922 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2924 TF/08 TLS error reporting now respects the incoming_interface and
2925 incoming_port log selectors.
2927 TF/09 Produce a more useful error message if an SMTP transport's hosts
2928 setting expands to an empty string.
2930 NM/06 Bugzilla 744: EXPN did not work under TLS.
2931 Patch provided by Phil Pennock.
2933 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2934 Patch provided by Richard Godbee.
2936 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2937 acl_smtp_notquit, added index entry.
2939 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2940 Patch provided by Eugene Bujak.
2942 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2943 accept(). Patch provided by Maxim Dounin.
2945 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2946 Patch provided by Phil Pennock.
2948 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2950 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2951 Patch provided by Brad "anomie" Jorsch.
2953 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2954 Patch provided by Dean Brooks.
2956 TK/05 Add native DKIM support (does not depend on external libraries).
2958 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2959 Patch provided by Graeme Fowler.
2961 NM/16 Bugzilla 851: Documentation example syntax fix.
2963 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2965 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2968 NM/19 Bugzilla 745: TLS version reporting.
2969 Patch provided by Phil Pennock.
2971 NM/20 Bugzilla 167: bool: condition support.
2972 Patch provided by Phil Pennock.
2974 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2975 clients. Patch provided by Phil Pennock.
2977 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2978 Patch provided by Brad "anomie" Jorsch.
2980 NM/23 Bugzilla 687: Fix misparses in eximstats.
2981 Patch provided by Heiko Schlittermann.
2983 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2984 Patch provided by Heiko Schlittermann.
2986 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2987 plus update to original patch.
2989 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2991 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2992 Patch provided by David Brownlee.
2994 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2996 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2998 NM/30 Bugzilla 888: TLS documentation bugfixes.
3000 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3002 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3003 Unlike the original bugzilla I have changed all shell scripts in src tree.
3005 NM/33 Bugzilla 898: Transport filter timeout fix.
3006 Patch by Todd Rinaldo.
3008 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3009 Patch by Serge Demonchaux.
3011 NM/35 Bugzilla 39: Base64 decode bug fixes.
3012 Patch by Jakob Hirsch.
3014 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3016 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3018 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3020 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3026 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3027 ALT-N's libdkim that I have put here:
3028 http://duncanthrax.net/exim-experimental/
3030 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3031 were called 'true' and it seems that C99 defines that as a reserved
3032 keyword to be used with 'bool' variable types. That means you could
3033 not include C99-style headers which use bools without triggering
3034 build errors in sieve.c.
3036 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3037 as mailq or other aliases. Changed the --help handling significantly
3038 to do whats expected. exim_usage() emits usage/help information.
3040 SC/01 Added the -bylocaldomain option to eximstats.
3042 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3044 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3046 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3052 PH/01 Another patch from the Sieve maintainer.
3054 PH/02 When an IPv6 address is converted to a string for single-key lookup
3055 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3056 dots are used instead of colons so that keys in lsearch files need not
3057 contain colons. This was done some time before quoting was made available
3058 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3059 (notated using the quote facility) so as to distinguish them from IPv4
3060 keys. This meant that lookups for IP addresses in host lists did not work
3061 for iplsearch lookups.
3063 This has been fixed by arranging for IPv6 addresses to be expressed with
3064 colons if the lookup type is iplsearch. This is not incompatible, because
3065 previously such lookups could never work.
3067 The situation is now rather anomalous, since one *can* have colons in
3068 ordinary lsearch keys. However, making the change in all cases is
3069 incompatible and would probably break a number of configurations.
3071 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3074 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3075 conversion specification without a maximum field width, thereby enabling
3076 a rogue spamd server to cause a buffer overflow. While nobody in their
3077 right mind would setup Exim to query an untrusted spamd server, an
3078 attacker that gains access to a server running spamd could potentially
3079 exploit this vulnerability to run arbitrary code as the Exim user.
3081 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3082 $primary_hostname instead of what libspf2 thinks the hosts name is.
3084 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3085 a directory entry by the name of the lookup key. Previously, if a
3086 symlink pointed to a non-existing file or a file in a directory that
3087 Exim lacked permissions to read, a lookup for a key matching that
3088 symlink would fail. Now it is enough that a matching directory entry
3089 exists, symlink or not. (Bugzilla 503.)
3091 PH/03 The body_linecount and body_zerocount variables are now exported in the
3094 PH/04 Added the $dnslist_matched variable.
3096 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3097 This means they are set thereafter only if the connection becomes
3100 PH/06 Added the client_condition to authenticators so that some can be skipped
3101 by clients under certain conditions.
3103 PH/07 The error message for a badly-placed control=no_multiline_responses left
3104 "_responses" off the end of the name.
3106 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3108 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3109 (without spaces) instead of just copying the configuration text.
3111 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3113 PH/11 Added $max_received_linelength.
3115 PH/12 Added +ignore_defer and +include_defer to host lists.
3117 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3118 way in which PCRE > 7.0 is built.
3120 PH/14 Implemented queue_only_load_latch.
3122 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3123 MAIL command. The effect was to mangle the value on 64-bit systems.
3125 PH/16 Another patch from the Sieve maintainer.
3127 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3129 PH/18 If a system quota error occurred while trying to create the file for
3130 a maildir delivery, the message "Mailbox is full" was not appended to the
3131 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3132 only to a quota excession during the actual writing of the file.
3134 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3135 characters?) which causes problems in log lines. The DN values are now
3136 passed through string_printing() before being added to log lines.
3138 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3139 and InterBase are left for another time.)
3141 PH/21 Added message_body_newlines option.
3143 PH/22 Guard against possible overflow in moan_check_errorcopy().
3145 PH/23 POSIX allows open() to be a macro; guard against that.
3147 PH/24 If the recipient of an error message contained an @ in the local part
3148 (suitably quoted, of course), incorrect values were put in $domain and
3149 $local_part during the evaluation of errors_copy.
3155 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3156 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3159 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3160 issue a MAIL command.
3162 PH/02 In an ACL statement such as
3164 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3166 if a client was not listed at all, or was listed with a value other than
3167 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3168 the condition was not true (as it should be), so access was not denied.
3169 The bug was that the ! inversion was incorrectly passed on to the second
3170 item. This has been fixed.
3172 PH/03 Added additional dnslists conditions == and =& which are different from
3173 = and & when the dns lookup returns more than one IP address.
3175 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3176 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3178 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3179 FSYNC, which compiles an option called disable_fsync that allows for
3180 bypassing fsync(). The documentation is heavily laced with warnings.
3182 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3184 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3185 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3186 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3187 including adding "make clean"; (3) Added -fPIC when compiling the test
3188 dynamically loaded module, to get rid of a warning.
3190 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3191 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3192 = 0s. The bug is otherwise harmless.
3194 PH/07 There was a bug in the dovecot authenticator such that the value of
3195 $auth1 could be overwritten, and so not correctly preserved, after a
3196 successful authentication. This usually meant that the value preserved by
3197 the server_setid option was incorrect.
3199 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3201 PH/09 Installed PCRE release 7.0.
3203 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3204 run for batched SMTP input. It is now run at the start of every message
3205 in the batch. While fixing this I discovered that the process information
3206 (output by running exiwhat) was not always getting set for -bs and -bS
3207 input. This is fixed, and it now also says "batched" for BSMTP.
3209 PH/11 Added control=no_pipelining.
3211 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3212 patch, slightly modified), and move the expansion of helo_data till after
3213 the connection is made in the smtp transport (so it can use these
3216 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3218 PH/14 Added log_selector = +pid.
3220 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3222 PH/16 Add ${if forany and ${if forall.
3224 PH/17 Added dsn_from option to vary the From: line in DSNs.
3226 PH/18 Flush SMTP output before performing a callout, unless control =
3227 no_callout_flush is set.
3229 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3230 was true (the default) a successful delivery failed to delete the retry
3231 item, thus causing premature timeout of the address. The bug is now
3234 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3236 PH/21 Long custom messages for fakedefer and fakereject are now split up
3237 into multiline responses in the same way that messages for "deny" and
3238 other ACL rejections are.
3240 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3241 with slight modification.
3243 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3244 draft, changing the syntax and factoring some duplicate code".
3246 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3247 for deliveries of the second and subsequent messages over the same SMTP
3250 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3251 ${reduce, with only minor "tidies".
3253 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3255 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3256 expansion side effects.
3258 PH/27 When a message times out after an over-quota error from an Exim-imposed
3259 quota, the bounce message says "mailbox is full". This message was not
3260 being given when it was a system quota that was exceeded. It now should
3263 MH/03 Made $recipients available in local_scan(). local_scan() already has
3264 better access to the recipient list through recipients_list[], but
3265 $recipients can be useful in postmaster-provided expansion strings.
3267 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3268 in the case of a MAIL command with additional options following the
3269 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3270 were accidentally chopped off.
3272 PH/29 SMTP synchronization checks are implemented when a command is read -
3273 there is a check that no more input is waiting when there shouldn't be
3274 any. However, for some commands, a delay in an ACL can mean that it is
3275 some time before the response is written. In this time, more input might
3276 arrive, invalidly. So now there are extra checks after an ACL has run for
3277 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3278 pipelining has not been advertised.
3280 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3282 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3283 returned in the SMTP response when smtp_return_error_details was false.
3284 This has been fixed.
3286 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3287 instead of the C I/O that was originally supplied, because problems were
3288 reported on Solaris.
3290 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3291 Exim which did not show up earlier: it was assuming that a call to
3292 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3293 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3294 was a macro that became an assignment, so it seemed to work. This has
3295 changed to a proper function call with a void return, hence the compile
3296 error. Exim's code has been fixed.
3298 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3301 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3303 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3305 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3306 because it turns out to be too big or there is a timeout) there is no
3307 "Completed" line in the log. When some messages of this type were
3308 selected by exigrep, they were listed as "not completed". Others were
3309 picked up by some special patterns. I have improved the selection
3310 criteria to be more general.
3312 PH/38 The host_find_failed option in the manualroute router can now be set
3313 to "ignore", to completely ignore a host whose IP address cannot be
3314 found. If all hosts are ignored, the behaviour is controlled by the new
3315 host_all_ignored option.
3317 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3318 homing or because of multiple MX records with /mx) generated more than
3319 one IP address, and the following item turned out to be the local host,
3320 all the secondary addresses of the first item were incorrectly removed
3321 from the list, along with the local host and any following hosts (which
3322 is what is supposed to happen).
3324 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3325 whoever called Exim into the -H file. In the case of the daemon it was
3326 behaving confusingly. When first started, it used values for whoever
3327 started the daemon, but after a SIGHUP it used the Exim user (because it
3328 calls itself on a restart). I have changed the code so that it now always
3331 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3332 message are rejected with the same error (e.g. no authentication or bad
3333 sender address), and a DATA command is nevertheless sent (as can happen
3334 with PIPELINING or a stupid MUA), the error message that was given to the
3335 RCPT commands is included in the rejection of the DATA command. This is
3336 intended to be helpful for MUAs that show only the final error to their
3339 PH/42 Another patch from the Sieve maintainer.
3341 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3342 Eximstats - Fixed some broken HTML links and added missing column headers
3344 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3345 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3347 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3349 PH/43 Yet another patch from the Sieve maintainer.
3351 PH/44 I found a way to check for a TCP/IP connection going away before sending
3352 the response to the final '.' that terminates a message, but only in the
3353 case where the client has not sent further data following the '.'
3354 (unfortunately, this is allowed). However, in many cases there won't be
3355 any further data because there won't be any more messages to send. A call
3356 to select() can be used: if it shows that the input is "ready", there is
3357 either input waiting, or the socket has been closed. An attempt to read
3358 the next input character can distinguish the two cases. Previously, Exim
3359 would have sent an OK response which the client would never have see.
3360 This could lead to message repetition. This fix should cure that, at
3361 least in a lot of common cases.
3363 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3364 advertised in response to EHLO.
3370 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3371 fixed by 4.65/MH/01 (is this a record?) are fixed:
3373 (i) An empty string was always treated as zero by the numeric comparison
3374 operators. This behaviour has been restored.
3376 (ii) It is documented that the numeric comparison operators always treat
3377 their arguments as decimal numbers. This was broken in that numbers
3378 starting with 0 were being interpreted as octal.
3380 While fixing these problems I realized that there was another issue that
3381 hadn't been noticed. Values of message_size_limit (both the global option
3382 and the transport option) were treated as octal if they started with 0.
3383 The documentation was vague. These values are now always treated as
3384 decimal, and I will make that clear in the documentation.
3390 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3391 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3394 MH/01 Don't check that the operands of numeric comparison operators are
3395 integers when their expansion is in "skipping" mode (fixes bug
3396 introduced by 4.64-PH/07).
3398 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3399 child addresses, Exim now panics and dies. Previously, because the count
3400 is held in a short int, deliveries were likely to be lost. As such a
3401 large number of recipients for a single message is ridiculous
3402 (performance will be very, very poor), I have chosen to impose a limit
3403 rather than extend the field.
3409 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3410 leftover -K file (the existence of which was triggered by #402).
3411 While we were at it, introduced process PID as part of the -K
3412 filename. This should rule out race conditions when creating
3415 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3416 processing considerably. Previous code took too long for large mails,
3417 triggering a timeout which in turn triggers #401.
3419 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3420 in the DK code in transports.c. sendfile() is not really portable,
3421 hence the _LINUX specificness.
3423 TF/01 In the add_headers option to the mail command in an Exim filter,
3424 there was a bug that Exim would claim a syntax error in any
3425 header after the first one which had an odd number of characters
3428 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3429 callout verification, Exim cached a "reject" for the entire domain. This
3430 is correct for most verifications, but it is not correct for a recipient
3431 verification with use_sender or use_postmaster set, because in that case
3432 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3433 case of MAIL FROM:<> rejection from other early rejections (e.g.
3434 rejection of HELO). When verifying a recipient using a non-null MAIL
3435 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3436 Whatever the result of the callout, the value of the domain cache is
3437 left unchanged (for any other kind of callout, getting as far as trying
3438 RCPT means that the domain itself is ok).
3440 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3443 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3444 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3445 session, the daemon ignores ECONNECT errors and logs others; it now
3446 ignores EPIPE as well.
3448 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3449 (quoted-printable decoding).
3451 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3452 later the small subsequent patch to fix an introduced bug.
3454 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3456 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3458 PH/08 An error is now given if message_size_limit is specified negative.
3460 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3461 to be given (somewhat) arbitrary names.
3463 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3466 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3467 miscellaneous code fixes
3469 PH/10 Added the log_reject_target ACL modifier to specify where to log
3472 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3473 hostname. This is wrong, because it relates to the incoming message (and
3474 probably the interface on which it is arriving) and not to the outgoing
3475 callout (which could be using a different interface). This has been
3476 changed to use the value of the helo_data option from the smtp transport
3477 instead - this is what is used when a message is actually being sent. If
3478 there is no remote transport (possible with a router that sets up host
3479 addresses), $smtp_active_hostname is used.
3481 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3482 tweaks were necessary in order to get it to work (see also 21 below):
3483 (a) The code assumed that strncpy() returns a negative number on buffer
3484 overflow, which isn't the case. Replaced with Exim's string_format()
3486 (b) There were several signed/unsigned issues. I just did the minimum
3487 hacking in of casts. There is scope for a larger refactoring.
3488 (c) The code used strcasecmp() which is not a standard C function.
3489 Replaced with Exim's strcmpic() function.
3490 (d) The code set only $1; it now sets $auth1 as well.
3491 (e) A simple test gave the error "authentication client didn't specify
3492 service in request". It would seem that Dovecot has changed its
3493 interface. Fortunately there's a specification; I followed it and
3494 changed what the client sends and it appears to be working now.
3496 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3499 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3500 address A is aliased to B and C, where B exists and C does not. Without
3501 -v the output is "A verified" because verification stops after a
3502 successful redirection if more than one address is generated. However,
3503 with -v the child addresses are also verified. Exim was outputting "A
3504 failed to verify" and then showing the successful verification for C,
3505 with its parentage. It now outputs "B failed to verify", showing B's
3506 parentage before showing the successful verification of C.
3508 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3509 look up a TXT record in a specific list after matching in a combined
3512 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3513 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3514 they consult the DNS. I had assumed they would set it the way they
3515 wanted; and indeed my experiments on Linux seem to show that in some
3516 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3517 To be on the safe side, however, I have now made the interface to
3518 host_find_byname() similar to host_find_bydns(), with an argument
3519 containing the DNS resolver options. The host_find_byname() function now
3520 sets these options at its start, just as host_find_bydns() does. The smtp
3521 transport options dns_qualify_single and dns_search_parents are passed to
3522 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3523 of host_find_byname() use the default settings of RES_DEFNAMES
3524 (qualify_single) but not RES_DNSRCH (search_parents).
3526 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3527 spool_read_header() do less string testing, by means of a preliminary
3528 switch on the second character of optional "-foo" lines. (This is
3529 overdue, caused by the large number of possibilities that now exist.
3530 Originally there were few.) While I was there, I also converted the
3531 str(n)cmp tests so they don't re-test the leading "-" and the first
3532 character, in the hope this might squeeze out yet more improvement.
3534 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3535 flag allowing group syntax was set by the header_syntax check but not
3536 turned off, possible causing trouble later; (2) The flag was not being
3537 set at all for the header_verify test, causing "group"-style headers to
3538 be rejected. I have now set it in this case, and also caused header_
3539 verify to ignore an empty address taken from a group. While doing this, I
3540 came across some other cases where the code for allowing group syntax
3541 while scanning a header line wasn't quite right (mostly, not resetting
3542 the flag correctly in the right place). These bugs could have caused
3543 trouble for malformed header lines. I hope it is now all correct.
3545 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3546 with the "reply" argument non-NULL. The code, however (which originally
3547 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3548 but it didn't always do it. This confused somebody who was copying the
3549 code for some other use. I have removed all the tests.
3551 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3552 feature that was used to support insecure browsers during the U.S. crypto
3553 embargo. It requires special client support, and Exim is probably the
3554 only MTA that supported it -- and would never use it because real RSA is
3555 always available. This code has been removed, because it had the bad
3556 effect of slowing Exim down by computing (never used) parameters for the
3557 RSA_EXPORT functionality.
3559 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3560 authenticator to fail if there's a tab character in the incoming data
3561 (there should never be unless someone is messing about, as it's supposed
3562 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3563 if the connection is using TLS or if the remote IP is the same as the
3564 local IP, and the "valid-client-cert option" if a client certificate has
3567 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3568 authenticators. This can be used for authorization after authentication
3569 succeeds. (In the case of plaintext, it servers for both authentication
3572 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3573 if any retry times were supplied.
3575 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3576 connection, where there is no client IP address to check. In this
3577 situation, the verify now always succeeds.
3579 PH/25 Applied John Jetmore's -Mset patch.
3581 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3583 PH/27 In a string expansion for a processed (not raw) header when multiple
3584 headers of the same name were present, leading whitespace was being
3585 removed from all of them, but trailing whitespace was being removed only
3586 from the last one. Now trailing whitespace is removed from each header
3587 before concatenation. Completely empty headers in a concatenation (as
3588 before) are ignored.
3590 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3591 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3593 PH/29 [Removed. This was a change that I later backed out, and forgot to
3594 correct the ChangeLog entry (that I had efficiently created) before
3595 committing the later change.]
3597 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3598 address errors (4xx response to RCPT) over the same connection as other
3599 messages routed to the same hosts. Such deliveries are always "forced",
3600 so retry times are not inspected. This resulted in far too many retries
3601 for the affected addresses. The effect occurred only when there were more
3602 hosts than the hosts_max_try setting in the smtp transport when it had
3603 the 4xx errors. Those hosts that it had tried were not added to the list
3604 of hosts for which the message was waiting, so if all were tried, there
3605 was no problem. Two fixes have been applied:
3607 (i) If there are any address or message errors in an SMTP delivery, none
3608 of the hosts (tried or untried) are now added to the list of hosts
3609 for which the message is waiting, so the message should not be a
3610 candidate for sending over the same connection that was used for a
3611 successful delivery of some other message. This seems entirely
3612 reasonable: after all the message is NOT "waiting for some host".
3613 This is so "obvious" that I'm not sure why it wasn't done
3614 previously. Hope I haven't missed anything, but it can't do any
3615 harm, as the worst effect is to miss an optimization.
3617 (ii) If, despite (i), such a delivery is accidentally attempted, the
3618 routing retry time is respected, so at least it doesn't keep
3619 hammering the server.
3621 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3622 in ${readsocket because some servers need this prod.
3624 PH/32 Added some extra debug output when updating a wait-xxx database.
3626 PH/33 The hint "could be header name not terminated by colon", which has been
3627 given for certain expansion errors for a long time, was not being given
3628 for the ${if def:h_colon_omitted{... case.
3630 PH/34 The spec says: "With one important exception, whenever a domain list is
3631 being scanned, $domain contains the subject domain." There was at least
3632 one case where this was not true.
3634 PH/35 The error "getsockname() failed: connection reset by peer" was being
3635 written to the panic log as well as the main log, but it isn't really
3636 panic-worthy as it just means the connection died rather early on. I have
3637 removed the panic log writing for the ECONNRESET error when getsockname()
3640 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3641 runs only) independently of the message's sender address. This meant
3642 that, if the 4xx error was in fact related to the sender, a different
3643 message to the same recipient with a different sender could confuse
3644 things. In particular, this can happen when sending to a greylisting
3645 server, but other circumstances could also provoke similar problems.
3646 I have changed the default so that the retry time for these errors is now
3647 based a combination of the sender and recipient addresses. This change
3648 can be overridden by setting address_retry_include_sender=false in the
3651 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3652 remote server are returned as part of bounce messages. This was not
3653 happening for LMTP over a pipe (the lmtp transport), but now it is the
3654 same for both kinds of LMTP.
3656 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3657 in header lines that were in fact CNAMEs. This is no longer the case.
3659 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3660 and queue runs started by the daemon processed all messages. This has
3661 been fixed so that -R and -S can now usefully be given with -q<time>.
3663 PH/40 Import PCRE release 6.7 (fixes some bugs).
3665 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3667 PH/42 Give an error if -q is specified more than once.
3669 PH/43 Renamed the variables $interface_address and $interface_port as
3670 $received_ip_address and $received_port, to make it clear that these
3671 values apply to message reception, and not to the outgoing interface when
3672 a message is delivered. (The old names remain recognized, of course.)
3674 PH/44 There was no timeout on the connect() call when using a Unix domain
3675 socket in the ${readsocket expansion. There now is.
3677 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3678 be meaningful with "accept".
3680 SC/01 Eximstats V1.43
3681 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3683 SC/02 Eximstats V1.44
3684 Use a glob alias rather than an array ref in the generated
3685 parser. This improves both readability and performance.
3687 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3688 Collect SpamAssassin and rejection statistics.
3689 Don't display local sender or destination tables unless
3690 there is data to show.
3691 Added average volumes into the top table text output.
3693 SC/04 Eximstats V1.46
3694 Collect data on the number of addresses (recipients)
3695 as well as the number of messages.
3697 SC/05 Eximstats V1.47
3698 Added 'Message too big' to the list of mail rejection
3699 reasons (thanks to Marco Gaiarin).
3701 SC/06 Eximstats V1.48
3702 Mainlog lines which have GMT offsets and are too short to
3703 have a flag are now skipped.
3705 SC/07 Eximstats V1.49 (Alain Williams)
3706 Added the -emptyok flag.
3708 SC/08 Eximstats V1.50
3709 Fixes for obtaining the IP address from reject messages.
3711 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3712 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3713 whitespace changes from 4.64-PH/27
3715 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3718 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3719 are found, allow negative numbers in numeric criteria)
3721 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3723 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3724 to match changes made in 4.64-PH/43
3726 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3728 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3729 "domain" is actually an IP address, and force a failure. This locks out
3730 those revolvers/nameservers that support "A-for-A" lookups, in
3731 contravention of the specifications.
3733 PH/48 When a host name was looked up from an IP address, and the subsequent
3734 forward lookup of the name timed out, the host name was left in
3735 $sender_host_name, contrary to the specification.
3737 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3738 restricted to single-key lookups, Exim was not diagnosing an error if
3739 * or *@ was used with a query-style lookup.
3741 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3743 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3744 long ago, but noone interested enough thought of it. Let's just say that
3745 the "1.1" means that there are some new functions that weren't there at
3746 some point in the past.
3748 PH/51 Error processing for expansion failure of helo_data from an smtp
3749 transport during callout processing was broken.
3751 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3752 tested/used via the -bh/-bhc/-bs options.
3754 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3755 bug, fixed in subsequent PCRE releases).
3757 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3758 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3760 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3766 SC/01 Use a glob alias rather than an array ref in eximstats generated
3767 parser. This improves both readability and performance.
3769 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3770 Don't display local sender or destination tables in eximstats unless
3771 there is data to show.
3772 Added average volumes into the eximstats top table text output.
3774 SC/03 Collect data on the number of addresses (recipients) as well
3775 as the number of messages in eximstats.
3777 TF/01 Correct an error in the documentation for the redirect router. Exim
3778 does (usually) call initgroups() when daemonizing.
3780 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3781 with consistent privilege compared to when running as a daemon.
3783 TF/03 Note in the spec that $authenticated_id is not set for local
3784 submissions from trusted users.
3786 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3787 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3789 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3790 by adding some example configuration directives to the default
3791 configuration file. A little bit of work is required to uncomment the
3792 directives and define how usernames and passwords are checked, but
3793 there is now a framework to start from.
3795 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3796 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3797 without this. I don't know how relevant this is to other LDAP libraries.
3799 PH/02 Add the verb name to the "unknown ACL verb" error.
3801 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3803 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3805 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3806 directory not expanded when it should be if an expanded home directory
3807 was set for the address (which is overridden by the transport).
3809 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3812 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3813 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3814 because it is too late at that time, and has no effect.
3816 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3817 security issue with \' (bugzilla #107). I could not use the
3818 PQescapeStringConn() function, because it needs a PGconn value as one of
3821 PH/08 When testing addresses using -bt, indicate those final addresses that
3822 are duplicates that would not cause an additional delivery. At least one
3823 person was confused, thinking that -bt output corresponded to deliveries.
3824 (Suppressing duplicates isn't a good idea as you lose the information
3825 about possibly different redirections that led to the duplicates.)
3827 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3828 systems where poll() doesn't work, in particular OS X.
3830 PH/10 Added more information to debugging output for retry time not reached.
3832 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3833 operations in malware.c.
3835 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3838 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3839 syslog_duplication set false, log lines that would normally be written
3840 both the the main log and to the reject log were not written to syslog at
3843 PH/14 In the default configuration, change the use of "message" in ACL warn
3844 statements to "add_header".
3846 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3847 not followed by a command (e.g. "seen endif").
3849 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3850 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3853 PH/17 Added extra conditions to the default value of delay_warning_condition
3857 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3858 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3859 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3862 The Auto-Submitted: and various List- headers are standardised, whereas I
3863 don't think Precedence: ever was.
3865 PH/18 Refactored debugging code in route_finduser() to show more information,
3866 in particular, the error code if getpwnam() issues one.
3868 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3869 This is apparently needed in addition to the PH/07 change above to avoid
3870 any possible encoding problems.
3872 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3873 but not after initializing Perl.
3875 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3876 output them only if debugging. By default they are written stderr,
3877 apparently, which is not desirable.
3879 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3882 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3885 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3887 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3888 authenticated or an ident call has been made. Suppress the default
3889 values for $authenticated_id and $authenticated_sender (but permit -oMai
3890 and -oMas) when testing with -bh.
3892 PH/24 Re-jigged the order of the tests in the default configuration so that the
3893 tests for valid domains and recipients precede the DNS black list and CSA
3894 tests, on the grounds that those ones are more expensive.
3896 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3897 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3898 This bug exists in every version of Exim that I still have, right back to
3901 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3902 However, an attempt to turn on case-sensitivity in a regex key by
3903 including (?-i) didn't work because the subject string was already
3904 lowercased, and the effects were non-intuitive. It turns out that a
3905 one-line patch can be used to allow (?-i) to work as expected.
3911 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3912 other effects) broke the use of negated acl sub-conditions.
3914 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3917 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3918 "Deny" causes Exim to reject the incoming connection with a 554 error.
3919 Unfortunately, if there is a major crisis, such as a disk failure,
3920 tcp-wrappers gives "deny", whereas what one would like would be some
3921 kind of temporary error. A kludge has been added to help with this.
3922 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3923 554 error is used if errno is still zero or contains ENOENT (which occurs
3924 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3927 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3929 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3930 errors. Otherwise a message that provokes a temporary error (when other
3931 messages do not) can cause a whole host to time out.
3933 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3934 addresses were routed directly to files or pipes from a redirect router.
3935 File deliveries just didn't batch; pipe deliveries might have suffered
3938 PH/06 A failure to get a lock for a hints database would erroneously always say
3939 "Failed to get write lock", even when it was really a read lock.
3941 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3942 of 0600. This has been changed to use the value of the lockfile_mode
3943 option (which defaults to 0600).
3945 PH/08 Applied small patch from the Sieve maintainer.
3947 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3948 folder from quota calculations, a direct delivery into this folder messed
3949 up the contents of the maildirsize file. This was because the regex was
3950 used only to exclude .Trash (or whatever) when the size of the mailbox
3951 was calculated. There was no check that a delivery was happening into an
3952 excluded directory. This bug has been fixed by ignoring all quota
3953 processing for deliveries into excluded directories.
3955 PH/10 Added the maildirfolder_create_regex option to appendfile.
3961 PH/01 The code for finding all the local interface addresses on a FreeBSD
3962 system running IPv6 was broken. This may well have applied to all BSD
3963 systems, as well as to others that have similar system calls. The broken
3964 code found IPv4 interfaces correctly, but gave incorrect values for the
3965 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3966 that it would not match correctly against @[] and not recognize the IPv6
3969 PH/02 The ipliteral router was not recognizing addresses of the form user@
3970 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3972 PH/03 Added disable_ipv6.
3974 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3975 decoded form, because it is most often used to construct To: headers
3976 lines in autoreplies, and the decoded form may well be syntactically
3977 invalid. However, $reply_address has leading white space removed, and all
3978 newlines turned into spaces so that the autoreply transport does not
3981 PH/05 If group was specified without a user on a router, and no group or user
3982 was specified on a transport, the group from the router was ignored.
3984 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3985 for visible compile-time settings that can be used to change these
3986 numbers, for those that want even more. Backwards compatibility with old
3987 spool files has been maintained. However, going back to a previous Exim
3988 release will lost any variables that are in spool files.
3990 PH/07 Two small changes when running in the test harness: increase delay when
3991 passing a TCP/IP connection to a new process, in case the original
3992 process has to generate a bounce, and remove special handling of
3993 127.0.0.2 (sic), which is no longer necessary.
3995 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3996 be the same on different OS.
3998 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4001 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4002 whether --show-vars was specified or not
4004 JJ/02 exipick: Added support for new ACL variable spool format introduced
4007 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4008 syntactically invalid From: or Reply-to: line, and a filter used this to
4009 generate an autoreply, and therefore failed to obtain an address for the
4010 autoreply, Exim could try to deliver to a non-existent relative file
4011 name, causing unrelated and misleading errors. What now happens is that
4012 it logs this as a hard delivery error, but does not attempt to create a
4015 PH/11 The exinext utility has a -C option for testing purposes, but although
4016 the given file was scanned by exinext itself; it wasn't being passed on
4017 when Exim was called.
4019 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4020 an end-of-file indication when reading a command response.
4022 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4023 compiled. In many other places in Exim, IPv6 addresses are always
4024 recognized, so I have changed this. It also means that IPv4 domain
4025 literals of the form [IPV4:n.n.n.n] are now always recognized.
4027 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4028 used if the router is not running as root, for example, when verifying at
4029 ACL time, or when using -bh. The debugging output from this situation was
4030 non-existent - all you got was a failure to exec. I have made two
4033 (a) Failures to set uid/gid, the current directory, or a process leader
4034 in a subprocess such as that created by queryprogram now generate
4035 suitable debugging output when -d is set.
4037 (b) The queryprogram router detects when it is not running as root,
4038 outputs suitable debugging information if -d is set, and then runs
4039 the subprocess without attempting to change uid/gid.
4041 PH/15 Minor change to Makefile for building test_host (undocumented testing
4044 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4045 additional section of a DNS packet that returns MX or SRV records.
4046 Instead, it always explicitly searches for A/AAAA records. This avoids
4047 major problems that occur when a DNS server includes only records of one
4048 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4049 fixed another bug: if SRV records were looked up and the corresponding
4050 address records were *not* found in the additional section, the port
4051 values from the SRV records were lost.
4053 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4054 using the correct key (the original address) when searching the retry
4055 rules in order to find which one to use for generating the retry hint.
4057 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4058 adding the default one. Similarly, if it contains a Reply-To: header, the
4059 errors_reply_to option, if set, is not used.
4061 PH/19 When calculating a retry time, Exim used to measure the "time since
4062 failure" by looking at the "first failed" field in the retry record. Now
4063 it does not use this if it is later than than the arrival time of the
4064 message. Instead it uses the arrival time. This makes for better
4065 behaviour in cases where some deliveries succeed, thus re-setting the
4066 "first failed" field. An example is a quota failure for a huge message
4067 when small messages continue to be delivered. Without this change, the
4068 "time since failure" will always be short, possible causing more frequent
4069 delivery attempts for the huge message than are intended.
4070 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4072 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4073 $1, $2, $3) because the numerical variables can be reset during some
4074 expansion items (e.g. "match"), thereby losing the authentication data.
4076 PH/21 Make -bV show the size of off_t variables so that the test suite can
4077 decide whether to run tests for quotas > 2G.
4079 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4080 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4081 filecount value is greater than 2G or if a quota value is greater than 2G
4082 on a system where the size of off_t is not greater than 4, a panic error
4085 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4086 never match. The debug and -bh output now contains an explicit error
4087 message indicating a malformed IPv4 address or mask.
4089 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4090 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4091 PH/23 above applies.
4093 PH/25 Do not write to syslog when running in the test harness. The only
4094 occasion when this arises is a failure to open the main or panic logs
4095 (for which there is an explicit test).
4097 PH/26 Added the /no_tell option to "control=freeze".
4099 PH/27 If a host name lookup failed very early in a connection, for example, if
4100 the IP address matched host_lookup and the reverse lookup yielded a name
4101 that did not have a forward lookup, an error message of the form "no IP
4102 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4103 could be logged. Now it outputs the IP address instead of "NULL".
4105 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4106 allows the sender and the authenticated sender to be set when
4107 submitting a message from within Exim. Since child_open_exim() is
4108 documented for local_scan(), the new function should be too.
4110 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4111 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4112 results in an empty string is now treated as unset.
4114 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4116 PH/31 Added sender_verify_fail logging option.
4118 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4119 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4120 tidied the source and removed it altogether.
4122 PH/33 When a queue run was abandoned because the load average was too high, a
4123 log line was always written; now it is written only if the queue_run log
4124 selector is set. In addition, the log line for abandonment now contains
4125 information about the queue run such as the pid. This is always present
4126 in "start" and "stop" lines but was omitted from the "abandon" line.
4128 PH/34 Omit spaces between a header name and the colon in the error message that
4129 is given when verify = headers_syntax fails (if there are lots of them,
4130 the message gets confusing).
4132 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4133 names, as there are now some PTR records that contain slashes. This check
4134 is only to protect against broken name servers that fall over on strange
4135 characters, so the fact that it applies to all lookups doesn't matter.
4137 PH/36 Now that the new test suite is complete, we can remove some of the
4138 special code in Exim that was needed for the old test suite. For example,
4139 sorting DNS records because real resolvers return them in an arbitrary
4140 order. The new test suite's fake resolver always returns records in the
4143 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4144 bounces) except when queue_only is set, to avoid logging races between
4145 the different processes.
4147 PH/38 Panic-die if .include specifies a non-absolute path.
4149 PH/39 A tweak to the "H" retry rule from its user.
4151 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4152 a label. They prevented compilation on older perls.
4154 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4155 a warning to be raised on newish perls.
4157 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4158 on queue. Changes to match documented behaviour of showing count of
4159 messages matching specified criteria.
4161 PH/40 Changed the default ident timeout from 30s to 5s.
4163 PH/41 Added support for the use of login_cap features, on those BSD systems
4164 that have them, for controlling the resources used by pipe deliveries.
4166 PH/42 The content-scanning code uses fopen() to create files in which to put
4167 message data. Previously it was not paying any attention to the mode of
4168 the files. Exim runs with umask(0) because the rest of the code creates
4169 files with open(), and sets the required mode explicitly. Thus, these
4170 files were ending up world-writeable. This was not a big issue, because,
4171 being within the spool directory, they were not world-accessible. I have
4172 created a function called modefopen, which takes an additional mode
4173 argument. It sets umask(777), creates the file, chmods it to the required
4174 mode, then resets the umask. All the relevant calls to fopen() in the
4175 content scanning code have been changed to use this function.
4177 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4178 to 24 hours. This avoids potential overflow problems when processing G
4179 and H retry rules. I suspect nobody ever tinkers with this value.
4181 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4183 PH/45 When the plaintext authenticator is running as a client, the server's
4184 challenges are checked to ensure they are valid base64 strings. By
4185 default, the authentication attempt is cancelled if an invalid string is
4186 received. Setting client_ignore_invalid_base64 true ignores these errors.
4187 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4188 they are received. Thus, the responses can be made to depend on the
4189 challenges. If an invalid string is ignored, an empty string is placed in
4192 PH/46 Messages that are created by the autoreply transport now contains a
4193 References: header, in accordance with RFCs 2822 and 3834.
4195 PH/47 Added authenticated_sender_force to the smtp transport.
4197 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4199 PH/49 Installed latest patch from the Sieve maintainer.
4201 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4202 was also set, the appendfile transport was unnecessarily scanning a
4203 directory of message files (e.g. for maildir delivery) to find the count
4204 of files (along with the size), even though it did not need this
4205 information. It now does the scan only if it needs to find either the
4206 size of the count of files.
4208 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4210 PH/52 Two bugs concerned with error handling when the smtp transport is
4213 (i) Exim was not creating retry information for temporary errors given
4214 for individual recipients after the DATA command when the smtp transport
4215 was used in LMTP mode. This meant that they could be retried too
4216 frequently, and not timed out correctly.
4218 (ii) Exim was setting the flag that allows error details to be returned
4219 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4220 recipients that were returned after the DATA command.
4222 PH/53 This is related to PH/52, but is more general: for any failing address,
4223 when detailed error information was permitted to be returned to the
4224 sender, but the error was temporary, then after the final timeout, only
4225 "retry timeout exceeded" was returned. Now it returns the full error as
4226 well as "retry timeout exceeded".
4228 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4229 do this, and (what is worse) MTAs that accept it.
4231 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4232 will now be deprecated.
4234 PH/56 New os.c-cygwin from the Cygwin maintainer.
4236 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4237 formats (previously only available in exim formats via -bpr, -bpru,
4238 and -bpra. Now also available in native and exiqgrep formats)
4240 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4241 with very large, slow to parse queues
4243 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4245 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4247 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4248 responses to authentication challenges, though it was showing the
4249 challenges; (ii) I've removed the CR characters from the debug output for
4252 PH/58 Allow for the insertion of a newline as well as a space when a string
4253 is turned into more than one encoded-word during RFC 2047 encoding. The
4254 Sieve code now uses this.
4256 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4257 data_4xx, lost_connection, tls_required.
4259 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4260 message was being sent as an SMTP response.
4262 PH/61 Add -l and -k options to exicyclog.
4264 PH/62 When verifying, if an address was redirected to one new address, so that
4265 verification continued, and the new address failed or deferred after
4266 having set something in $address_data, the value of $address_data was not
4267 passed back to the ACL. This was different to the case when no
4268 redirection occurred. The value is now passed back in both cases.
4270 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4271 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4272 use login_cap.h, so on its own it isn't the distinguishing feature. The
4273 new name refers directly to the setclassresources() function.
4275 PH/65 Added configuration files for NetBSD3.
4277 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4279 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4280 is preferred over IPv4.
4282 PH/68 The bounce_return_message and bounce_return_body options were not being
4283 honoured for bounces generated during the reception of non-SMTP messages.
4284 In particular, this applied to messages rejected by the ACL. This bug has
4285 been fixed. However, if bounce_return_message is true and bounce_return_
4286 body is false, the headers that are returned for a non-SMTP message
4287 include only those that have been read before the error was detected.
4288 (In the case of an ACL rejection, they have all been read.)
4290 PH/69 The HTML version of the specification is now built in a directory called
4291 spec_html instead of spec.html, because the latter looks like a path with
4292 a MIME-type, and this confuses some software.
4294 PH/70 Catch two compiler warnings in sieve.c.
4296 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4297 function verify_get_ident() calls ip_connect() to connect a socket, but
4298 if the "connect()" function timed out, ip_connect() used to close the
4299 socket. However, verify_get_ident() also closes the socket later, and in
4300 between Exim writes to the log, which may get opened at this point. When
4301 the socket was closed in ip_connect(), the log could get the same file
4302 descriptor number as the socket. This naturally causes chaos. The fix is
4303 not to close the socket in ip_connect(); the socket should be closed by
4304 the function that creates it. There was only one place in the code where
4305 this was missing, in the iplookup router, which I don't think anybody now
4306 uses, but I've fixed it anyway.
4308 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4309 well as to direct DNS lookups. Otherwise the handling of names in host
4310 lists is inconsistent and therefore confusing.
4316 PH/01 Two changes to the default runtime configuration:
4318 (1) Move the checks for relay_from_hosts and authenticated clients from
4319 after to before the (commented out) DNS black list checks.
4321 (2) Add control=submission to the relay_from_hosts and authenticated
4322 clients checks, on the grounds that messages accepted by these
4323 statements are most likely to be submissions.
4325 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4327 (1) Generate an error if the third argument for the ${prvs expansion is
4330 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4333 (3) Reset the variables that are obtained from the first argument of
4334 ${prvscheck and used in the second argument before leaving the code,
4335 because their memory is reclaimed, so using them afterwards may do
4338 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4339 one (it's much easier than Tom thought :-).
4341 (5) Because of (4), we can now allow for the use of $prvscheck_result
4342 inside the third argument.
4344 PH/03 For some reason, the default setting of PATH when running a command from
4345 a pipe transport was just "/usr/bin". I have changed it to
4348 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4349 anything to be listed in the output from -bV.
4351 PH/05 When a filter generated an autoreply, the entire To: header line was
4352 quoted in the delivery log line, like this:
4354 => >A.N.Other <ano@some.domain> <original@ddress> ...
4356 This has been changed so that it extracts the operative address. There
4357 may be more than one such address. If so, they are comma-separated, like
4360 => >ano@some.domain,ona@other.domain <original@ddress> ...
4362 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4363 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4364 not being looked up in the rDNS to get a host name, Exim was showing the
4365 IP address twice in Received: lines, even though the IP addresses were
4366 identical. For example:
4368 Received: from [1.2.3.4] (helo=[1.2.3.4])
4370 However, if the real host name was known, it was omitting the HELO data
4371 if it matched the actual IP address. This has been tidied up so that it
4372 doesn't show the same IP address twice.
4374 PH/07 When both +timestamp and +memory debugging was on, the value given by
4375 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4376 called by the debug printing, thereby overwriting the timestamp buffer.
4377 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4379 PH/08 When the original message was included in an autoreply transport, it
4380 always said "this is a copy of the message, including all the headers",
4381 even if body_only or headers_only was set. It now gives an appropriate
4384 PH/09 Applied a patch from the Sieve maintainer which:
4386 o fixes some comments
4387 o adds the (disabled) notify extension core
4388 o adds some debug output for the result of if/elsif tests
4389 o points to the current vacation draft in the documentation
4390 and documents the missing references header update
4394 o fixes a bug in processing the envelope test (when testing
4395 multiple envelope elements, the last element determined the
4398 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4399 Electronic Mail") by including:
4401 Auto-submitted: auto-generated
4403 in the messages that it generates (bounce messages and others, such as
4404 warnings). In the case of bounce messages for non-SMTP messages, there was
4405 also a typo: it was using "Auto_submitted" (underscore instead of
4406 hyphen). Since every message generated by Exim is necessarily in response
4407 to another message, thes have all been changed to:
4409 Auto-Submitted: auto-replied
4411 in accordance with these statements in the RFC:
4413 The auto-replied keyword:
4415 - SHOULD be used on messages sent in direct response to another
4416 message by an automatic process,
4418 - MUST NOT be used on manually-generated messages,
4420 - MAY be used on Delivery Status Notifications (DSNs) and Message
4421 Disposition Notifications (MDNs),
4423 - MUST NOT be used on messages generated by automatic or periodic
4424 processes, except for messages which are automatic responses to
4427 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4428 to the default Received: header definition.
4430 PH/12 Added log selector acl_warn_skipped (default on).
4432 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4433 variables because (a) they are in the wrong storage pool and (b) even if
4434 they were copied, it wouldn't work properly because of the caching.
4436 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4437 checking when decoding. Apparently there are clients that generate
4438 overlong encoded strings. Why am I not surprised?
4440 PH/15 If the first argument of "${if match_address" was not empty, but did not
4441 contain an "@" character, Exim crashed. Now it writes a panic log message
4442 and treats the condition as false.
4444 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4446 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4447 extension "envelope-auth". The code is finished and in agreement with
4448 other implementations, but there is no documentation so far and in fact,
4449 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4450 not changing the active code.
4452 Print executed "if" and "elsif" statements when debugging is used. This
4453 helps a great deal to understand what a filter does.
4455 Document more things not specified clearly in RFC3028. I had all this
4456 sorted out, when out of a sudden new issues came to my mind. Oops."
4458 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4461 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4462 canonical form (as documented). However, after a host name lookup from
4463 the IP address, check_host() was doing a simple string comparison with
4464 addresses acquired from the DNS when checking that the found name did
4465 have the original IP as one of its addresses. Since any found IPv6
4466 addresses are likely to be in abbreviated form, the comparison could
4467 fail. Luckily, there already exists a function for doing the comparison
4468 by converting both addresses to binary, so now that is used instead of
4469 the text comparison.
4471 PH/20 There was another similar case to PH/19, when a complete host name was
4472 given in a host list; looking up its IP address could give an abbreviated
4473 form, whereas the current host's name might or might not be abbreviated.
4474 The same fix has been applied.
4480 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4481 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4484 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4485 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4487 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4489 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4490 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4491 Also, for "Auto-Submitted", treat anything other than "no" as
4492 non-personal, in accordance with RFC 3834. (Previously it treated
4493 anything starting "auto-" as non-personal.)
4495 TF/01 The control=submission/name=... option had a problem with syntax
4496 errors if the name included a slash character. The /name= option
4497 now slurps the rest of the string, so it can include any characters
4498 but it must come last in the list of options (after /sender_retain
4501 PH/05 Some modifications to the interface to the fake nameserver for the new
4509 TK/01 Added the "success_on_redirect" address verification option. See
4510 NewStuff for rationale and an example.
4512 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4514 PH/02 Patch to exigrep to allow it to work on syslog lines.
4516 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4517 fread() to skip over the body file's header line, because in Cygwin the
4518 header line is locked and is inaccessible.
4520 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4521 co-exist for some time) to make it clear that it is the Exim ID that is
4522 referenced, not the Message-ID: header line.
4524 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4525 string_format() function, because snprintf() does not exist on all
4528 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4529 ${stat: expansion item.
4531 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4532 protocol synchronization error", to keep the pedants happy.
4534 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4535 well as for IRIX systems, when gcc is being used. See the host.c source
4538 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4540 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4543 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4544 transport and to the smtp transport in LMTP mode.
4546 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4548 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4549 run a filter in a subprocess. This could lead to confusion in subsequent
4550 lookups in the parent process. There should also be a search_tidyup() at
4551 the end of the subprocess.
4553 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4554 only if the host matched helo_try_verify_hosts, which caused the
4555 verification to occur when the EHLO/HELO command was issued. The ACL just
4556 tested the remembered result. Now, if a previous verification attempt has
4557 not happened, "verify = helo" does it there and then.
4559 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4561 TK/03 Fix log output including CR from clamd.
4563 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4564 exist provoked a memory error which could cause a segfault.
4566 PH/15 Installed PCRE 6.2
4568 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4570 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4571 of the problem. Specifically, suggested +O2 rather than +O1 for the
4574 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4576 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4577 record, the port was not correctly logged when the outgoing_port log
4578 selector was set (it logged the transort's default port).
4580 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4581 fallback_hosts, and "hosts" in the smtp transport.
4583 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4584 host errors such as "Connection refused".
4586 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4587 authentication with radiusclient 0.4.9:
4589 - Error returned from rc_read_config was caught wrongly
4590 - Username/password not passed on to radius server due to wrong length.
4592 The presumption is that some radiusclient API changes for 4.51/PH/17
4593 were not taken care of correctly. The code is still untested by me (my
4594 Linux distribution still has 0.3.2 of radiusclient), but it was
4595 contributed by a Radius user.
4597 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4598 expanding the "port" option of the smtp transport.
4600 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4601 while reading a MIME header. Thanks to Tom Hughes for a patch.
4603 PH/24 Include config.h inside local_scan.h so that configuration settings are
4606 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4607 that in an ACL for RCPT (for example), you can examine exactly what was
4610 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4611 commands, but it was not correctly comparing the address with the actual
4612 client host address. Thus, it would show the EHLO address in Received:
4613 header lines when this was not necessary.
4615 PH/27 Added the % operator to ${eval:}.
4617 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4618 it should be ignoring failures (because with -C, for example, it has lost
4619 privilege). It wasn't ignoring creation failures other than "already
4622 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4625 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4626 because an input line was too long, either on its own, or by virtue of
4627 too many continuations, the temporary file was not being removed, and the
4628 return code was incorrect.
4630 PH/31 Missing "BOOL" in function definition in filtertest.c.
4632 PH/32 Applied Sieve patches from the maintainer.
4634 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4636 PH/33 Added "verify = not_blind".
4638 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4639 Local/Makefile (with some defaults set). These are used in built scripts
4640 such as exicyclog, but they have never been used in the exim_install
4641 script (though there are many overriding facilities there). I have
4642 arranged that the exim_install script now takes note of these two
4645 PH/35 Installed configuration files for Dragonfly.
4647 PH/36 When a locally submitted message by a trusted user did not contain a
4648 From: header, and the sender address was obtained from -f or from an SMTP
4649 MAIL command, and the trusted user did not use -F to supply a sender
4650 name, $originator_name was incorrectly used when constructing a From:
4651 header. Furthermore, $originator_name was used for submission mode
4652 messages from external hosts without From: headers in a similar way,
4653 which is clearly wrong.
4655 PH/37 Added control=suppress_local_fixups.
4657 PH/38 When log_selector = +received_sender was set, and the addition of the
4658 sender made the log line's construction buffer exactly full, or one byte
4659 less than full, an overflow happened when the terminating "\n" was
4662 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4663 when the result of a list match is failure because a DNS lookup failed.
4665 PH/40 RM_COMMAND is now used in the building process.
4667 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4668 the "build-* directories that it finds.
4670 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4671 address in a domain literal was a prefix of an interface address.
4673 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4674 when verifying a sender address, unless rewrite_headers is false.
4676 PH/44 Wrote a long comment about why errors_to addresses are verified as
4677 recipients, not senders.
4679 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4680 the ratelimit ACL was added.
4682 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4684 PH/46 Added extra information about PostgreSQL errors to the error string.
4686 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4687 avoiding the need to install special zones in a real server. This is
4688 backwards compatible; if it can't find the fake resolver, it drops back.
4689 Thus, both old and new test suites can be run.
4691 TF/02 Added util/ratelimit.pl
4693 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4694 clock is set back in time.
4696 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4697 Candler <B.Candler@pobox.com>.
4699 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4700 OK for addresses that are the result of redirections.
4702 PH/48 A number of further additions for the benefit of the new test suite,
4703 including a fake gethostbyname() that interfaces to the fake DNS resolver
4706 TF/06 The fix for widen_domains has also been applied to qualify_single and
4707 search_parents which are the other dnslookup options that can cause
4710 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4713 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4715 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4716 service" if it encountered an MX record with an empty target hostname.
4717 The message is now "an MX or SRV record indicated no SMTP service".
4719 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4720 if the DNS of the sending site is misconfigured. This is quite a
4721 common situation. This change restores the behaviour of treating a
4722 helo verification defer as a failure.
4724 PH/51 If self=fail was set on a router, the bounce message did not include the
4725 actual error message.
4731 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4733 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4734 command itself ended in error, the underlying message about the transport
4735 filter timeout was being overwritten with the pipe command error. Now the
4736 underlying error message should be appended to the second error message.
4738 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4740 PH/02 Reduce the amount of output that "make" produces by default. Full output
4741 can still be requested.
4743 PH/03 The warning log line about a condition test deferring for a "warn" verb
4744 was being output only once per connection, rather than after each
4745 occurrence (because it was using the same function as for successful
4746 "warn" verbs). This seems wrong, so I have changed it.
4748 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4749 it should not have, which might have caused a crash in the right
4750 circumstances, but probably never did.
4752 PH/04 Installed a modified version of Tony Finch's patch to make submission
4753 mode fix the return path as well as the Sender: header line, and to
4754 add a /name= option so that you can make the user's friendly name appear
4757 TF/03 Added the control = fakedefer ACL modifier.
4759 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4760 Mark Lowes for thorough testing.
4762 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4764 TK/03 Merged latest SRS patch from Miles Wilton.
4766 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4767 with the definition in sysexits.h (which is #included earlier).
4768 Fortunately, Exim does not actually use EX_OK. The code used to try to
4769 preserve the sysexits.h value, by assuming that macro definitions were
4770 scanned for macro replacements. I have been disabused of this notion,
4771 so now the code just undefines EX_OK before #including unistd.h.
4773 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4774 in the smtp transport. When a block could not be written in a single
4775 write() function, the timeout was being re-applied to each part-write.
4776 This seems wrong - if the receiver was accepting one byte at a time it
4777 would take for ever. The timeout is now adjusted when this happens. It
4778 doesn't have to be particularly precise.
4780 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4781 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4783 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4784 without a domain if the check to <postmaster@domain> fails.
4786 SC/01 Eximstats: added -xls and the ability to specify output files
4787 (patch written by Frank Heydlauf).
4789 SC/02 Eximstats: use FileHandles for outputting results.
4791 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4793 SC/04 Eximstats: fixed display of large numbers with -nvr option
4795 SC/05 Eximstats: fixed merging of reports with empty tables.
4797 SC/06 Eximstats: added the -include_original_destination flag
4799 SC/07 Eximstats: removed tabs and trailing whitespace.
4801 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4803 TK/06 MBOX spool code: Add real "From " MBOX separator line
4804 so the .eml file is really in mbox format (even though
4805 most programs do not really care). Patch from Alex Miller.
4807 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4808 The latter is generated from $received_to and is only set if the
4809 message has one envelope recipient. SA can use these headers,
4810 obviously out-of-the-box. Patch from Alex Miller.
4812 PH/08 The ${def test on a variable was returning false if the variable's
4813 value was "0", contrary to what the specification has always said!
4814 The result should be true unless the variable is empty.
4816 PH/09 The syntax error of a character other than { following "${if
4817 def:variable_name" (after optional whitespace) was not being diagnosed.
4818 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4819 accidental colon was present, for example, could give incorrect results.
4821 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4822 result is used (not including appendfile, where other changes are about
4825 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4826 This involved changing a lot of size variables from int to off_t. It
4827 should work with maildirs and everything.
4829 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4830 spamd dying while we are connected to it.
4832 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4835 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4836 function for BDB 4.3.
4838 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4840 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4841 the sources for PCRE have been re-arranged and more files are now
4844 PH/15 The code I had for printing potentially long long variables in PH/11
4845 above was not the best (it lost precision). The length of off_t variables
4846 is now inspected at build time, and an appropriate printing format (%ld
4847 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4848 to be "long long int" or "long int". This is needed for the internal
4849 formatting function string_vformat().
4851 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4852 the configuration file to be ":syslog", then the script "guesses" where
4853 the logs files are, rather than using the compiled in default. In our
4854 case the guess is not the same as the compiled default, so the script
4855 suddenly stopped working when I started to use syslog. The patch checks
4856 to see if log_file_path is "". If so, it attempts to read it from exim
4857 with no configuration file to get the compiled in version, before it
4858 falls back to the previous guessing code."
4860 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4861 implementing BATV in an Exim configuration. See NewStuff for the gory
4864 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4865 Makefile that are specific to HP-UX.
4867 PH/18 If the "use_postmaster" option was set for a recipient callout together
4868 with the "random" option, the postmaster address was used as the MAIL
4869 FROM address for the random test, but not for the subsequent recipient
4870 test. It is now used for both.
4872 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4873 patch removes a few documentation additions to RFC 3028, because the
4874 latest draft now contains them. It adds the new en;ascii-case comparator
4875 and a new error check for 8bit text in MIME parts. Comparator and
4876 require names are now matched exactly. I enabled the subaddress
4877 extension, but it is not well tested yet (read: it works for me)."
4879 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4880 rework some of the code of TK/09 above to avoid the hardwired use of
4881 "%lld" and "long long". Replaced the call to snprintf() with a call to
4884 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4885 records point to non-existent hosts", "retry timeout exceeded", and
4886 "retry time not reached for any host after a long failure period".
4888 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4889 experimental DomainKeys support:
4891 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4892 (2) On an error such as an illegally used "control", the wrong name for
4893 the control was given.
4895 These problems did NOT occur unless DomainKeys support was compiled.
4897 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4899 PH/24 Added ${if match_ip condition.
4901 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4902 databases so that it will be absolutely obvious if a crash occurs in the
4903 DB library. This is a regular occurrence (often caused by mis-matched
4906 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4907 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4908 picked up on a user's system that detects such things. There doesn't seem
4909 to be a gcc warning option for this - only an attribute that has to be
4910 put on the function's prototype. It seems that in Fedora Core 4 they have
4911 set this on a number of new functions. No doubt there will be more in due
4914 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4915 specify a transport. However, if an address that was verified by such a
4916 router was the subject of a callout, Exim crashed because it tried to
4917 read the rcpt_include_affixes from the non-existent transport. Now it
4918 just assumes that the setting of that option is false. This bug was
4919 introduced by 4.51/PH/31.
4921 PH/28 Changed -d+all to exclude +memory, because that information is very
4922 rarely of interest, but it makes the output a lot bigger. People tend to
4923 do -d+all out of habit.
4925 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4926 code in os-type was giving problems when libc.so lives in lib64, like on
4929 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4930 aren't the modern standard, and it seems that some systems' include files
4931 don't always have them. Exim was already checking for some of the newer
4932 ones like T_AAAA, and defining it itself. I've added checks for all the
4933 record types that Exim uses.
4935 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4936 not automatically generating a new one, as it is supposed to. This
4937 prevented TLS from working. If the file did exist, but contained invalid
4938 data, a new version was generated, as expected. It was only the case of a
4939 non-existent file that was broken.
4941 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4942 with a change in libdomainkeys > 0.64.
4944 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4945 from DNS. If the selector record carries the flag, it now has
4946 precedence over the domain-wide flag.
4948 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4950 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4951 the use of a port name that isn't defined in /etc/services) occurred, the
4952 message was deferred as in a normal delivery, and thus remained on the
4953 spool, instead of being failed because of the mua_wrapper setting. This
4954 is now fixed, and I tidied up some of the mua_wrapper messages at the
4957 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4958 the messages in a hash of arrays rather than using individual hashes.
4959 This is a bit cleaner and results in dramatic memory savings, albeit
4960 at a slight CPU cost.
4962 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4963 as requested by Marc Sherman.
4965 SC/10 Eximstats: added histograms for user specified patterns as requested
4968 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4970 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4971 fopen() in the content-scanning modules that did not already have it.
4977 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4978 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4980 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4982 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4984 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4985 to test Sieve filters that use "vacation".
4987 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4988 that changes the way the GnuTLS parameters are stored in the cache file.
4989 The new format can be generated externally. For backward compatibility,
4990 if the data in the cache doesn't make sense, Exim assumes it has read an
4991 old-format file, and it generates new data and writes a new file. This
4992 means that you can't go back to an older release without removing the
4995 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4996 work if there are any delivery delays because "one_time" forces the
4997 parent to be marked "delivered", so its unseen clone is never tried
4998 again. For this reason, Exim now forbids the simultaneous setting of
5001 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5002 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5003 entry if you want to know the details. The fix, however, neglected to
5004 consider the case when local delivery batching is involved. The test for
5005 "previously delivered" was not happening when checking to see if an
5006 address could be batched with a previous (undelivered) one; under
5007 certain circumstances this could lead to multiple deliveries to the same
5010 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5011 in its include files, and this causes problems building Exim.
5013 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5014 header_syntax) but Exim was just ignoring anything given after a slash.
5015 In particular, this caused confusion with an attempt to use "verify =
5016 reverse_host_lookup/defer_ok". An error is now given when options are
5017 supplied for verify items that do not have them. (Maybe reverse_host_
5018 lookup should have a defer_ok option, but that's a different point.)
5020 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5021 defined by RFC 821) to 2048, because there were problems with some AUTH
5022 commands, and RFC 1869 says the size should be increased for extended
5023 SMTP commands that take arguments.
5025 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5028 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5029 "unknown" error; now it says that the functionality isn't in the binary.
5031 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5032 an address' error message when a string expansion fails (syntax or
5033 whatever). Otherwise the password may appear in the log. Following change
5034 PH/42 below, there is no longer a chance of it appearing in a bounce
5037 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5039 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5040 of its addresses was ever tried. (Bugzilla bug #2.)
5042 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5043 the result incorrectly in the debug output. (It correctly added a newline
5044 to what was transported.)
5046 TF/01 Added $received_time.
5048 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5049 commented out examples of how to interface to a virus scanner and to
5050 SpamAssassin. Also added commented examples of av_scanner and
5051 spamd_address settings.
5053 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5054 and controls are allowed in which ACLs. There were a couple of minor
5055 errors. Some of the entries in the conditions table (which is a table of
5056 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5057 negation of where the condition IS allowed.
5059 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5061 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5062 header file does not have a version number, so I've had to invent a new
5063 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5064 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5065 radiusclient), but it was contributed by a Radius user.
5067 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5068 files or directories, for OpenSSL.
5070 PH/19 When an Exim process that is running as root has to create an Exim log
5071 file, it does so in a subprocess that runs as exim:exim so as to get the
5072 ownership right at creation (otherwise, other Exim processes might see
5073 the file with the wrong ownership). There was no test for failure of this
5074 fork() call, which would lead to the process getting stuck as it waited
5075 for a non-existent subprocess. Forks do occasionally fail when resources
5076 run out. I reviewed all the other calls to fork(); they all seem to check
5079 PH/20 When checking for unexpected SMTP input at connect time (before writing
5080 the banner), Exim was not dealing correctly with a non-positive return
5081 from the read() function. If the client had disconnected by this time,
5082 the result was a log entry for a synchronization error with an empty
5083 string after "input=" when read() returned zero. If read() returned -1
5084 (an event I could not check), uninitialized data bytes were printed.
5085 There were reports of junk text (parts of files, etc) appearing after
5088 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5090 PH/22 Added support for macro redefinition, and (re)definition in between
5091 driver and ACL definitions.
5093 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5094 forgetting to use the resulting value; it was using the unexpanded value.
5096 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5097 hadn't been configured. The fix is from Juergen Kreileder, who
5098 understands it better than I do:
5100 "Here's what I see happening with three configured cyrus_sasl
5101 authenticators configured (plain, login, cram-md5):
5103 On startup auth_cyrus_sasl_init() gets called for each of these.
5104 This means three calls to sasl_listmech() without a specified mech_list.
5105 => SASL tests which mechs of all available mechs actually work
5106 => three warnings about OTP not working
5107 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5109 With the patch, sasl_listmech() also gets called three times. But now
5110 SASL's mech_list option is set to the server_mech specified in the the
5111 authenticator. Or in other words, the answer from sasl_listmech()
5112 gets limited to just the mech you're testing for (which is different
5114 => the return list contains just 'plain' or 'login', 'cram-md5' or
5115 nothing depending on the value of ob->server_mech.
5117 I've just tested the patch: Authentication still works fine,
5118 unavailable mechs specified in the exim configuration are still
5119 caught, and the auth.log warnings about OTP are gone."
5121 PH/25 When debugging is enabled, the contents of the command line are added
5122 to the debugging output, even when log_selector=+arguments is not
5125 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5126 answer is "GNU", and only if the return is "GNU/something" is the answer
5129 PH/27 $acl_verify_message is now set immediately after the failure of a
5130 verification in an ACL, and so is available in subsequent modifiers. In
5131 particular, the message can be preserved by coding like this:
5133 warn !verify = sender
5134 set acl_m0 = $acl_verify_message
5136 Previously, $acl_verify_message was set only while expanding "message"
5137 and "log_message" when a very denied access.
5139 PH/28 Modified OS/os.c-Linux with
5141 -#ifndef OS_LOAD_AVERAGE
5142 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5144 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5145 nomenclature these days.)
5147 PH/29 Installed patch from the Sieve maintainer that adds the options
5148 sieve_useraddress and sieve_subaddress to the redirect router.
5150 PH/30 In these circumstances:
5151 . Two addresses routed to the same list of hosts;
5152 . First host does not offer TLS;
5153 . First host accepts first address;
5154 . First host gives temporary error to second address;
5155 . Second host offers TLS and a TLS session is established;
5156 . Second host accepts second address.
5157 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5158 and peerdn, if requested) that were in fact used only for the second
5161 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5162 attention to any local part prefix or suffix that was matched by the
5163 router that accepted the address. It now behaves in the same way as it
5164 does for delivery: the affixes are removed from the local part unless
5165 rcpt_include_affixes is set on the transport.
5167 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5168 timeout during the DATA phase of an incoming message.
5170 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5171 applied a patch sanctioned by the Sieve maintainer.
5173 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5174 the uid or gid is negative. A case of a negative gid caused this to be
5175 noticed. The fix allows for either to be negative.
5177 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5178 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5181 PH/36 The change PH/12 above was broken. Fixed it.
5183 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5184 the grounds that routing the same address twice would always produce the
5185 same answer. This might have been true once, but it is certainly no
5186 longer true now. Routing a child address may depend on the previous
5187 routing that produced that child. Some complicated redirection strategies
5188 went wrong when messages had multiple recipients, and made Exim's
5189 behaviour dependent on the order in which the addresses were given.
5191 I have moved the duplicate checking until after the routing is complete.
5192 Exim scans the addresses that are assigned to local and remote
5193 transports, and removes any duplicates. This means that more work will be
5194 done, as duplicates will always all be routed, but duplicates are
5195 presumably rare, so I don't expect this is of any significance.
5197 For deliveries to pipes, files, and autoreplies, the duplicate checking
5198 still happens during the routing process, since they are not going to be
5201 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5202 It corrects a timeout issue with spamd. This is Ian's comment: "The
5203 background is that sometimes spamd either never reads data from a
5204 connection it has accepted, or it never writes response data. The exiscan
5205 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5206 blindly assumes that writes won't block so it may never time out."
5208 PH/39 Allow G after quota size as well as K and M.
5210 PH/40 The value set for $authenticated_id in an authenticator may not contain
5211 binary zeroes or newlines because the value is written to log lines and
5212 to spool files. There was no check on this. Now the value is run through
5213 the string_printing() function so that such characters are converted to
5214 printable escape sequences.
5216 PH/41 $message_linecount is a new variable that contains the total number of
5217 lines in the message. Compare $body_linecount, which is the count for the
5220 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5221 bounce and delay warning messages, except in certain special cases, which
5224 (a) An SMTP error message from a remote host;
5225 (b) A message specified in a :fail: redirection;
5226 (c) A message specified in a "fail" command in a system filter;
5227 (d) A message specified in a FAIL return from the queryprogram router;
5228 (e) A message specified by the cannot_route_message router option.
5230 In these cases only, Exim does include the error details in bounce and
5231 warning messages. There are also a few cases where bland messages such
5232 as "unrouteable address" or "local delivery error" are given.
5234 PH/43 $value is now also set for the "else" part of a ${run expansion.
5236 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5237 being worked on, but at least Exim now implements the latest version to
5240 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5241 process to complete was treated as a delivery failure, a timeout while
5242 writing the message to the pipe was logged, but erroneously treated as a
5243 successful delivery. Such timeouts include transport filter timeouts. For
5244 consistency with the overall process timeout, these timeouts are now
5245 treated as errors, giving rise to delivery failures by default. However,
5246 there is now a new Boolean option for the pipe transport called
5247 timeout_defer, which, if set TRUE, converts the failures into defers for
5248 both kinds of timeout. A transport filter timeout is now identified in
5251 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5252 systems where "make" and "gmake" are different, calling "gmake" at top
5253 level broke things. I've arranged for the value of $(MAKE) to be passed
5254 from the Makefile to this script so that it can call the same version of
5258 A note about Exim versions 4.44 and 4.50
5259 ----------------------------------------
5261 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5262 changes of various kinds. As a consequence, a big documentation update was
5263 needed. This delayed the release for rather longer than seemed good, especially
5264 in the light of a couple of (minor) security issues. Therefore, the changes
5265 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5266 release. So 4.44 and 4.50 are in effect two different branches that both start
5269 I have left the 4.50 change log unchanged; it contains all the changes since
5270 4.43. The change log for 4.44 is below; many of its items are identical to
5271 those for 4.50. This seems to be the most sensible way to preserve the
5272 historical information.
5278 1. Minor wording change to the doc/README.SIEVE file.
5280 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5281 computation of the current number of files was incorrect.
5283 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5284 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5287 4. Give more explanation in the error message when the command for a transport
5288 filter fails to execute.
5290 5. There are several places where Exim runs a non-Exim command in a
5291 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5292 was being done only for the command run by the queryprogram router. It is
5293 now done for all such subprocesses. The other cases are: ${run, transport
5294 filters, and the commands run by the lmtp and pipe transports.
5296 6. Added CONFIGURE_GROUP build-time option.
5298 7. Some older OS have a limit of 256 on the maximum number of file
5299 descriptors. Exim was using setrlimit() to set 1000 as a large value
5300 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5301 systems. I've change it so that if it can't get 1000, it tries for 256.
5303 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5304 was an oversight, and furthermore, ever since the addition of extra
5305 controls (e.g. 4.43/32), the checks on when to allow different forms of
5306 "control" were broken. There should now be diagnostics for all cases when a
5307 control that does not make sense is encountered.
5309 9. Added the /retain_sender option to "control=submission".
5311 10. $recipients is now available in the predata ACL (oversight).
5313 11. Tidy the search cache before the fork to do a delivery from a message
5314 received from the command line. Otherwise the child will trigger a lookup
5315 failure and thereby defer the delivery if it tries to use (for example) a
5316 cached ldap connection that the parent has called unbind on.
5318 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5319 of $address_data from the recipient verification was clobbered by the
5320 sender verification.
5322 13. The value of address_data from a sender verification is now available in
5323 $sender_address_data in subsequent conditions in the ACL statement.
5325 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5327 15. Added a new option "connect=<time>" to callout options, to set a different
5330 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5331 was its contents. (It was OK if the option was not defined at all.)
5333 17. A "Completed" log line is now written for messages that are removed from
5334 the spool by the -Mrm option.
5336 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5337 information about exactly what failed.
5339 19. Added -dd to debug only the daemon process.
5341 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5342 handles timeouts, both on the server side and network timeouts. Renamed the
5343 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5345 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5346 It is now set to "smtps".
5348 22. $host_address is now set to the target address during the checking of
5349 ignore_target_hosts.
5351 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5352 being passed; this would have caused $sender_host_name to have been used if
5353 matching the list had actually called for a host name (not very likely,
5354 since this list is usually IP addresses). A host name is now passed as
5357 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5358 code that specifies a non-restarting handler (typically sigaction() in
5359 modern systems) in an attempt to fix a rare and obscure crash bug.
5361 25. Narrowed the window for a race in the daemon that could cause it to ignore
5362 SIGCHLD signals. This is not a major problem, because they are used only to
5363 wake it up if nothing else does.
5365 26. A malformed maildirsize file could cause Exim to calculate negative values
5366 for the mailbox size or file count. Odd effects could occur as a result.
5367 The maildirsize information is now recalculated if the size or filecount
5370 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5371 support for a long time. Removed HAVE_SYS_VFS_H.
5373 28. Installed the latest version of exipick from John Jetmore.
5375 29. In an address list, if the pattern was not a regular expression, an empty
5376 subject address (from a bounce message) matched only if the pattern was an
5377 empty string. Non-empty patterns were not even tested. This was the wrong
5378 because it is perfectly reasonable to use an empty address as part of a
5379 database query. An empty address is now tested by patterns that are
5380 lookups. However, all the other forms of pattern expect the subject to
5381 contain a local part and a domain, and therefore, for them, an empty
5382 address still always fails if the pattern is not itself empty.
5384 30. Exim went into a mad DNS loop when attempting to do a callout where the
5385 host was specified on an smtp transport, and looking it up yielded more
5386 than one IP address.
5388 31. Re-factored the code for checking spool and log partition space into a
5389 function that finds that data and another that does the check. The former
5390 is then used to implement four new variables: $spool_space, $log_space,
5391 $spool_inodes, and $log_inodes.
5393 32. The RFC2047 encoding function was originally intended for short strings
5394 such as real names; it was not keeping to the 75-character limit for
5395 encoded words that the RFC imposes. It now respects the limit, and
5396 generates multiple encoded words if necessary. To be on the safe side, I
5397 have increased the buffer size for the ${rfc2047: expansion operator from
5400 33. It is now permitted to omit both strings after an "if" condition; if the
5401 condition is true, the result is "true". As before, when the second string
5402 is omitted, a false condition yields an empty string. This makes it less
5403 cumbersome to write custom ACL and router conditions.
5405 34. 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 35. 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 36. The table in the code that translates DNS record types into text (T_A to
5414 "A" for instance) was missing entries for NS and CNAME. It is just possible
5415 that this could have caused confusion if both these types were looked up
5416 for the same domain, because the text type is used as part of Exim's
5417 per-process caching. But the chance of anyone hitting this buglet seems
5420 37. The dnsdb lookup has been extended in a number of ways.
5422 (1) There is a new type, "zns", which walks up the domain tree until it
5423 finds some nameserver records. It should be used with care.
5425 (2) There is a new type, "mxh", which is like "mx" except that it returns
5426 just the host names, not the priorities.
5428 (3) It is now possible to give a list of domains (or IP addresses) to be
5429 looked up. The behaviour when one of the lookups defers can be
5430 controlled by a keyword.
5432 (4) It is now possible to specify the separator character for use when
5433 multiple records are returned.
5435 38. The dnslists ACL condition has been extended: it is now possible to supply
5436 a list of IP addresses and/or domains to be looked up in a particular DNS
5439 39. Added log_selector=+queue_time_overall.
5441 40. When running the queue in the test harness, wait just a tad after forking a
5442 delivery process, to get repeatability of debugging output.
5444 41. Include certificate and key file names in error message when GnuTLS fails
5445 to set them up, because the GnuTLS error message doesn't include the name
5446 of the failing file when there is a problem reading it.
5448 42. Allow both -bf and -bF in the same test run.
5450 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5452 44. The "Exiscan patch" is now merged into the mainline Exim source.
5454 45. Sometimes the final signoff response after QUIT could fail to get
5455 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5456 < 0 before doing a fflush(). This bug looks as though it goes back to the
5457 introduction of TLS in release 3.20, but "sometimes" must have been rare
5458 because the tests only now provoked it.
5460 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5461 (this can affect the format of dates).
5463 47. exim_tidydb, when checking for the continued existence of a message for
5464 which it has found a message-specific retry record, was not finding
5465 messages that were in split spool directories. Consequently, it was
5466 deleting retry records that should have stayed in existence.
5468 48. Steve fixed some bugs in eximstats.
5470 49. The SPA authentication driver was not abandoning authentication and moving
5471 on to the next authenticator when an expansion was forced to fail,
5472 contradicting the general specification for all authenticators. Instead it
5473 was generating a temporary error. It now behaves as specified.
5475 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5476 (the order specifies the preference for clients). The order is now AES256,
5477 AES128, 3DES, ARCFOUR128.
5479 51. Small patch to Sieve code - explicitly set From: when generating an
5482 52. Exim crashed if a remote delivery caused a very long error message to be
5483 recorded - for instance if somebody sent an entire SpamAssassin report back
5484 as a large number of 550 error lines. This bug was coincidentally fixed by
5485 increasing the size of one of Exim's internal buffers (big_buffer) that
5486 happened as part of the Exiscan merge. However, to be on the safe side, I
5487 have made the code more robust (and fixed the comments that describe what
5490 53. Now that there can be additional text after "Completed" in log lines (if
5491 the queue_time_overall log selector is set), a one-byte patch to exigrep
5492 was needed to allow it to recognize "Completed" as not the last thing in
5495 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5496 patch that reportedly fixes this has been added. I am not expert enough to
5497 create a test for it. This is what the patch creator wrote:
5499 "I found a little strange behaviour of ldap code when working with
5500 Windows 2003 AD Domain, where users was placed in more than one
5501 Organization Units. When I tried to give exim partial DN, the exit code
5502 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5503 But simultaneously result of request was absolutely normal ldap result,
5504 so I produce this patch..."
5506 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5507 REFERENCE, so I have modified the code to exclude the patch when that macro
5510 55. Some experimental protocols are using DNS PTR records for new purposes. The
5511 keys for these records are domain names, not reversed IP addresses. The
5512 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5513 leaves it alone. Component reversal etc. now happens only for IP addresses.
5516 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5518 57. Double the size of the debug message buffer (to 2048) so that more of very
5519 long debug lines gets shown.
5521 58. The exicyclog utility now does better if the number of log files to keep
5522 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5524 59. Two changes related to the smtp_active_hostname option:
5526 (1) $smtp_active_hostname is now available as a variable.
5527 (2) The default for smtp_banner uses $smtp_active_hostname instead
5528 of $primary_hostname.
5530 60. The host_aton() function is supposed to be passed a string that is known
5531 to be a valid IP address. However, in the case of IPv6 addresses, it was
5532 not checking this. This is a hostage to fortune. Exim now panics and dies
5533 if the condition is not met. A case was found where this could be provoked
5534 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5535 components; fortuitously, this particular loophole had already been fixed
5536 by change 4.50/55 above.
5538 If there are any other similar loopholes, the new check in host_aton()
5539 itself should stop them being exploited. The report I received stated that
5540 data on the command line could provoke the exploit when Exim was running as
5541 exim, but did not say which command line option was involved. All I could
5542 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5543 running as the user.
5546 61. There was a buffer overflow vulnerability in the SPA authentication code
5547 (which came originally from the Samba project). I have added a test to the
5548 spa_base64_to_bits() function which I hope fixes it.
5551 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5552 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5554 63. The daemon start-up calls getloadavg() while still root for those OS that
5555 need the first call to be done as root, but it missed one case: when
5556 deliver_queue_load_max is set with deliver_drop_privilege. This is
5557 necessary for the benefit of the queue runner, because there is no re-exec
5558 when deliver_drop_privilege is set.
5560 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5561 This has been fixed.
5563 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5564 was in use, was not putting the data itself into the right store pool;
5565 consequently, it could be overwritten for a subsequent message in the same
5566 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5569 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5571 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5572 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5573 the function were treating the return as a boolean value, which happened to
5574 work because 0=false and not-0=true, but is not correct code.
5576 68. The host_aton() function was not handling scoped IPv6 addresses (those
5577 with, for example, "%eth0" on the end) correctly.
5579 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5580 negated items (that is, ~something) in unsigned ints. Some compilers
5581 apparently mutter when there is no cast.
5583 70. If an address verification called from an ACL failed, and did not produce a
5584 user-specific message (i.e. there was only a "system" message), nothing was
5585 put in $acl_verify_message. In this situation, it now puts the system
5588 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5589 session; change 4.31/43 added the unwanted input to the log line - except
5590 that it did not do this in the start of session case. It now does.
5592 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5593 This is wrong and can cause the other end to generate a synchronization
5594 error if it is another Exim or anything else that does the synchronization
5595 check. A QUIT command is no longer sent after a timeout.
5597 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5598 during host lookups.
5600 74. The defer_ok option of callout verification was not working if it was used
5601 when verifying addresses in header lines, that is, for this case:
5603 verify = header_sender/callout=defer_ok
5605 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5606 those file descriptors could be used for SMTP connections. If anything
5607 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5608 could be sent to the SMTP client, causing chaos. The daemon now opens
5609 stdin, stdout, and stderr to /dev/null when it puts itself into the
5612 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5613 log by default. The user can override this with suitable Perl magic.
5615 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5616 the log message when discard triggers, was not working for the DATA ACL or
5617 for the non-SMTP ACL.
5619 78. Error message wording change in sieve.c.
5621 79. If smtp_accept_max_per_host was set, the number of connections could be
5622 restricted to fewer than expected, because the daemon was trying to set up
5623 a new connection before checking whether the processes handling previous
5624 connections had finished. The check for completed processes is now done
5625 earlier. On busy systems, this bug wouldn't be noticed because something
5626 else would have woken the daemon, and it would have reaped the completed
5629 80. If a message was submitted locally by a user whose login name contained one
5630 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5631 It caused a spool format error. I have fixed the spool reading code. A
5632 related problem was that the "from" clause in the Received: line became
5633 illegal because of the space(s). It is now covered by ${quote_local_part.
5635 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5638 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5640 83. Patch from Sieve maintainer to fix unterminated string problem in
5641 "vacation" handling.
5643 84. Some minor changes to the Linux configuration files to help with other
5644 OS variants using glibc.
5646 85. One more patch for Sieve to update vacation handling to latest spec.
5649 ----------------------------------------------------
5650 See the note above about the 4.44 and 4.50 releases.
5651 ----------------------------------------------------
5657 1. Change 4.43/35 introduced a bug that caused file counts to be
5658 incorrectly computed when quota_filecount was set in an appendfile
5661 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5662 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5665 3. Give more explanation in the error message when the command for a transport
5666 filter fails to execute.
5668 4. There are several places where Exim runs a non-Exim command in a
5669 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5670 was being done only for the command run by the queryprogram router. It is
5671 now done for all such subprocesses. The other cases are: ${run, transport
5672 filters, and the commands run by the lmtp and pipe transports.
5674 5. Some older OS have a limit of 256 on the maximum number of file
5675 descriptors. Exim was using setrlimit() to set 1000 as a large value
5676 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5677 systems. I've change it so that if it can't get 1000, it tries for 256.
5679 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5680 was an oversight, and furthermore, ever since the addition of extra
5681 controls (e.g. 4.43/32), the checks on when to allow different forms of
5682 "control" were broken. There should now be diagnostics for all cases when a
5683 control that does not make sense is encountered.
5685 7. $recipients is now available in the predata ACL (oversight).
5687 8. Tidy the search cache before the fork to do a delivery from a message
5688 received from the command line. Otherwise the child will trigger a lookup
5689 failure and thereby defer the delivery if it tries to use (for example) a
5690 cached ldap connection that the parent has called unbind on.
5692 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5693 of $address_data from the recipient verification was clobbered by the
5694 sender verification.
5696 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5697 was its contents. (It was OK if the option was not defined at all.)
5699 11. A "Completed" log line is now written for messages that are removed from
5700 the spool by the -Mrm option.
5702 12. $host_address is now set to the target address during the checking of
5703 ignore_target_hosts.
5705 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5706 being passed; this would have caused $sender_host_name to have been used if
5707 matching the list had actually called for a host name (not very likely,
5708 since this list is usually IP addresses). A host name is now passed as
5711 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5712 code that specifies a non-restarting handler (typically sigaction() in
5713 modern systems) in an attempt to fix a rare and obscure crash bug.
5715 15. Narrowed the window for a race in the daemon that could cause it to ignore
5716 SIGCHLD signals. This is not a major problem, because they are used only to
5717 wake it up if nothing else does.
5719 16. A malformed maildirsize file could cause Exim to calculate negative values
5720 for the mailbox size or file count. Odd effects could occur as a result.
5721 The maildirsize information is now recalculated if the size or filecount
5724 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5725 support for a long time. Removed HAVE_SYS_VFS_H.
5727 18. Updated exipick to current release from John Jetmore.
5729 19. Allow an empty sender to be matched against a lookup in an address list.
5730 Previously the only cases considered were a regular expression, or an
5733 20. Exim went into a mad DNS lookup loop when doing a callout where the
5734 host was specified on the transport, if the DNS lookup yielded more than
5737 21. The RFC2047 encoding function was originally intended for short strings
5738 such as real names; it was not keeping to the 75-character limit for
5739 encoded words that the RFC imposes. It now respects the limit, and
5740 generates multiple encoded words if necessary. To be on the safe side, I
5741 have increased the buffer size for the ${rfc2047: expansion operator from
5744 22. Failure to deliver a bounce message always caused it to be frozen, even if
5745 there was an errors_to setting on the router. The errors_to setting is now
5748 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5749 canonical form (fully expanded) before being placed in
5750 $sender_host_address.
5752 24. Updated eximstats to version 1.33
5754 25. Include certificate and key file names in error message when GnuTLS fails
5755 to set them up, because the GnuTLS error message doesn't include the name
5756 of the failing file when there is a problem reading it.
5758 26. Expand error message when OpenSSL has problems setting up cert/key files.
5761 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5762 (this can affect the format of dates).
5764 28. exim_tidydb, when checking for the continued existence of a message for
5765 which it has found a message-specific retry record, was not finding
5766 messages that were in split spool directories. Consequently, it was
5767 deleting retry records that should have stayed in existence.
5769 29. eximstats updated to version 1.35
5770 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5771 1.35 - bugfix such that pie charts by volume are generated correctly
5773 30. The SPA authentication driver was not abandoning authentication and moving
5774 on to the next authenticator when an expansion was forced to fail,
5775 contradicting the general specification for all authenticators. Instead it
5776 was generating a temporary error. It now behaves as specified.
5778 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5779 (the order specifies the preference for clients). The order is now AES256,
5780 AES128, 3DES, ARCFOUR128.
5782 31. Small patch to Sieve code - explicitly set From: when generating an
5785 32. Exim crashed if a remote delivery caused a very long error message to be
5786 recorded - for instance if somebody sent an entire SpamAssassin report back
5787 as a large number of 550 error lines. This bug was coincidentally fixed by
5788 increasing the size of one of Exim's internal buffers (big_buffer) that
5789 happened as part of the Exiscan merge. However, to be on the safe side, I
5790 have made the code more robust (and fixed the comments that describe what
5793 33. Some experimental protocols are using DNS PTR records for new purposes. The
5794 keys for these records are domain names, not reversed IP addresses. The
5795 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5796 leaves it alone. Component reversal etc. now happens only for IP addresses.
5799 34. The host_aton() function is supposed to be passed a string that is known
5800 to be a valid IP address. However, in the case of IPv6 addresses, it was
5801 not checking this. This is a hostage to fortune. Exim now panics and dies
5802 if the condition is not met. A case was found where this could be provoked
5803 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5804 components; fortuitously, this particular loophole had already been fixed
5805 by change 4.50/55 or 4.44/33 above.
5807 If there are any other similar loopholes, the new check in host_aton()
5808 itself should stop them being exploited. The report I received stated that
5809 data on the command line could provoke the exploit when Exim was running as
5810 exim, but did not say which command line option was involved. All I could
5811 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5812 running as the user.
5815 35. There was a buffer overflow vulnerability in the SPA authentication code
5816 (which came originally from the Samba project). I have added a test to the
5817 spa_base64_to_bits() function which I hope fixes it.
5820 36. The daemon start-up calls getloadavg() while still root for those OS that
5821 need the first call to be done as root, but it missed one case: when
5822 deliver_queue_load_max is set with deliver_drop_privilege. This is
5823 necessary for the benefit of the queue runner, because there is no re-exec
5824 when deliver_drop_privilege is set.
5826 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5827 was in use, was not putting the data itself into the right store pool;
5828 consequently, it could be overwritten for a subsequent message in the same
5829 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5832 38. Sometimes the final signoff response after QUIT could fail to get
5833 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5834 < 0 before doing a fflush(). This bug looks as though it goes back to the
5835 introduction of TLS in release 3.20, but "sometimes" must have been rare
5836 because the tests only now provoked it.
5842 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5843 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5844 Now it returns an integer. A number of calls were still expecting a T/F
5845 return. Fortuitously, in all cases, the tests worked in OK situations,
5846 which is the norm. However, things would have gone wrong on any write
5847 failures on the smtp file descriptor. This function is used when sending
5848 messages over SMTP and also when doing verify callouts.
5850 2. When Exim is called to do synchronous delivery of a locally submitted
5851 message (the -odf or -odi options), it no longer closes stderr before doing
5854 3. Implemented the mua_wrapper option.
5856 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5858 5. Implemented the functions header_remove(), header_testname(),
5859 header_add_at_position(), and receive_remove_recipient(), and exported them
5862 6. If an ACL "warn" statement specified the addition of headers, Exim already
5863 inserted X-ACL-Warn: at the start if there was no header name. However, it
5864 was not making this test for the second and subsequent header lines if
5865 there were newlines in the string. This meant that an invalid header could
5866 be inserted if Exim was badly configured.
5868 7. Allow an ACL "warn" statement to add header lines at the start or after all
5869 the Received: headers, as well as at the end.
5871 8. Added the rcpt_4xx retry error code.
5873 9. Added postmaster_mailfrom=xxx to callout verification option.
5875 10. Added mailfrom=xxxx to the callout verification option, for verify=
5878 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5879 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5881 12. Inserted some casts to stop certain compilers warning when using pointer
5882 differences as field lengths or precisions in printf-type calls (mostly
5883 affecting debugging statements).
5885 13. Added optional readline() support for -be (dynamically loaded).
5887 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5888 same clock tick as a message's arrival, so that its received time was the
5889 same as the "first fail" time on the retry record, and that message
5890 remained on the queue past the ultimate address timeout, every queue runner
5891 would try a delivery (because it was past the ultimate address timeout) but
5892 after another failure, the ultimate address timeout, which should have then
5893 bounced the address, did not kick in. This was a "< instead of <=" error;
5894 in most cases the first failure would have been in the next clock tick
5895 after the received time, and all would be well.
5897 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5898 being recognized when the domain list was tested by the match_domain
5899 condition in an expansion string.
5901 16. Added the ${str2b64: operator.
5903 17. Exim was always calling setrlimit() to set a large limit for the number of
5904 processes, without checking whether the existing limit was already
5905 adequate. (It did check for the limit on file descriptors.) Furthermore,
5906 errors from getrlimit() and setrlimit() were being ignored. Now they are
5907 logged to the main and panic logs, but Exim does carry on, to try to do its
5908 job under whatever limits there are.
5910 18. Imported PCRE 5.0.
5912 19. Trivial typo in log message " temporarily refused connection" (the leading
5915 20. If the log selector return_path_on_delivery was set and an address was
5916 redirected to /dev/null, the delivery process crashed because it assumed
5917 that a return path would always be set for a "successful" delivery. In this
5918 case, the whole delivery is bypassed as an optimization, and therefore no
5921 21. Internal re-arrangement: the function for sending a challenge and reading
5922 a response while authentication was assuming a zero-terminated challenge
5923 string. It's now changed to take a pointer and a length, to allow for
5924 binary data in such strings.
5926 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5928 23. Exim was not respecting finduser_retries when seeking the login of the
5929 uid under which it was called; it was always trying 10 times. (The default
5930 setting of finduser_retries is zero.) Also, it was sleeping after the final
5931 failure, which is pointless.
5933 24. Implemented tls_on_connect_ports.
5935 25. Implemented acl_smtp_predata.
5937 26. If the domain in control=submission is set empty, Exim assumes that the
5938 authenticated id is a complete email address when it generates From: or
5939 Sender: header lines.
5941 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5942 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5943 chown and chgrp in /bin and hostname in /usr/bin.
5945 28. Exim was keeping the "process log" file open after each use, just as it
5946 does for the main log. This opens the possibility of it remaining open for
5947 long periods when the USR1 signal hits a daemon. Occasional processlog
5948 errors were reported, that could have been caused by this. Anyway, it seems
5949 much more sensible not to leave this file open at all, so that is what now
5952 29. The long-running daemon process does not normally write to the log once it
5953 has entered its main loop, and it closes the log before doing so. This is
5954 so that log files can straightforwardly be renamed and moved. However,
5955 there are a couple of unusual error situations where the daemon does write
5956 log entries, and I had neglected to close the log afterwards.
5958 30. The text of an SMTP error response that was received during a remote
5959 delivery was being truncated at 512 bytes. This is too short for some of
5960 the long messages that one sometimes sees. I've increased the limit to
5963 31. It is now possible to make retry rules that apply only when a message has a
5964 specific sender, in particular, an empty sender.
5966 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5967 it possible to be selective about when SMTP synchronization is enforced.
5969 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5971 32. Implemented hosts_connection_nolog.
5973 33. Added an ACL for QUIT.
5975 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5978 35. Added mailbox_size and mailbox_filecount to appendfile.
5980 36. Added control = no_multiline_responses to ACLs.
5982 37. There was a bug in the logic of the code that waits for the clock to tick
5983 in the case where the clock went backwards by a substantial amount such
5984 that the microsecond fraction of "now" was more than the microsecond
5985 fraction of "then" (but the whole seconds number was less).
5987 38. Added support for the libradius Radius client library this is found on
5988 FreeBSD (previously only the radiusclient library was supported).
5994 1. When certain lookups returned multiple values in the form name=value, the
5995 quoting of the values was not always being done properly. Specifically:
5996 (a) If the value started with a double quote, but contained no whitespace,
5998 (b) If the value contained whitespace other than a space character (i.e.
5999 tabs or newlines or carriage returns) it was not quoted.
6000 This fix has been applied to the mysql and pgsql lookups by writing a
6001 separate quoting function and calling it from the lookup code. The fix
6002 should probably also be applied to nisplus, ibase and oracle lookups, but
6003 since I cannot test any of those, I have not disturbed their existing code.
6005 2. A hit in the callout cache for a specific address caused a log line with no
6006 reason for rejecting RCPT. Now it says "Previous (cached) callout
6007 verification failure".
6009 3. There was an off-by-one bug in the queryprogram router. An over-long
6010 return line was truncated at 256 instead of 255 characters, thereby
6011 overflowing its buffer with the terminating zero. As well as fixing this, I
6012 have increased the buffer size to 1024 (and made a note to document this).
6014 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6015 when Exim is waiting for an SMTP response from a remote server, Exim
6016 restarts its select() call on the socket, thereby resetting its timeout.
6017 This is not a problem when such interrupts are rare. Somebody set up a cron
6018 job to run exiwhat every 2 minutes, which is less than the normal select()
6019 timeout (5 or 10 minutes). This meant that the select() timeout never
6020 kicked in because it was always reset. I have fixed this by comparing the
6021 time when an interrupt arrives with the time at the start of the first call
6022 to select(). If more time than the timeout has elapsed, the interrupt is
6023 treated as a timeout.
6025 5. Some internal re-factoring in preparation for the addition of Sieve
6026 extensions (by MH). In particular, the "personal" test is moved to a
6027 separate function, and given an option for scanning Cc: and Bcc: (which is
6028 not set for Exim filters).
6030 6. When Exim created an email address using the login of the caller as the
6031 local part (e.g. when creating a From: or Sender: header line), it was not
6032 quoting the local part when it contained special characters such as @.
6034 7. Installed new OpenBSD configuration files.
6036 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6037 try to make them clearer.
6039 9. Callout options, other than the timeout value, were being ignored when
6040 verifying sender addresses in header lines. For example, when using
6042 verify = header_sender/callout=no_cache
6044 the cache was (incorrectly) being used.
6046 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6047 only the Cygwin environment.
6049 11. When return_path_on_delivery was set as a log selector, if different remote
6050 addresses in the same message used different return paths and parallel
6051 remote delivery occurred, the wrong values would sometimes be logged.
6052 (Whenever a remote delivery process finished, the return path value from
6053 the most recently started remote delivery process was logged.)
6055 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6056 lines when AUTH and/or TLS are in use. This is the "received protocol"
6057 field. Exim used to use "asmtp" for authenticated SMTP, without any
6058 indication (in the protocol name) for TLS use. Now it follows the RFC and
6059 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6060 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6061 names appear in log lines as well as in Received: header lines.
6063 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6064 extensions, and comparison tests, and to fix some bugs.
6066 14. Changes to the "personal" filter test:
6068 (1) The test was buggy in that it was just doing the equivalent of
6069 "contains" tests on header lines. For example, if a user's address was
6070 anne@some.where, the "personal" test would incorrectly be true for
6072 To: susanne@some.where
6074 This test is now done by extracting each address from the header in turn,
6075 and checking the entire address. Other tests that are part of "personal"
6076 are now done using regular expressions (for example, to check local parts
6077 of addresses in From: header lines).
6079 (2) The list of non-personal local parts in From: addresses has been
6080 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6081 taken from the Sieve specification recommendations.
6083 (3) If the message contains any header line starting with "List-" it is
6084 treated as non-personal.
6086 (4) The test for "circular" in the Subject: header line has been removed
6087 because it now seems ill-conceived.
6089 15. Minor typos in src/EDITME comments corrected.
6091 16. Installed latest exipick from John Jetmore.
6093 17. If headers_add on a router specified a text string that was too long for
6094 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6095 of string_sprintf() is now avoided.
6097 18. $message_body_size was not set (it was always zero) when running the DATA
6098 ACL and the local_scan() function.
6100 19. For the "mail" command in an Exim filter, no default was being set for
6101 the once_repeat time, causing a random time value to be used if "once" was
6102 specified. (If the value happened to be <= 0, no repeat happened.) The
6103 default is now 0s, meaning "never repeat". The "vacation" command was OK
6104 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6105 (I found it when inspecting the code).
6107 20. There is now an overall timeout for performing a callout verification. It
6108 defaults to 4 times the callout timeout, which applies to individual SMTP
6109 commands during the callout. The overall timeout applies when there is more
6110 than one host that can be tried. The timeout is checked before trying the
6111 next host. This prevents very long delays if there are a large number of
6112 hosts and all are timing out (e.g. when the network connections are timing
6113 out). The value of the overall timeout can be changed by specifying an
6114 additional sub-option for "callout", called "maxwait". For example:
6116 verify = sender/callout=5s,maxwait=20s
6118 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6119 to the end before writing, but this should make it even safer).
6121 22. Exim was forgetting that it had advertised PIPELINING for the second and
6122 subsequent messages on an SMTP connection. It was also not resetting its
6123 memory on STARTTLS and an internal HELO.
6125 23. When Exim logs an SMTP synchronization error within a session, it now
6126 records whether PIPELINING has been advertised or not.
6128 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6129 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6130 rather than long int.
6132 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6134 26. Added the never_mail option to autoreply.
6140 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6141 crash if the getsockname() call failed; this can happen if a connection is
6142 closed very soon after it is established. The problem was simply in the
6143 order in which certain operations were done, causing Exim to try to write
6144 to the SMTP stream before it had set up the file descriptor. The bug has
6145 been fixed by making things happen in the correct order.
6151 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6152 before the connection was closed, thus losing the rejection response.
6154 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6155 some early Solaris releases, but causes trouble in current releases where
6156 socklen_t is defined.
6158 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6161 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6164 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6165 that releasing the top of it at the end releases what was used for sub-
6166 expansions (unless the block got too big). However, discard this block if
6167 the first thing is a variable or header, so that we can use its block when
6168 it is dynamic (useful for very large $message_headers, for example).
6170 6. Lookups now cache *every* query, not just the most recent. A new, separate
6171 store pool is used for this. It can be recovered when all lookup caches are
6172 flushed. Lookups now release memory at the end of their result strings.
6173 This has involved some general refactoring of the lookup sources.
6175 7. Some code has been added to the store_xxx() functions to reduce the amount
6176 of flapping under certain conditions.
6178 8. log_incoming_interface used to affect only the <= reception log lines. Now
6179 it causes the local interface and port to be added to several more SMTP log
6180 lines, for example "SMTP connection from", and rejection lines.
6182 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6184 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6186 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6187 could be overwritten at the end of the current message (or the start of a
6188 new message if it was set in a HELO ACL). The value is now preserved for
6189 the duration of the SMTP connection.
6191 12. If a transport had a headers_rewrite setting, and a matching header line
6192 contained an unqualified address, that address was qualified, even if it
6193 did not match any rewriting rules. The underlying bug was that the values
6194 of the flags that permit the existence of unqualified sender and recipient
6195 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6196 non-local messages, and by -bnq for local messages) were not being
6197 preserved with the message after it was received.
6199 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6200 "next input=" as part of the text comprising the host identity instead of
6201 the correct text. The code was using the same buffer for two different
6202 strings. However, depending on which order the printing function evaluated
6203 its arguments, the bug did not always show up. Under Linux, for example, my
6204 test suite worked just fine.
6206 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6207 doesn't work with some older versions of Perl. It has been changed to "my",
6208 which in any case is probably the better facility to use.
6210 15. A really picky compiler found some instances of statements for creating
6211 error messages that either had too many or two few arguments for the format
6214 16. The size of the buffer for calls to the DNS resolver has been increased
6215 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6216 for addresses that have a lot of PTR records. This alleviates a problem; it
6217 does not fully solve it.
6219 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6220 buffer now truncates the list and logs the incident, which is the same
6221 action as happens when Exim is looking up a host name and its aliases.
6222 Previously in this situation something unpredictable would happen;
6223 sometimes it was "internal error: store_reset failed".
6225 18. If a server dropped the connection unexpectedly when an Exim client was
6226 using GnuTLS and trying to read a response, the client delivery process
6227 crashed while trying to generate an error log message.
6229 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6230 string, for example:
6232 warn message = H1: something\nH2: something
6234 the text was added as a single header line from Exim's point of view
6235 though it ended up OK in the delivered message. However, searching for the
6236 second and subsequent header lines using $h_h2: did not work. This has been
6237 fixed. Similarly, if a system filter added multiple headers in this way,
6238 the routers could not see them.
6240 20. Expanded the error message when iplsearch is called with an invalid key to
6241 suggest using net-iplsearch in a host list.
6243 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6244 ACLs are no longer actually imposed (and a message to that effect is
6247 22. If a "gecos" field in a passwd entry contained escaped characters, in
6248 particular, if it contained a \" sequence, Exim got it wrong when building
6249 a From: or a Sender: header from that name. A second bug also caused
6250 incorrect handling when an unquoted " was present following a character
6251 that needed quoting.
6253 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6254 was not being matched caselessly.
6256 24. Arranged for all hyphens in the exim.8 source to be escaped with
6259 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6260 properly. Recipient callout cache records were still being keyed to include
6261 the sender, even when use_sender was set false. This led to far more
6262 callouts that were necessary. The sender is no longer included in the key
6263 when use_sender is false.
6265 26. Added "control = submission" modifier to ACLs.
6267 27. Added the ${base62d: operator to decode base 62 numbers.
6269 28. dnsdb lookups can now access SRV records.
6271 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6272 the configuration file.
6274 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6275 (-v) mode. This makes the output for a verbose queue run more intelligible.
6277 31. Added a use_postmaster feature to recipient callouts.
6279 32. Added the $body_zerocount variable, containing the number of binary zero
6280 bytes in the message body.
6282 33. The time of last modification of the "new" subdirectory is now used as the
6283 "mailbox time last read" when there is a quota error for a maildir
6286 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6288 35. Added +ignore_unknown as a special item in host lists.
6290 36. Code for decoding IPv6 addresses in host lists is now included, even if
6291 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6292 address was recognized as an IP address, but was then not correctly decoded
6293 into binary, causing unexpected and incorrect effects when compared with
6300 1. Very minor rewording of debugging text in manualroute to say "list of
6301 hosts" instead of "hostlist".
6303 2. If verify=header_syntax was set, and a header line with an unqualified
6304 address (no domain) and a large number of spaces between the end of the
6305 name and the colon was received, the reception process suffered a buffer
6306 overflow, and (when I tested it) crashed. This was caused by some obsolete
6307 code that should have been removed. The fix is to remove it!
6309 3. When running in the test harness, delay a bit after writing a bounce
6310 message to get a bit more predictability in the log output.
6312 4. Added a call to search_tidyup() just before forking a reception process. In
6313 theory, someone could use a lookup in the expansion of smtp_accept_max_
6314 per_host which, without the tidyup, could leave open a database connection.
6316 5. Added the variables $recipient_data and $sender_data which get set from a
6317 lookup success in an ACL "recipients" or "senders" condition, or a router
6318 "senders" option, similar to $domain_data and $local_part_data.
6320 6. Moved the writing of debug_print from before to after the "senders" test
6323 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6324 problems for message scanning, either using a data ACL, or using
6325 local_scan() because the Received: header was not generated till after they
6326 were called (in order to set the time as the time of reception completion).
6327 I have revised the way this works. The header is now generated after the
6328 body is received, but before the ACL or local_scan() are called. After they
6329 are run, the timestamp in the header is updated.
6335 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6336 before starting a queue runner without re-exec. This happened only when
6337 deliver_drop_privilege was set or when the Exim user was set to root. The
6338 effect of the bug was that timeouts during subsequent deliveries caused
6339 crashes instead of being properly handled. The handler is now left at its
6340 default (and expected) setting.
6342 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6343 message, again when deliver_drop_privilege is set or Exim is run as root.
6344 The bug described in (1) was not present in this case, but the tidying up
6345 of the other signals was missing. I have made the two cases consistent.
6347 3. The ignore_target_hosts setting on a manualroute router was being ignored
6348 for hosts that were looked up using the /MX notation.
6350 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6353 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6354 operated on the sender address. After changing the $sender_address to <>
6355 for the sender address verify, Exim was re-instated it as the original
6356 (before rewriting) address, but remembering that it had rewritten it, so it
6357 wasn't rewriting it again. This bug also had the effect of breaking the
6358 sender address verification caching when the sender address was rewritten.
6360 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6361 This has been changed so that if the ip literal address matches
6362 ignore_target_hosts, the router declines.
6364 7. Added expansion conditions match_domain, match_address, and match_local_
6365 part (NOT match_host).
6367 8. The placeholder for the Received: header didn't have a length field set.
6369 9. Added code to Exim itself and to exim_lock to test for a specific race
6370 condition that could lead to file corruption when using MBX delivery. The
6371 issue is with the lockfile that is created in /tmp. If this file is removed
6372 after a process has opened it but before that process has acquired a lock,
6373 there is the potential for a second process to recreate the file and also
6374 acquire a lock. This could lead to two Exim processes writing to the file
6375 at the same time. The added code performs the same test as UW imapd; it
6376 checks after acquiring the lock that its file descriptor still refers to
6377 the same named file.
6379 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6380 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6383 11. Added the smtp_active_hostname option. If used, this will typically be made
6384 to depend on the incoming interface address. Because $interface_address is
6385 not set up until the daemon has forked a reception process, error responses
6386 that can happen earlier (such as "too many connections") no longer contain
6389 12. If an expansion in a condition on a "warn" statement fails because a lookup
6390 defers, the "warn" statement is abandoned, and the next ACL statement is
6391 processed. Previously this caused the whole ACL to be aborted.
6393 13. Added the iplsearch lookup type.
6395 14. Added ident_timeout as a log selector.
6397 15. Added tls_certificate_verified as a log selector.
6399 16. Added a global option tls_require_ciphers (compare the smtp transport
6400 option of the same name). This controls incoming TLS connections.
6402 17. I finally figured out how to make tls_require_ciphers do a similar thing
6403 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6404 before starting the TLS session.
6406 18. Tabs are now shown as \t in -bP output.
6408 19. If the log selector return_path_on_delivery was set, Exim crashed when
6409 bouncing a message because it had too many Received: header lines.
6411 20. If two routers both had headers_remove settings, and the first one included
6412 a superfluous trailing colon, the final name in the first list and the
6413 first name in the second list were incorrectly joined into one item (with a
6414 colon in the middle).
6420 1. Added -C and -D options to the exinext utility, mainly to make it easier
6421 to include in the automated testing, but these could be helpful when
6422 multiple configurations are in use.
6424 2. The exinext utility was not formatting the output nicely when there was
6425 an alternate port involved in the retry record key, nor when there was a
6426 message id as well (for retries that were specific to a specific message
6427 and a specific host). It was also confused by IPv6 addresses, because of
6428 the additional colons they contain. I have fixed the IPv4 problem, and
6429 patched it up to do a reasonable job for IPv6.
6431 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6432 delivery, the log line now contains "pipelined" if PIPELINING was used.
6434 4. An SMTP transport process used to panic and die if the bind() call to set
6435 an explicit outgoing interface failed. This has been changed; it is now
6436 treated in the same way as a connect() failure.
6438 5. A reference to $sender_host_name in the part of a conditional expansion
6439 that was being skipped was still causing a DNS lookup. This no longer
6442 6. The def: expansion condition was not recognizing references to header lines
6443 that used bh_ and bheader_.
6445 7. Added the _cache feature to named lists.
6447 8. The code for checking quota_filecount in the appendfile transport was
6448 allowing one more file than it should have been.
6450 9. For compatibility with Sendmail, the command line option
6458 and sets the incoming protocol and host name (for trusted callers). The
6459 host name and its colon can be omitted when only the protocol is to be set.
6460 Note the Exim already has two private options, -pd and -ps, that refer to
6461 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6462 "s", but I don't think that's a major issue.
6464 10. A number of refactoring changes to the code, none of which should affect
6467 (a) The number of logging options was getting close to filling up the
6468 32-bit word that was used as a bit map. I have split them into two classes:
6469 those that are passed in the argument to log_write(), and those that are
6470 only ever tested independently outside of that function. These are now in
6471 separate 32-bit words, so there is plenty of room for expansion again.
6472 There is no change in the user interface or the logging behaviour.
6474 (b) When building, for example, log lines, the code previously used a
6475 macro that called string_cat() twice, in order to add two strings. This is
6476 not really sufficiently general. Furthermore, there was one instance where
6477 it was actually wrong because one of the argument was used twice, and in
6478 one call a function was used. (As it happened, calling the function twice
6479 did not affect the overall behaviour.) The macro has been replaced by a
6480 function that can join an arbitrary number of extra strings onto a growing
6483 (c) The code for expansion conditions now uses a table and a binary chop
6484 instead of a serial search (which was left over from when there were very
6485 few conditions). Also, it now recognizes conditions like "pam" even when
6486 the relevant support is not compiled in: a suitably worded error message is
6487 given if an attempt is made to use such a condition.
6489 11. Added ${time_interval:xxxxx}.
6491 12. A bug was causing one of the ddress fields not to be passed back correctly
6492 from remote delivery subprocesses. The field in question was not being
6493 subsequently used, so this caused to problems in practice.
6495 13. Added new log selectors queue_time and deliver_time.
6497 14. Might have fixed a bug in maildirsizefile handling that threw up
6498 "unexpected character" debug warnings, and recalculated the data
6499 unnecessarily. In any case, I expanded the warning message to give more
6502 15. Added the message "Restricted characters in address" to the statements in
6503 the default ACL that block characters like @ and % in local parts.
6505 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6506 Three changes have been made:
6508 (a) There was a serious bug; a negative response to MAIL caused the whole
6509 recipient domain to be cached as invalid, thereby blocking all messages
6510 to all local parts at the same domain, from all senders. This bug has
6511 been fixed. The domain is no longer cached after a negative response to
6512 MAIL if the sender used is not empty.
6514 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6517 (c) A new callout option, "use_sender" has been added for people who want
6518 the modified behaviour.
6524 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6527 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6528 indeed breaks things for older releases.
6530 3. Added additional logging to the case where there is a problem reading data
6531 from a filter that is running in a subprocess using a pipe, in order to
6532 try to track down a specific problem.
6534 4. Testing facility fudge: when running in the test harness and attempting
6535 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6536 getting "No route to host". Convert this to a timeout.
6538 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6541 6. Some OS don't have socklen_t but use size_t instead. This affects the
6542 fifth argument of getsockopt() amongst other things. This is now
6543 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6544 can be set for individual OS. I have set it for SunOS5, OSF1, and
6545 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6546 some earlier ones do not.
6548 7. Change 4.30/15 was not doing the test caselessly.
6550 8. The standard form for an IPv6 address literal was being rejected by address
6551 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6552 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6553 this, as well as the form without the "IPv6" on the front (but only when
6554 address literals are enabled, of course).
6556 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6558 10. Exim crashed if a message with an empty sender address specified by -f
6559 encountered a router with an errors_to setting. This could be provoked only
6560 by a command such as
6564 where an empty string was supplied; "<>" did not hit this bug.
6566 11. Installed PCRE release 4.5.
6568 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6569 remained set. It is now erased.
6571 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6572 times from message ids (which are base 36 rather than the normal 62).
6574 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6575 were being counted as actual protocol errors, and logged if the log
6576 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6577 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6578 and DATA following a set of rejected RCPTs do not count as protocol errors.
6579 In other words, Exim assumes they were pipelined, though this may not
6580 actually be the case. Of course, in all cases the client gets an
6581 appropriate error code.
6583 15. If a lookup fails in an ACL condition, a message about the failure may
6584 be available; it is used if testing the ACL cannot continue, because most
6585 such messages specify what the cause of the deferral is. However, some
6586 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6587 that caused an old message to be retained and used if a later statement
6588 caused a defer, replacing the real cause of the deferral.
6590 16. If an IP address had so many PTR records that the DNS lookup buffer
6591 was not large enough to hold them, Exim could crash while trying to process
6592 the truncated data. It now detects and logs this case.
6594 17. Further to 4.21/58, another change has been made: if (and only if) the
6595 first line of a message (the first header line) ends with CRLF, a bare LF
6596 in a subsequent header line has a space inserted after it, so as not to
6597 terminate the header.
6599 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6600 unnecessarily, used atoi() instead of strtol(), and didn't check the
6601 termination when getting file sizes from file names by regex.
6603 19. Completely re-implemented the support for maildirsize files, in the light
6604 of a number of problems with the previous contributed implementation
6605 (4.30/29). In particular:
6607 . If the quota is zero, the maildirsize file is maintained, but no quota is
6610 . If the maildir directory does not exist, it is created before any attempt
6611 to write a maildirsize file.
6613 . The quota value in the file is just a cache; if the quota is changed in
6614 the transport, the new value overrides.
6616 . A regular expression is available for excluding directories from the
6619 20. The autoreply transport checks the characters in options that define the
6620 message's headers; it allows continued headers, but it was checking with
6621 isspace() after an embedded newline instead of explicitly looking for a
6624 21. If all the "regular" hosts to which an address was routed had passed their
6625 expiry times, and had not reached their retry times, the address was
6626 bounced, even if fallback hosts were defined. Now Exim should go on to try
6629 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6630 equivalent code in the SMTP transport. Some hosts send humungous responses
6631 to HELO/EHLO, more than 1024 it seems.
6633 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6634 gives compiler warnings in some environments. I've now done it "properly",
6637 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6638 (because of problems with the built-in one) was declared to return uschar *
6639 instead of char *, causing compiler failure.
6641 25. Fixed a file descriptor leak when processing alias/forward files.
6643 26. Fixed a minor format string issue in dbfn.c.
6645 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6647 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6648 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6651 29. When a sender address is verified, it is cached, to save repeating the test
6652 when there is more than one recipient in a message. However, when the
6653 verification involves a callout, it is possible for different callout
6654 options to be set for different recipients. It is too complicated to keep
6655 track of this in the cache, so now Exim always runs a verification when a
6656 callout is required, relying on the callout cache for the optimization.
6657 The overhead is duplication of the address routing, but this should not be
6660 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6661 to be verified with callout=postmaster, and the main callout worked but the
6662 postmaster check failed, the verification correctly failed. However, if a
6663 subsequent RCPT command asked for sender verification *without* the
6664 postmaster check, incorrect caching caused this verification also to fail,
6667 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6668 it was not caching the DNS options (qualify_single, search_parents) that
6669 were used when the lookup failed. A subsequent lookup with different
6670 options therefore always gave the same answer, though there were cases
6671 where it should not have. (Example: a "domains = !$mx_any" option on a
6672 dnslookup router: the "domains" option is always processed without any
6673 widening, but the router might have qualify_single set.) Now Exim uses the
6674 cached value only when the same options are set.
6676 32. Added John Jetmore's "exipick" utility to the distribution.
6678 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6679 than a timeout (e.g. a certificate is required, and is not provided), an
6680 Exim server now closes the connection immediately. Previously it waited for
6681 the client to close - but if the client is SSL, it seems that they each
6682 wait for each other, leading to a delay before one of them times out.
6684 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6685 maintained 0.8.x compatibility because I don't think many are using it, and
6686 it is clearly obsolete.
6688 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6691 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6692 client certificate was expired. A simple patch fixes this, though I don't
6693 understand the full logic of why the verify callback is called multiple
6696 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6697 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6698 which causes problems with some clients (such as the Certicom SSL Plus
6699 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6700 disables the coutermeasure allowing Eudora to connect."
6702 38. Exim was not checking that a write() to a log file succeeded. This could
6703 lead to Bad Things if a log got too big, in particular if it hit a file
6704 size limit. Exim now panics and dies if it cannot write to a log file, just
6705 as it does if it cannot open a log file.
6707 39. Modified OS/Makefile-Linux so that it now contains
6709 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6711 The two -D definitions ensure that Exim is compiled with large file
6712 support, which makes it possible to handle log files that are bigger than
6715 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6716 instance) a domain was checked against a named list that involved a lookup,
6717 causing $domain_data to be set, then another domain was checked against the
6718 same list, then the first domain was re-checked, the value of $domain_data
6719 after the final check could be wrong. In particular, if the second check
6720 failed, it could be set empty. This bug probably also applied to
6723 41. The strip_trailing_dot option was not being applied to the address given
6724 with the -f command-line option.
6726 42. The code for reading a message's header from the spool was incrementing
6727 $received_count, but never initializing it. This meant that the value was
6728 incorrect (doubled) while delivering a message in the same process in which
6729 it was received. In the most common configuration of Exim, this never
6730 happens - a fresh exec is done - but it can happen when
6731 deliver_drop_privilege is set.
6733 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6734 it now includes up to 150 characters of the unexpected data in the log
6737 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6738 and building data strings. The size of both of these buffers was 10 000
6739 bytes - far larger than anybody would *ever* want, thought I. Needless to
6740 say, somebody hit the limit. I have increased the maximum line length to
6741 20 000 and the maximum data length of concatenated lines to 100 000. I have
6742 also fixed two bugs, because there was no checking on these buffers. Tsk,
6743 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6744 buffer is too small.
6746 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6747 lsearch lookups. Now it does.
6749 46. When parsing a route_list item in a manualroute router, a fixed-length
6750 buffer was used for the list of hosts. I made this 1024 bytes long,
6751 thinking that nobody would ever have a list of hosts that long. Wrong.
6752 Somebody had a whole pile of complicated expansion conditions, and the
6753 string was silently truncated, leading to an expansion error. It turns out
6754 that it is easier to change to an unlimited length (owing to other changes
6755 that have happened since this code was originally written) than to build
6756 structure for giving a limitation error. The length of the item that
6757 expands into the list of hosts is now unlimited.
6759 47. The lsearch lookup could not handle data where the length of text line was
6760 more than 4095 characters. Such lines were truncated, leading to shortened
6761 data being returned. It should now handle lines of any length.
6763 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6764 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6767 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6768 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6769 no longer happens in comment lines. A list of replacements is now placed
6770 at the head of all of the source files, except those whose only change is
6771 to replace PERL_COMMAND in the very first #! line.
6773 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6774 messages on the queue, with a bottom-up merge sort, using code contributed
6775 by Michael Haardt. This should make operations like -bp somewhat faster on
6776 large queues. It won't affect queue runners, except when queue_run_in_order
6779 51. Installed eximstats 1.31 in the distribution.
6781 52. Added support for SRV lookups to the dnslookup router.
6783 53. If an ACL referred to $message_body or $message_body_end, the value was not
6784 reset for any messages that followed in the same SMTP session.
6786 54. The store-handling optimization for building very long strings was not
6787 differentiating between the different store pools. I don't think this
6788 actually made any difference in practice, but I've tidied it.
6790 55. While running the routers to verify a sender address, $sender_address
6791 was still set to the sender address. This is wrong, because when routing to
6792 send a bounce to the sender, it would be empty. Therefore, I have changed
6793 it so that, while verifying a sender address, $sender_address is set to <>.
6794 (There is no change to what happens when verifying a recipient address.)
6796 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6797 target A or AAAA records (if not already returned) without resetting the
6798 qualify_single or search_parents options of the DNS resolver. These are
6799 inappropriate in this case because the targets of MX and SRV records must
6800 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6801 target that, when qualified, matched something in the local domain. These
6802 two options are now turned off when doing these lookups.
6804 57. It seems that at least some releases of Reiserfs (which does not have the
6805 concept of a fixed number of inodes) returns zero and not -1 for the
6806 number of available inodes. This interacted badly with check_spool_inodes,
6807 which assumed that -1 was the "no such thing" setting. What I have done is
6808 to check that the total number of inodes is greater than zero before doing
6809 the test of how many are available.
6811 58. When a "warn" ACL statement has a log_message modifier, the message is
6812 remembered, and not repeated. This is to avoid a lot of repetition when a
6813 message has many recipients that cause the same warning to be written.
6814 However, Exim was preserving the list of already written lines for an
6815 entire SMTP session, which doesn't seem right. The memory is now reset if a
6816 new message is started.
6818 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6819 debugging output unless log_rewrite was also set.
6821 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6822 of (int)(handle) when we know that handle contains (void *)(-1).
6824 61. The Exim daemon panic-logs an error return when it closes the incoming
6825 connection. However "connection reset by peer" seems to be common, and
6826 isn't really an error worthy of noting specially, so that particular error
6829 62. When Exim is trying to find all the local interfaces, it used to panic and
6830 die if the ioctl to get the interface flags failed. However, it seems that
6831 on at least one OS (Solaris 9) it is possible to have an interface that is
6832 included in the list of interfaces, but for which you get a failure error
6833 for this call. This happens when the interface is not "plumbed" into a
6834 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6835 failure of the "get flags" call assumes that the interface is down.
6837 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6838 makes life easier for people who are doing arithmetic on fields extracted
6839 from dates, where you often get leading zeros that should not be
6840 interpreted as octal.
6842 64. Added qualify_domain to the redirect router, to override the global
6845 65. If a pathologically long header line contained very many addresses (the
6846 report of this problem mentioned 10 000) and each of them was rewritten,
6847 Exim could use up a very large amount of memory. (It kept on making new
6848 copies of the header line as it rewrote, and never released the old ones.)
6849 At the expense of a bit more processing, the header rewriting function has
6850 been changed so that it no longer eats memory in this way.
6852 66. The generation of the Received: header has been moved from the time that a
6853 message starts to be received, to the time that it finishes. The timestamp
6854 in the Received: header should now be very close to that of the <= log
6855 line. There are two side-effects of this change:
6857 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6858 logged header lines no longer include the local Received: line, because
6859 it has not yet been created. The same applies to a copy of the message
6860 that is returned to a non-SMTP sender when a message is rejected.
6862 (b) When a filter file is tested using -bf, no additional Received: header
6863 is added to the test message. After some thought, I decided that this
6866 This change does not affect the value of $received_for. It is still set
6867 after address rewriting, but before local_scan() is called.
6869 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6871 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6872 gave an unhelpful panic error message, and a defer error. I have managed to
6873 change this behaviour so that it now rejects any supplied certificate,
6874 which seems right, as the list of acceptable certificates is empty.
6876 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6877 gave an unhelpful defer error. I have not managed to make this reject any
6878 supplied certificates, but the error message it gives is "no certificate
6879 supplied", which is not helpful.
6881 70. exigrep's output now also includes lines that are not associated with any
6882 message, but which match the given pattern. Implemented by a patch from
6883 Martin Sluka, which also tidied up the Perl a bit.
6885 71. Recipient callout verification, like sender verification, was using <> in
6886 the MAIL FROM command. This isn't really the right thing, since the actual
6887 sender may affect whether the remote host accepts the recipient or not. I
6888 have changed it to use the actual sender in the callout; this means that
6889 the cache record is now keyed on a recipient/sender pair, not just the
6890 recipient address. There doesn't seem to be a real danger of callout loops,
6891 since a callout by the remote host to check the sender would use <>.
6892 [SEE ABOVE: changed after hitting problems.]
6894 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6895 temporary errors. However, in the case of such a code being given after
6896 the end of a data transmission (i.e. after ".") Exim was failing to write
6897 a retry record for the message. (Yes, there was some broken host that was
6898 actually sending 8xx at this point.)
6900 73. An unknown lookup type in a host list could cause Exim to panic-die when
6901 the list was checked. (An example that provoked this was putting <; in the
6902 middle of a list instead of at the start.) If this happened during a DATA
6903 ACL check, a -D file could be left lying around. This kind of configuration
6904 error no longer causes Exim to die; instead it causes a defer error. The
6905 incident is still logged to the main and panic logs.
6907 74. Buglet left over from Exim 3 conversion. The message "too many messages
6908 in one connection" was written to the rejectlog but not the mainlog, except
6909 when address rewriting (yes!) was being logged.
6911 75. Added write_rejectlog option.
6913 76. When a system filter was run not as root (that is, when system_filter_user
6914 was set), the values of the $n variables were not being returned to the
6915 main process; thus, they were not subsequently available in the $sn
6918 77. Added +return_path_on_delivery log selector.
6920 78. A connection timeout was being treated differently from recipients deferred
6921 when testing hosts_max_try with a message that was older than the host's
6922 retry timeout. (The host should not be counted, thus allowing all hosts to
6923 be tried at least once before bouncing.) This may have been the cause of an
6924 occasionally reported bug whereby a message would remain on the queue
6925 longer than the retry timeout, but would be bounced if a delivery was
6926 forced. I say "may" because I never totally pinned down the problem;
6927 setting up timeout/retry tests is difficult. See also the next item.
6929 79. The ultimate address timeout was not being applied to errors that involved
6930 a combination of host plus message (for example, a timeout on a MAIL
6931 command). When an address resolved to a number of possible hosts, and they
6932 were not all tried for each delivery (e.g. because of hosts_max_try), a
6933 message could remain on the queue longer than the retry timeout.
6935 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6938 81. Fixed an obscure SMTP outgoing bug which required at least the following
6939 conditions: (a) there was another message waiting for the same server;
6940 (b) the server returned 5xx to all RCPT commands in the first message so
6941 that the message was not completed; (c) the server dropped the connection
6942 or gave a negative response to the RSET that Exim sends to abort the
6943 transaction. The observed case was a dropped connection after DATA that had
6944 been sent in pipelining mode. That is, the server had advertised PIPELINING
6945 but was not implementing it correctly. The effect of the bug was incorrect
6946 behaviour, such as trying another host, and this could lead to a crash.
6952 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6953 and daemon.c were passed as pointers to ints; they should have been
6954 pointers to socklen_t variables (which are typically unsigned ints).
6956 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6959 3. Fixed a really odd bug that affected only the testing scheme; patching a
6960 certain fixed string in the binary changed the value of another string that
6961 happened to be identical to the end of the original first string.
6963 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6964 name", it returns that address as the IP address. On some operating
6965 systems (e.g. Solaris), it also passes back the IP address string as the
6966 "host name". However, on others (e.g. Linux), it passes back an empty
6967 string. Exim wasn't checking for this, and was changing the host name to an
6968 empty string, assuming it had been canonicalized.
6970 5. Although rare, it is permitted to have more than one PTR record for a given
6971 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6972 all the names associated with an address, because they do in Solaris.
6973 However, it seems that they do not in Linux for data that comes from the
6974 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6975 I found this out when I moved to a new Linux workstation and tried to run
6976 the Exim test suite.
6978 To get round this problem I have changed the code so that it now does its
6979 own call to the DNS to look up PTR records when searching for a host name.
6980 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6981 addresses that are only in /etc/hosts are still found.
6983 This behaviour is, however, controlled by an option called host_lookup_
6984 order, which defaults to "bydns:byaddr". If people want to use the other
6985 order, or indeed, just use one or the other means of lookup, they can
6986 specify it in this variable.
6988 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6989 some operating systems, this comes back from gethostbyaddr() as an empty
6990 string, and this is what Exim used to test for. However, it seems that in
6991 other systems, "." is yielded. Exim now tests for this case too.
6993 7. The values of check_spool_space and check_log_space are now held internally
6994 as a number of kilobytes instead of an absolute number of bytes. If a
6995 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6996 kilobyte. This means that much larger values can be stored.
6998 8. Exim monitor: an attempt to get the action menu when not actually pointing
6999 at a message produces an empty menu entitled "No message selected". This
7000 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7001 no entries in it ("Shell widget menu has zero width and/or height"). So I
7002 have added a single, blank menu entry in this case.
7004 9. Added ${quote_local_part.
7006 10. MIME decoding is now applied to the contents of Subject: header lines when
7009 11. Now that a reference to $sender_host_address automatically causes a reverse
7010 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7011 host lookup before query-style lookups in lists that might use this
7012 variable. This has therefore been abolished, and the "net-" prefix is no
7013 longer necessary for query-style lookups.
7015 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7016 have been a typo for LFLAGS, so it has been changed.
7018 13. The install script calls Exim with "-C /dev/null" in order to find the
7019 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7020 to be output. However, since Exim outputs its version number before the
7021 error, it didn't break the script. It just looked ugly. I fixed this by
7022 always allowing "-C /dev/null" if the caller is root.
7024 14. Ignore overlarge ACL variable number when reading spool file - insurance
7025 against a later release with more variables having written the file.
7027 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7028 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7029 this, as well as the form without the "IPv6" on the front.
7031 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7032 OS/Makefile-Darwin file.
7034 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7035 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7037 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7038 variable. However, some OS use ints for inodes. I've added cast to long int
7039 to get rid of the compiler warning.
7041 19. I had forgotten to lock out "/../" in configuration file names when
7042 ALT_CONFIG_PREFIX was set.
7044 20. Routers used for verification do not need to specify transports. However,
7045 if such a router generated a host list, and callout was configured, Exim
7046 crashed, because it could not find a port number from the (non-existent)
7047 transport. It now assumes port 25 in this circumstance.
7049 21. Added the -t option to exigrep.
7051 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7052 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7053 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7054 them is set, LOOKUP_LSEARCH is forced.
7056 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7058 24. Added sender and host information to the "rejected by local_scan()" log
7059 line; previously there was no indication of these.
7061 25. Added .include_if_exists.
7063 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7064 Linux. It turns out that not all file systems support this. Apparently some
7065 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7066 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7067 not supported on the file descriptor, is now ignored when Exim is trying to
7068 sync a directory. This applies only to Linux.
7070 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7072 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7073 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7076 29. Added support for maildirsize files from supplied patch (modified a bit).
7078 30. The use of :fail: followed by an empty string could lead Exim to respond to
7079 sender verification failures with (e.g.):
7081 550 Verification failed for <xxx>
7082 550 Sender verify failed
7084 where the first response line was missing the '-' that indicates it is not
7085 the final line of the response.
7087 31. The loop for finding the name of the user that called Exim had a hardwired
7088 limit of 10; it now uses the value of finduser_retries, which is used for
7089 all other user lookups.
7091 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7094 33. Exim was neglecting to zero errno before one call of strtol() when
7095 expanding a string and expecting an integer value. On some systems this
7096 resulted in spurious "integer overflow" errors. Also, it was casting the
7097 result into an int without checking.
7099 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7100 did not work. The code looks as if it has *never* worked, though it appears
7101 to have been documented since at least release 1.62. I have made it work.
7103 35. The "timeout_DNS" error in retry rules, also documented since at least
7104 1.62, also never worked. As it isn't clear exactly what this means, and
7105 clearly it isn't a major issue, I have abolished the feature by treating it
7106 as "timeout", and writing a warning to the main and panic logs.
7108 36. The display of retry rules for -brt wasn't always showing the error code
7111 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7112 timeout_connect_A, timeout_connect_MX.
7114 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7115 to the empty sender.
7117 39. The daemon was not analysing the content of -oX till after it had closed
7118 stderr and disconnected from the controlling terminal. This meant that any
7119 syntax errors were only noted on the panic log, and the return code from
7120 the command was 0. By re-arranging the code a little, I've made the
7121 decoding happen first, so such errors now appear on stderr, and the return
7122 code is 1. However, the actual setting up of the sockets still happens in
7123 the disconnected process, so errors there are still only recorded on the
7126 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7127 connections (as happens on some IP stacks) was logged at start up time as
7128 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7129 it from "IPv6 and IPv4", which means that two separate sockets are being
7132 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7133 says whether AF_INET or AF_INET6 was passed as an argument.
7135 42. Exiwhat output was messed up when time zones were included in log
7138 43. Exiwhat now gives more information about the daemon's listening ports,
7139 and whether -tls-on-connect was used.
7141 44. The "port" option of the smtp transport is now expanded.
7143 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7144 silently ignored. Now an error message is written to the main and panic
7147 46. There's a new ACL modifier called "logwrite" which writes to a log file
7148 as soon as it is encountered.
7150 47. Added $local_user_uid and $local_user_gid at routing time.
7152 48. Exim crashed when trying to verify a sender address that was being
7155 49. Exim was recognizing only a space character after ".include". It now also
7156 recognizes a tab character.
7158 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7159 extracting the relevant information from the specification. The man page no
7160 longer contains scrambled data for the -d option, and I've added a section
7161 at the front about calling Exim under different names.
7163 51. Added "extra_headers" argument to the "mail" command in filter files.
7165 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7168 53. Installed eximstats 1.29.
7170 54. Added transport_filter_timeout as a generic transport option.
7172 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7173 Cc: header lines. This was required by RFC 822, but it not required by RFC
7176 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7177 incoming messages that did not have them. Now it does so only if the
7178 message originates locally, that is, if there is no associated remote host
7179 address. When Resent- header lines are present, this applies to the Resent-
7180 lines rather than the non-Resent- lines.
7182 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7183 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7185 58. Messages for configuration errors now include the name of the main
7186 configuration file - useful now that there may be more than one file in a
7187 list (.included file names were always shown).
7189 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7190 for those rare installations that do not start the daemon as root or run it
7191 setuid root. I've cut out the call to initgroups() if the daemon is not
7194 60. The Exim user and group can now be bound into the binary as text strings
7195 that are looked up at the start of Exim's processing.
7197 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7199 62. Added $mailstore_basename variable.
7201 63. Installed patch to sieve.c from Michael Haardt.
7203 64. When Exim failed to open the panic log after failing to open the main log,
7204 the original message it was trying to log was written to stderr and debug
7205 output, but if they were not available (the usual case in production), it
7206 was lost. Now it is written to syslog before the two lines that record the
7207 failures to open the logs.
7209 65. Users' Exim filters run in subprocesses under the user's uid. It is
7210 possible for a "deliver" command or an alias in a "personal" command to
7211 provoke an address rewrite. If logging of address rewriting is configured,
7212 this fails because the process is not running as root or exim. There may be
7213 a better way of dealing with this, but for the moment (because 4.30 needs
7214 to be released), I have disabled address rewrite logging when running a
7215 filter in a non-root, non-exim process.
7221 1. The buildconfig auxiliary program wasn't quoting the value set for
7222 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7223 not defined. This bug was masked in 4.22 by the effect that was fixed in
7226 2. Some messages that were rejected after a message id was allocated were
7227 shown as "incomplete" by exigrep. It no longer does this for messages that
7228 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7230 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7231 have allow_domain_literals set, the ID did not get logged in the <= line.
7232 Domain literals are now always recognized in Message-ID: header lines.
7234 4. The first argument for a ${extract expansion item is the key name or field
7235 number. Leading and trailing spaces in this item were not being ignored,
7236 causing some misleading effects.
7238 5. When deliver_drop_privilege was set, single queue runner processes started
7239 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7240 same command in the process it spins off) were not dropping privilege.
7242 6. When the daemon running as "exim" started a queue runner, it always
7243 re-executed Exim in the spun-off process. This is a waste of effort when
7244 deliver_drop_privilege is set. The new process now just calls the
7245 queue-runner function directly.
7251 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7254 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7255 ignored. Though the use of -oP was forcing the writing of a pid file, it
7256 was always written to the default place.
7258 3. If the message "no IP address found for host xxxx" is generated during
7259 incoming verification, it is now followed by identification of the incoming
7260 connection (so you can more easily find what provoked it).
7262 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7264 5. Added some features to "harden" Exim a bit more against certain attacks:
7266 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7267 be put in Local/Makefile. This is like the never_users runtime option,
7268 but it cannot be overridden. The default setting is "root".
7270 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7271 prefix string with which any file named in a -C command line option
7274 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7275 is retained for -C and -D only if the caller of Exim is root. Without
7276 it, the exim user may also use -C and -D and retain privilege.
7278 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7279 command line option is disabled.
7281 6. Macro names set by the -D option must start with an upper case letter, just
7282 like macro names defined in the configuration file.
7284 7. Added "dereference=" facility to LDAP.
7286 8. Two instances of the typo "uknown" in the source files are fixed.
7288 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7289 the Configure-Makefile script screwed up while processing it.
7291 10. Incorporated PCRE 4.4.
7293 11. The SMTP synchronization check was not operating right at the start of an
7294 SMTP session. For example, it could not catch a HELO sent before the client
7295 waited for the greeting. There is now a check for outstanding input at the
7296 point when the greeting is written. Because of the duplex, asynchronous
7297 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7298 way, but not yet received, when the check is performed.
7300 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7301 on TCP/IP sockets, because this apparently causes some broken clients to
7304 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7305 unchanged) from the Cygwin maintainer.
7307 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7308 was supported instead of testing for mbx. Effectively a typo.
7310 15. The spa authenticator server code was not checking that the input it
7311 received was valid base64.
7313 16. The debug output line for the "set" modifier in ACLs was not showing the
7314 name of the variable that was being set.
7316 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7318 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7319 lookup on its own. Something else was needed to trigger the lookup. For
7320 example, a match in host_lookup or the need for a host name in a host list.
7321 Now, if $sender_host_name is referenced and the host name has not yet been
7322 looked up, a lookup is performed. If the lookup fails, the variable remains
7323 empty, and $host_lookup_failed is set to "1".
7325 19. Added "eqi" as a case-independent comparison operator.
7327 20. The saslauthd authentication condition could segfault if neither service
7328 nor realm was specified.
7330 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7331 error message that was logged was misleading, and incoming SMTP
7332 connections were dropped. The message is now more accurate, and temporary
7333 errors are given to SMTP connections.
7335 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7336 (except RSET) with a 421 error, until QUIT is received. However, it was
7337 failing to send a response to QUIT.
7339 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7340 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7341 ACL is now run after the helo_try_verify_hosts code.
7343 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7344 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7345 case-independent, so other case variants are also recognized). Apparently
7346 some systems use these upper case variants.
7348 25. If more than two messages were waiting for the same host, and a transport
7349 filter was specified for the transport, Exim sent two messages over the
7350 same TCP/IP connection, and then failed with "socket operation on non-
7351 socket" when it tried to send the third.
7353 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7355 27. The extern definition of crypt16() in expand.c was not being excluded when
7356 the OS had its own crypt16() function.
7358 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7359 as a preferred synonym for return_size_limit, both as an option and as an
7362 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7364 30. Changed the default configuration ACL to relax the local part checking rule
7365 for addresses that are not in any local domains. For these addresses,
7366 slashes and pipe symbols are allowed within local parts, but the sequence
7367 /../ is explicitly forbidden.
7369 31. SPA server authentication was not clearing the challenge buffer before
7372 32. log_message in a "warn" ACL statement was writing to the reject log as
7373 well as to the main log, which contradicts the documentation and doesn't
7374 seem right (because no rejection is happening). So I have stopped it.
7376 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7377 However, I am unable to do any testing of this.
7379 34. Fixed an infelicity in the appendfile transport. When checking directories
7380 for a mailbox, to see if any needed to be created, it was accidentally
7381 using path names with one or more superfluous leading slashes; tracing
7382 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7384 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7385 contents are added to the log line that is written for every discarded
7386 recipient. (Previously a log_message setting was ignored.)
7388 36. The ${quote: operator now quotes the string if it is empty.
7390 37. The install script runs exim in order to find its version number. If for
7391 some reason other than non-existence or emptiness, which it checks, it
7392 could not run './exim', it was installing it with an empty version number,
7393 i.e. as "exim-". This error state is now caught, and the installation is
7396 38. An argument was missing from the function that creates an error message
7397 when Exim fails to connect to the socket for saslauthd authentication.
7398 This could cause Exim to crash, or give a corrupted message.
7400 39. Added isip, isip4, and isip6 to ${if conditions.
7402 40. The ACL variables $acl_xx are now saved with the message, and can be
7403 accessed later in routers, transports, and filters.
7405 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7406 strings in the file are not string-expanded.
7408 42. If a MAIL command specified a SIZE value that was too large to fit into an
7409 int variable, the check against message_size_limit failed. Such values are
7410 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7411 day this will have to be increased, but I don't think I want to be around
7412 when emails are that large.
7419 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7420 iconv() is not standard in FreeBSD.
7422 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7423 IPv6 enabled. The observed symptom was a segmentation fault on return from
7424 the function os_common_find_running_interfaces() in src/os.c.
7426 3. In the check_special_case() function in daemon.c I had used "errno" as an
7427 argument name, which causes warnings on some systems. This was basically a
7428 typo, since it was named "eno" in the comments!
7430 4. The code that waits for the clock to tick (at a resolution of some fraction
7431 of a second) so as to ensure message-id uniqueness was always waiting for
7432 at least one whole tick, when it could have waited for less. [This is
7433 almost certainly not relevant at current processor speeds, where it is
7434 unlikely to ever wait at all. But we try to future-proof.]
7436 5. The function that sleeps for a time interval that includes fractions of a
7437 second contained a race. It did not block SIGALRM between setting the
7438 timer, and suspending (a couple of lines later). If the interval was short
7439 and the sigsuspend() was delayed until after it had expired, the suspension
7440 never ended. On busy systems this could lead to processes getting stuck for
7443 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7444 process, before it forks any delivery processes. The open lookup caching
7445 mechanism meant that the open file or database connection was passed into
7446 the delivery process. The problem was that delivery processes always tidy
7447 up cached lookup data. This could cause a problem for the next delivery
7448 process started by the queue runner, because the external queue runner
7449 process does not know about the closure. So the next delivery process
7450 still has data in the lookup cache. In the case of a file lookup, there was
7451 no problem because closing a file descriptor in a subprocess doesn't affect
7452 the parent. However, if the lookup was caching a connection to a database,
7453 the connection was closed, and the second delivery process was likely to
7454 see errors such as "PGSQL: query failed: server closed the connection
7455 unexpectedly". The problem has been fixed by closing all cached lookups
7456 in a queue runner before running a delivery process.
7458 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7459 seem to have the "const" qualifier which it has on other OS. I've
7462 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7463 *of the same type* (client or server) with the same public name that an
7464 error should be diagnosed.
7466 9. When Exim looked up a host name for an IP address, but failed to find the
7467 original IP address when looking up the host name (a safety check), it
7468 output the message "<ip address> does not match any IP for NULL", which was
7469 confusing, to say the least. The bug was that the host name should have
7470 appeared instead of "NULL".
7472 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7473 user that is built into the binary, and the -C or -D options is used, root
7474 privilege is dropped before the configuration file is read. In addition,
7475 logging is switched to stderr instead of the normal log files. If the
7476 configuration then re-defines the Exim user, the unprivileged environment
7477 is probably not what is expected, so Exim logs a panic warning message (but
7480 However, if deliver_drop_privilege is set, the unprivileged state may well
7481 be exactly what is intended, so the warning has been cut out in that case,
7482 and Exim is allowed to try to write to its normal log files.
7488 1. smtp_return_error_details was not giving details for temporary sender
7489 or receiver verification errors.
7491 2. Diagnose a configuration error if two authenticators have the same public
7494 3. Exim used not to create the message log file for a message until the first
7495 delivery attempt. This could be confusing when incoming messages were held
7496 for policy or load reasons. The message log file is now created at the time
7497 the message is received, and an initial "Received" line is written to it.
7499 4. The automatically generated man page for command line options had a minor
7500 bug that caused no ill effects; however, a more serious problem was that
7501 the procedure for building the man page automatically didn't always
7502 operate. Consequently, release 4.20 contains an out-of-date version. This
7503 shouldn't happen again.
7505 5. When building Exim with embedded Perl support, the script that builds the
7506 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7507 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7509 6. The freeze_tell option was not being used for messages that were frozen on
7510 arrival, either by an ACL or by local_scan().
7512 7. Added the smtp_incomplete_transaction log selector.
7514 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7515 was accepting AUTH without a new EHLO.
7517 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7518 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7521 10. Logging of TCP/IP connections (when configured) now happens in the main
7522 daemon process instead of the child process, so that the TCP/IP connection
7523 count is more accurate (but it can never be perfect).
7525 11. The use of "drop" in a nested ACL was not being handled correctly in the
7526 outer ACL. Now, if condition failure induced by the nested "drop" causes
7527 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7528 or "require"), the connection is dropped.
7530 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7531 that yield "discard" can now be used with an "accept" or a "discard" verb,
7532 but an error is generated for any others (because I can't see a useful way
7533 to define what should happen).
7535 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7536 are now processed in the same way as lines in the Exim configuration file.
7537 In particular, continuation lines are supported.
7539 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7541 15. Added -ti meaning -t -i.
7543 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7544 domains, and warn by logging if others are found.
7546 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7547 returned from GIFCONF when Exim is trying to find the list of interfaces on
7548 a host. The code in os.c has been modified to copy each ifreq to an aligned
7549 structure in all cases.
7551 Also, in some cases, the returned ifreq's were being copied to a 'struct
7552 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7553 means the last couple of bytes of an IPv6 address could be chopped if the
7554 ifreq contained only a normal sockaddr (14 bytes storage).
7556 18. Named domain lists were not supported in the hosts_treat_as_local option.
7557 An entry such as +xxxx was not recognized, and was treated as a literal
7560 19. Ensure that header lines added by a DATA ACL are included in the reject log
7561 if the ACL subsequently rejects the message.
7563 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7564 MD5 (which is deprecated).
7566 21. When testing a filter file using -bf, Exim was writing a message when it
7567 took the sender from a "From " line in the message, but it was not doing so
7568 when it took $return_path from a Return-Path: header line. It now does.
7570 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7571 with a valid header line field name (a series of printing characters
7572 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7574 23. Changed "disc" in the source to "disk" to conform to the documentation and
7575 the book and for uniformity.
7577 24. Ignore Sendmail's -Ooption=value command line item.
7579 25. When execve() failed while trying to run a command in a pipe transport,
7580 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7581 could be confused with a return value of 69 from the command itself. This
7582 has been changed to 127, the value the shell returns if it is asked to run
7583 a non-existent command. The wording for the related log line suggests a
7584 non-existent command as the problem.
7586 26. If received_header_text expands to an empty string, do not add a Received:
7587 header line to the message. (Well, it adds a token one on the spool, but
7588 marks it "old" so that it doesn't get used or transmitted.)
7590 27. Installed eximstats 1.28 (addition of -nt option).
7592 28. There was no check for failure on the call to getsockname() in the daemon
7593 code. This can fail if there is a shortage of resources on the system, with
7594 ENOMEM, for example. A temporary error is now given on failure.
7596 29. Contrary to the C standard, it seems that in some environments, the
7597 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7598 program. Exim now does this explicitly; it affects the formatting of
7599 timestamps using strftime().
7601 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7602 complaints. I've now initialized all the variables, to avoid this.
7604 32. Header lines added by a system filter were not being "seen" during
7605 transport-time rewrites.
7607 33. The info_callback() function passed to OpenSSL is set up with type void
7608 (*)(SSL *, int, int), as described somewhere. However, when calling the
7609 function (actually a macro) that sets it up, the type void(*)() is
7610 expected. I've put in a cast to prevent warnings from picky compilers.
7612 34. If a DNS black list lookup found a CNAME record, but there were no A
7613 records associated with the domain it pointed at, Exim crashed.
7615 35. If a DNS black list lookup returned more than one A record, Exim ignored
7616 all but the first. It now scans all returned addresses if a particular IP
7617 value is being sought. In this situation, the contents of the
7618 $dnslist_value variable are a list of all the addresses, separated by a
7621 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7622 to accept a host name and all its aliases if the forward lookup for any of
7623 them yielded the IP address of the incoming connection. Now it accepts only
7624 those names whose forward lookup yields the correct IP address. Any other
7625 names are discarded. This closes a loophole whereby a rogue DNS
7626 administrator could create reverse DNS records to break through a
7627 wildcarded host restriction in an ACL.
7629 37. If a user filter or a system filter that ran in a subprocess used any of
7630 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7631 the wrong values were passed to the pipe command ($thisaddress had the
7632 value of $0, $0 had the value of $1, etc). This bug was introduced by
7633 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7635 38. Improved the line breaking for long SMTP error messages from ACLs.
7636 Previously, if there was no break point between 40 and 75 characters, Exim
7637 left the rest of the message alone. Two changes have been made: (a) I've
7638 reduced the minimum length to 35 characters; (b) if it can't find a break
7639 point between 35 and 75 characters, it looks ahead and uses the first one
7640 that it finds. This may give the occasional overlong line, but at least the
7641 remaining text gets split now.
7643 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7644 file descriptors might be low, and that setting 1000 would always raise it.
7645 It turns out that in some environments, the limit is already over 1000 and
7646 that lowering it causes trouble. So now Exim takes care not to decrease it.
7648 40. When delivering a message, the value of $return_path is set to $sender_
7649 address at the start of routing (routers may change the value). By an
7650 oversight, this default was not being set up when an address was tested by
7651 -bt or -bv, which affected the outcome if any router or filter referred to
7654 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7655 log only when "message" or "log_message" are set. However, if one of the
7656 conditions was an address verification, or a call to a nested ACL, the
7657 messages generated by the underlying test were being passed through. This
7658 no longer happens. The underlying message is available in $acl_verify_
7659 message for both "message" and "log_message" expansions, so it can be
7660 passed through if needed.
7662 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7663 new expansion $bh_ to give the encoded byte string without charset
7664 translation. Translation happens only if iconv() is available; HAVE_ICONV
7665 indicates this at build time. HEADERS_CHARSET gives the charset to
7666 translate to; headers_charset can change it in the configuration, and
7667 "headers charset" can change it in an individual filter file.
7669 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7670 that creates RFC 2047 encoded "words" labels them as that charset instead
7671 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7672 expansion operator; (ii) when Exim creates a From: line for a local
7673 message; (iii) when a header line is rewritten to include a "phrase" part.
7675 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7676 buggy, causing it to skip the first lines of messages whose message ID
7677 ended in 'D'. This would not have bitten before Exim release 4.14, because
7678 message IDs were unlikely to end in 'D' before then. The effect was to have
7679 incorrect size information for certain domains.
7681 45. #include "config.h" was missing at the start of the crypt16.c module. This
7682 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7685 46. If there was a timeout during a "random" callout check, Exim treated it as
7686 a failure of the random address, and carried on sending RSET and the real
7687 address. If the delay was just some slowness somewhere, the response to the
7688 original RCPT would be taken as a response to RSET and so on, causing
7689 mayhem of various kinds.
7691 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7692 when I implemented it. It didn't allow for the fact that some option values
7693 may legitimately be negative (e.g. size_addition), and it didn't even do
7694 the right test for positive values.
7696 48. Domain names in DNS records are case-independent. Exim always looks them up
7697 in lower case. Some resolvers return domain names in exactly the case they
7698 appear in the zone file, that is, they may contain uppercase letters. Not
7699 all resolvers do this - some return always lower case. Exim was treating a
7700 change of case by a resolver as a change of domain, similar to a widening
7701 of a domain abbreviation. This triggered its re-routing code and so it was
7702 trying to route what was effectively the same domain again. This normally
7703 caused routing to fail (because the router wouldn't handle the domain
7704 twice). Now Exim checks for this case specially, and just changes the
7705 casing of the domain that it ultimately uses when it transmits the message
7708 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7711 50. If a filter generated a file delivery with a non-absolute name (possible if
7712 no home directory exists for the router), the forbid_file option was not
7715 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7716 the existing equality matching.
7718 52. Exim was using ints instead of ino_t variables in some places where it was
7719 dealing with inode numbers.
7721 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7722 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7723 called TMPDIR, and if it finds it is different, it changes its value.
7725 54. The smtp_printf() function is now made available to local_scan() so
7726 additional output lines can be written before returning. There is also an
7727 smtp_fflush() function to enable the detection of a dropped connection.
7728 The variables smtp_input and smtp_batched_input are exported to
7731 55. Changed the default runtime configuration: the message "Unknown user"
7732 has been removed from the ACL, and instead placed on the localuser router,
7733 using the cannot_route_message feature. This means that any verification
7734 failures that generate their own messages won't get overridden. Similarly,
7735 the "Unrouteable address" message that was in the ACL for unverifiable
7736 relay addresses has also been removed.
7738 56. Added hosts_avoid_esmtp to the smtp transport.
7740 57. The exicyclog script was not checking for the esoteric option
7741 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7742 will work only if exicyclog is run under the appropriate euid.
7744 58. Following a discussion on the list, the rules by which Exim recognises line
7745 endings on incoming messages have been changed. The -dropcr and drop_cr
7746 options are now no-ops, retained only for backwards compatibility. The
7747 following line terminators are recognized: LF CRLF CR. However, special
7748 processing applies to CR:
7750 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7751 nor a local message in the state where . is a terminator.
7753 (ii) If a bare CR is encountered in a header line, an extra space is added
7754 after the line terminator so as not to end the header. The reasoning
7755 behind this is that bare CRs in header lines are most likely either
7756 to be mistakes, or people trying to play silly games.
7758 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7759 was being incorrectly given as 18 bytes larger than it should have been.
7760 This is a VOB (very old bug).
7762 60. This may never have affected anything current, but just in case it has:
7763 When the local host is found other than at the start of a list of hosts,
7764 the local host, those with the same MX, and any that follow, are discarded.
7765 When the list in question was part of a longer list of hosts, the following
7766 hosts (not currently being processed) were also being discarded. This no
7767 longer happens. I'm not sure if this situation could ever has previously
7770 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7773 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7774 header. This is something that is recommended in a new Internet Draft, and
7775 is something that is documented as being done by Sendmail. There are two
7776 possible values. For messages generated by the autoreply transport, Exim
7779 Auto-Submitted: auto-replied
7781 whereas for all other generated messages (e.g. bounces) it adds
7783 Auto-Submitted: auto-generated
7785 63. The "personal" condition in filters now includes a test for the
7786 Auto-Submitted: header. If it contains the string "auto-" the message it
7787 not considered personal.
7789 64. Added rcpt_include_affixes as a generic transport option.
7791 65. Added queue_only_override (default true).
7793 66. Added the syslog_duplication option.
7795 67. If what should have been the first header line of a message consisted of
7796 a space followed by a colon, Exim was mis-interpreting it as a header line.
7797 It isn't of course - it is syntactically invalid and should therefore be
7798 treated as the start of the message body. The misbehaviour could have
7799 caused a number of strange effects, including loss of data in subsequent
7800 header lines, and spool format errors.
7802 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7803 client host had authenticated. This control can now be exercised by an ACL
7804 for more flexibility.
7806 69. By default, callouts do not happen when testing with -bh. There is now a
7807 variant, -bhc, which does actually run the callout code, including
7808 consulting and updating the callout cache.
7810 70. Added support for saslauthd authentication, courtesy of Alexander
7813 71. If statvfs() failed on the spool or log directories while checking their
7814 size for availability, Exim confusingly gave the error "space shortage".
7815 Furthermore, in debugging mode it crashed with a floating point exception.
7816 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7817 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7818 serious problem, Exim now writes to the main and panic logs when this
7819 happens, with details of the failure. It then refuses to accept the
7820 incoming message, giving the message "spool directory problem" or "log
7821 directory problem" with a 421 code for SMTP messages.
7823 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7824 0, 1, and 2 exist, because some OS complain for execs without them (see
7825 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7826 descriptors. However, the code omitted to check that the open succeeded,
7827 causing mysterious errors if for some reason the permissions on /dev/null
7828 got screwed. Now Exim writes a message to the main and panic logs, and
7829 bombs out if it can't open /dev/null.
7831 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7832 interact so that it is all more flexible. It is supposed to remain
7833 backwards compatible. Also added extra_local_interfaces.
7835 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7836 to bomb out with an assertion failure - to the client this appears as a
7837 connection drop. This problem occurs in the part of the code that was taken
7838 from the Samba project. Fortunately, the assertion is in a very simple
7839 function, so I have fixed this by reproducing the function inline in the
7840 one place where it is called, and arranging for authentication to fail
7841 instead of killing the process with assert().
7843 75. The SPA client code was not working when the server requested OEM rather
7844 than Unicode encoding.
7846 76. Added code to make require_files with a specific uid setting more usable in
7847 the case where statting the file as root fails - usually a non-root-mounted
7848 NFS file system. When this happens and the failure is EACCES, Exim now
7849 forks a subprocess and does the per-uid checking as the relevant uid.
7851 77. Added process_log_path.
7853 78. If log_file_path was not explicitly set, a setting of check_log_space or
7854 check_log_inodes was ignored.
7856 79. If a space check for the spool or log partitions fails, the incident is now
7857 logged. Of course, in the latter case the data may get lost...
7859 80. Added the %p formatting code to string_format() so that it can be used to
7860 print addresses in debug_print(). Adjusted all the address printing in the
7861 debugging in store.c to use %p rather than %d.
7863 81. There was a concern that a line of code in smtp_in.c could overflow a
7864 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7865 initially expressed, the concern was not well-founded, because trailing
7866 spaces are removed early. However, if the trailing spaces were followed by
7867 a NULL, they did not get removed, so the overflow was possible. Two fixes
7870 (a) I re-wrote the offending code in a cleaner fashion.
7871 (b) If an incoming SMTP command contains a NULL character, it is rejected
7874 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7875 runs initgroups(), so that if the Exim user is in any additional groups,
7876 they will be used during message reception.
7882 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob