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 Avoid costly startup code when not strictly needed. This reduces time
10 for some exim process initialisations. It does mean that the logging
11 of TLS configuration problems is only done for the daemon startup.
13 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
15 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
16 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
19 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
20 DKIM signing is being done. Previously a transport_filter would always
21 disable CHUNKING, falling back to traditional DATA.
23 JH/05 Regard command-line receipients as tainted.
25 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
27 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
28 PAM library frees one of the arguments given to it, despite the
29 documentation. Therefore a plain malloc must be used.
31 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
32 on-stack buffers were used, resulting in a taint trap when DSN information
33 copied from a received message was written into the buffer.
35 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
36 the ordering of its ARC headers. This caused a crash.
38 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
39 a new record was being constructed with information from the peer, a trap
42 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
43 installation would get error messages from DMARC verify, when it hit the
44 nonexistent file indicated by the default. Distros wanting DMARC enabled
45 should both provide the file and set the option.
46 Also enforce no DMARC verification for command-line sourced messages.
48 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
49 could, depending on the platform, hang at the STARTTLS response.
51 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
52 message on a connection. Previously if one message had ARC headers and
53 the following one did not, a crash could result when adding an
54 Authentication-Results: header.
56 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
57 Exim main code and Exim-related utities. The introduction of taint
58 tracking also did many adjustments to string handling. Since then, eximon
59 frequently terminated with an assert failure.
61 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
62 check for 452 responses. This slightly helps the inefficieny of doing
63 a large alias-expansion into a recipient-limited target. The max_rcpt
64 transport option still applies (and at the current default, will override
65 the new feature). The check is done for either cause of synch, and forces
66 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
67 connection. The new facility is not tunable at this time.
69 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
70 library live data was being used, so the results became garbage. Make
71 copies while it is still usable.
73 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
74 on delivery deferred (==) and failed (**) lines (if a delivery was
75 attemtped). Previously it was only on completion (=>) lines.
77 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
78 for the expansion of the server_scram_iter and server_scram_salt options.
80 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
81 are now specifically given a NO_DATA response without hitting the system
82 resolver. The library goes on to do the now-standard TXT lookup.
83 Use of dnsdb lookups is not affected.
85 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
86 only retrieve the errormessage once. Previously two calls to dlerror()
87 were used, and the second one (for mainlog/paniclog) retrieved null
90 JH/20 Taint checking: disallow use of tainted data for
91 - the appendfile transport file and directory options
92 - the pipe transport command
93 - the autoreply transport file, log and once options
94 - file names used by the redirect router (including filter files)
96 - paths used by single-key lookups
97 Previously this was permitted.
99 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
100 adjusted the size of a major service buffer; this failed because the
101 buffer was in use at the time. Change to a compile-time increase in the
102 buffer size, when this authenticator is compiled into exim.
104 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
105 previous fast-mode was untenable in the face of glibs using mmap to
106 support larger malloc requests.
108 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
109 New values supported, if defined on system where compiled:
110 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
111 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
113 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
114 running a limited number of proceses in parallel, a benefit is gained. The
115 amount varies with the platform hardware and load. The use of the option
116 queue_run_in_order means we cannot do this, as ordering becomes
119 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
120 had introduced a string-copy (for ensuring NUL-termination) which was not
121 appropriate for that case, which can include embedded NUL bytes in the
122 block of data. Investigation showed the copy to actually be needless, the
123 data being length-specified.
125 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
126 done during a receiving connection, and both used TLS, global info was
127 used rather than per-connection info for tracking the state of data
128 queued for transmission. This could result in a connection hang.
130 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
131 Previously, when delivering serveral messages down a single connection
132 only the first would provide a SIZE. This was due to the size information
133 not being properly tracked.
135 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
136 TAI (at 37 seconds currently), pretend to be in UTC for time-related
137 expansion and logging. Previously, spurious values such as a future
138 minute could be seen.
140 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
141 it could crash from a null-deref. This could also affect the
142 ${addresses: } operator and ${readsock } item.
144 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
145 message following a mime one, the variable was not reset.
147 JH/30 When an pipelined-connect fails at the first response, assume incorrect
148 cached capability (perhaps the peer reneged?) and immediately retry in
151 JH/31 Fix spurious detection of timeout while writing to transport filter.
153 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
154 an attempt to copy the string was made before checking it.
156 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
157 taint of the lookup key was maintained; we now regard the presence in the
158 filesystem as sufficient validation.
160 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
161 argument is supplied.
163 JH/35 The dsearch lookup now requires that the directory is an absolute path.
164 Previously this was not checked, and nonempty relative paths made an
165 access under Exim's current working directory.
167 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
168 Previously no event was raised.
170 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
171 parameter supplied by the sender MAIL FROM command. Previously it was
172 ignored, and only the check_spool_space option value for the required
175 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
176 the size of the signing public-key. Previously it was instead giving
177 the size of the signature hash.
179 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
180 the default. See the (new) dkim_verify_min_keysizes option.
186 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
187 side, to decode a TLS 1.3 packet capture.
189 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
190 Previously the default library behaviour applied, sending two, each in
193 JH/03 Debug output for ACL now gives the config file name and line number for
196 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
198 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
200 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
201 buffer overrun for (non-chunking) other transports.
203 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
204 TLS1.3, means that a server rejecting a client certificate is not visible
205 to the client until the first read of encrypted data (typically the
206 response to EHLO). Add detection for that case and treat it as a failed
207 TLS connection attempt, so that the normal retry-in-clear can work (if
208 suitably configured).
210 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
211 and/or domain. Found and fixed by Jason Betts.
213 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
214 configuration). If a CNAME target was not a wellformed name pattern, a
217 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
218 the OS reports them interleaved with other addresses.
220 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
221 used both for input and for a verify callout, both encrypted, SMTP
222 responses being sent by the server could be lost. This resulted in
223 dropped connections and sometimes bounces generated by a peer sending
226 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
227 string. Previously it was possible to cause undefined behaviour in a
228 library routine (usually a crash). Found by "zerons".
230 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
233 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
234 API was removed, so update to use the newer ones.
236 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
237 any timeout set, is taking a long time. Previously we would hang on to a
238 rotated logfile "forever" if the input was arriving with long gaps
239 (a previous attempt to fix addressed lack, for a long time, of initial
242 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
243 shared (NFS) environment. The length of the tempfile name is now
244 4 + 16 ("hdr.$message_exim_id") which might break on file
245 systems which restrict the file name length to lower values.
248 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
249 shared (NFS) environment.
251 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
252 did for all versions <4.90). Notably -M, -m, --invert, -I may be
255 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
256 on some platforms for bit 31.
258 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
259 to changes apparently associated with TLS1.3 handling some of the APIs
260 previously used were either nonfunctional or inappropriate. Strings
261 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
262 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
263 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
264 This affects log line X= elements, the $tls_{in,out}_cipher variables,
265 and the use of specific cipher names in the encrypted= ACL condition.
267 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
269 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
270 verification result was not updated unless hosts_require_ocsp applied.
272 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
273 queue_list_requires_admin set to false, non-admin users were denied the
276 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
277 directory-of-certs mode. Previously they were advertised despite the
280 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
281 A single TCP connection by a client will now hold a TLS connection open
282 for multiple message deliveries, by default. Previoud the default was to
285 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
286 default. If built with the facility, DANE will be used. The facility
287 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
289 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
290 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
291 must be defined and you must still, unless you define DISABLE_TLS, manage
292 the the include-dir and library-file requirements that go with that
293 choice. Non-TLS builds are still supported.
295 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
298 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
299 default. If the platform supports and has the facility enabled, it will
300 be requested on all coneections.
302 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
303 controlled by the build-time option SUPPORT_PIPE_CONNECT.
305 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
307 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
308 success-DSN messages. Previously the From: header was always the default
309 one for these; the option was ignored.
311 JH/28 Fix the timeout on smtp response to apply to the whole response.
312 Previously it was reset for every read, so a teergrubing peer sending
313 single bytes within the time limit could extend the connection for a
314 long time. Credit to Qualsys Security Advisory Team for the discovery.
316 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
317 delivery address, which leaked information of the results of local
318 forwarding. Change to the original envelope recipient address, per
321 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
322 requested. Previously not bounce was generated and a log entry of
323 error ignored was made.
325 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
327 JH/32 Introduce a general tainting mechanism for values read from the input
328 channel, and values derived from them. Refuse to expand any tainted
329 values, to catch one form of exploit.
331 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
332 was unused and the unexpanded text used for the test. Found and
333 fixed by Ruben Jenster.
335 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
336 an attempt to use a TLS library read routine dereffed a nul pointer,
339 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
340 them in our smtp response.
342 JH/36 Have the general router option retry_use_local_part default to true when
343 any of the restrictive preconditions are set (to anything). Previously it
344 was only for check_local user. The change removes one item of manual
345 configuration which is required for proper retries when a remote router
346 handles a subset of addresses for a domain.
348 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
349 link count into consideration.
351 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
352 caused the extension of big_buffer, the following lines were ignored.
354 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
355 accordance with RFC 2308. Previously there was no expiry, so a longlived
356 receive process (eg. due to ACL delays) versus a short SOA value could
359 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
361 JH/39 Promote DMARC support to mainline.
363 JH/40 Bug 2452: Add a References: header to DSNs.
365 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
366 parameters. The relevant library call is documented as "Deprecated: This
367 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
368 3.6.0, DH parameters are negotiated following RFC7919."
370 HS/06 Change the default of dnssec_request_domains to "*"
372 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
373 carried on and emitted a BDAT command, even when PIPELINING was not
376 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
377 buffer was used for the filename, resulting in a trap when tainted
378 arguments (eg. $domain) were used.
380 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
381 recommended to avoid a possible server-load attack. The feature can be
382 re-enabled via the openssl_options main cofiguration option.
384 JH/45 local_scan API: documented the current smtp_printf() call. This changed
385 for version 4.90 - adding a "more data" boolean to the arguments.
386 Bumped the ABI version number also, this having been missed previously;
387 release versions 4.90 to 4.92.3 inclusive were effectively broken in
388 respect of usage of smtp_printf() by either local_scan code or libraries
389 accessed via the ${dlfunc } expansion item. Both will need coding
390 adjustment for any calls to smtp_printf() to match the new function
391 signature; a FALSE value for the new argument is always safe.
393 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
394 the file-offset (which the Linux syscall does, and exim expects); this
395 resulted in an indefinite loop.
397 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
398 to do ARC verification. The Authentication-Results: header line added
399 by the configuration then had no ARC item.
405 JH/01 Remove code calling the customisable local_scan function, unless a new
406 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
408 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
409 non-signal-safe functions being used.
411 JH/03 Bug 2269: When presented with a received message having a stupidly large
412 number of DKIM-Signature headers, disable DKIM verification to avoid
413 a resource-consumption attack. The limit is set at twenty.
415 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
416 report of oldest_pass in ${authres } in consequence, and separate out
417 some descriptions of reasons for verification fail.
419 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
420 files in the spool were present and unlocked. A queue-runner could spot
421 them, resulting in a duplicate delivery. Fix that by doing the unlock
422 after the unlink. Investigation by Tim Stewart. Take the opportunity to
423 add more error-checking on spoolfile handling while that code is being
426 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
427 No known attacks, no CVE, this is defensive hardening.
429 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
430 a queue-runner could start a delivery while other operations were ongoing.
431 Cutthrough delivery was a common victim, resulting in duplicate delivery.
432 Found and investigated by Tim Stewart. Fix by using the open message data
433 file handle rather than opening another, and not locally closing it (which
434 releases a lock) for that case, while creating the temporary .eml format
435 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
437 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
438 $sender_verify_failure/$recipient_verify_failure to "random".
440 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
443 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
444 Previously this would segfault.
446 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
449 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
450 like zero, since the resolver should be doing this for us, But we need one
451 as a CNAME but no MX presence gets the CNAME returned; we need to check
452 that doesn't point to an MX to declare it "no MX returned" rather than
453 "error, loop". A new main option is added so the older capability of
454 following some limited number of chain links is maintained.
456 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
458 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
459 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
460 now-deprecated earlier definitions used only specified the range up to TLS
461 1.2 (in the older-version library docs).
463 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
465 JH/15 Rework TLS client-side context management. Stop using a global, and
466 explicitly pass a context around. This enables future use of TLS for
467 connections to service-daemons (eg. malware scanning) while a client smtp
468 connection is using TLS; with cutthrough connections this is quite likely.
470 JH/16 Fix ARC verification to do AS checks in reverse order.
472 JH/17 Support a "tls" option on the ${readsocket } expansion item.
474 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
475 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
476 Previously the "utf8" would be re-prepended for every additional message.
478 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
479 Previously thery were accepted, resulting in issues when attempting to
480 forward messages to a non-supporting MTA.
482 PP/02 Let -n work with printing macros too, not just options.
484 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
485 one parent address was copied, and bogus data was used at delivery-logging
486 time. Either a crash (after delivery) or bogus log data could result.
487 Discovery and analysis by Tim Stewart.
489 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
490 Previously if the string ended mid-character, we did not insert the
491 promised '?' replacement.
493 PP/04 Documentation: current string operators work on bytes, not codepoints.
495 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
496 should pack well giving a smaller memory footprint so better caching and
497 therefore performance. Group the declarations where this can't be done so
498 that the byte-sized flag variables are not interspersed among pointer
499 variables, giving a better chance of good packing by the compiler.
501 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
502 non-null, to avoid issues with sites running BATV. Previously reports were
503 sent with an empty envelope sender so looked like bounces.
505 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
506 The ignore_error flag wasn't being returned from the filter subprocess so
507 was not set for later routers. Investigation and fix by Matthias Kurz.
509 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
510 and a msg:complete for the whole, when a message is manually removed using
511 -Mrm. Developement by Matthias Kurz, hacked on by JH.
513 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
514 a "Gnu special" function, asprintf() in the DB utility binary builds; I
515 hope that is portable enough.
517 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
518 requiring a known-CA anchor certificate; make it now rely entirely on the
519 TLSA as an anchor. Checking the name on the leaf cert against the name
520 on the A-record for the host is still done for TA (but not for EE mode).
522 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
523 included in delivery lines for non-proxied connections, when compiled with
524 SUPPORT_SOCKS and running with proxy logging enabled.
526 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
527 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
528 move the existing event to fire before the normal logging of message
529 failure so that custom logging is bracketed by normal logging.
531 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
532 msg:fail:internal event. Developement by Matthias Kurz.
534 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
535 far too small for todays use of crypto signatures stored there. Go all
536 the way to the max DNS message size of 64kB, even though this might be
537 overmuch for IOT constrained device use.
539 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
540 copy a string over itself. The library routine is documented as not
541 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
543 JH/32 For main options check_spool_space and check_inode_space, where the
544 platform supports 64b integers, support more than the previous 2^31 kB
545 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
546 the previous G, M, k.
548 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
549 $authenticated_fail_id variable on authentication failure. Previously
552 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
553 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
554 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
555 GNUTLS_SEC_PARAM_MEDIUM.
557 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
558 side. Previously we would continue as if no SNI had been received.
560 JH/36 Harden the handling of string-lists. When a list consisted of a sole
561 "<" character, which should be a list-separator specification, we walked
562 off past the nul-terimation.
564 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
565 causes) even when the retry time is not yet met. Previously they were
566 not, meaning that when (say) an account was over-quota and temp-rejecting,
567 and multiple senders' messages were queued, only one sender would get
568 notified on each configured delay_warning cycle.
570 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
572 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
573 cipher-suites, an error can be left on the stack even for a succeeding
574 accept; this results in impossible error messages when a later operation
577 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
578 return error codes indicating retry. Under TLS1.3 this becomes required.
580 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
581 it only wrote the new authenticators, resulting in a lack of tracking of
582 peer changes of ESMTP extensions until the next cache flush.
584 JH/41 Fix the loop reading a message header line to check for integer overflow,
585 and more-often against header_maxsize. Previously a crafted message could
586 induce a crash of the recive process; now the message is cleanly rejected.
588 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
589 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
595 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
596 When redis_servers is set to a list of > 1 element, and the Redis servers
597 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
598 case of MOVED into a DEFER case instead, thus moving the query onto the
599 next server in the list. For a cluster of N elements, all N servers must
600 be defined in redis_servers.
602 GF/02 Catch and remove uninitialized value warning in exiqsumm
603 Check for existence of @ARGV before looking at $ARGV[0]
605 JH/01 Replace the store_release() internal interface with store_newblock(),
606 which internalises the check required to safely use the old one, plus
607 the allocate and data copy operations duplicated in both (!) of the
608 extant use locations.
610 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
611 modifier. This matches the restriction on the commandline.
613 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
614 Previously only the last row was returned.
616 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
617 we assumed that tags in the header were well-formed, and parsed the
618 element content after inspecting only the first char of the tag.
619 Assumptions at that stage could crash the receive process on malformed
622 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
623 While running the DKIM ACL we operate on the Permanent memory pool so that
624 variables created with "set" persist to the DATA ACL. Also (at any time)
625 DNS lookups that fail create cache records using the Permanent pool. But
626 expansions release any allocations made on the current pool - so a dnsdb
627 lookup expansion done in the DKIM ACL releases the memory used for the
628 DNS negative-cache, and bad things result. Solution is to switch to the
629 Main pool for expansions.
630 While we're in that code, add checks on the DNS cache during store_reset,
631 active in the testsuite.
632 Problem spotted, and debugging aided, by Wolfgang Breyha.
634 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
635 When none of the hosts presented to a transport match an already-open
636 connection, close it and proceed with the list. Previously we would
637 queue the message. Spotted by Lena with Yahoo, probably involving
640 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
641 Previously a spurious "250 OK id=" response was appended to the proper
644 JH/08 The "support for" informational output now, which built with Content
645 Scanning support, has a line for the malware scanner interfaces compiled
646 in. Interface can be individually included or not at build time.
648 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
649 by the template makefile "src/EDITME". The "STREAM" support for an older
650 ClamAV interface method is removed.
652 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
653 rows affected is given instead).
655 JH/11 The runtime Berkeley DB library version is now additionally output by
656 "exim -d -bV". Previously only the compile-time version was shown.
658 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
659 SMTP connection. Previously, when one had more receipients than the
660 first, an abortive onward connection was made. Move to full support for
661 multiple onward connections in sequence, handling cutthrough connection
662 for all multi-message initiating connections.
664 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
665 routers. Previously, a multi-recipient message would fail to match the
666 onward-connection opened for the first recipient, and cause its closure.
668 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
669 a timeout on read on a GnuTLS initiating connection, resulting in the
670 initiating connection being dropped. This mattered most when the callout
671 was marked defer_ok. Fix to keep the two timeout-detection methods
674 JH/15 Relax results from ACL control request to enable cutthrough, in
675 unsupported situations, from error to silently (except under debug)
676 ignoring. This covers use with PRDR, frozen messages, queue-only and
679 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
681 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
682 metadata, resulting in a crash in free().
684 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
685 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
686 Broken also in d185889f4, with init system revamp.
688 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
689 Previously we abruptly closed the connection after reading a malware-
690 found indication; now we go on to read the "scan ok" response line,
693 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
694 ACL. Previously, a crash would result.
696 JH/19 Speed up macro lookups during configuration file read, by skipping non-
697 macro text after a replacement (previously it was only once per line) and
698 by skipping builtin macros when searching for an uppercase lead character.
700 JH/20 DANE support moved from Experimental to mainline. The Makefile control
701 for the build is renamed.
703 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
704 was allocated for every new TLS startup, meaning one per message. Fix
705 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
707 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
708 reported the original. Fix to report (as far as possible) the ACL
709 result replacing the original.
711 JH/23 Fix memory leak during multi-message connections using STARTTLS under
712 OpenSSL. Certificate information is loaded for every new TLS startup,
713 and the resources needed to be freed.
715 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
717 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
720 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
721 DATA response info to the (existing) per-recipient response info for
722 the "C=" log element. It can have useful tracking info from the
723 destination system. Patch from Simon Arlott.
725 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
726 length value. Previously this would segfault.
728 HS/02 Support Avast multiline protoocol, this allows passing flags to
729 newer versions of the scanner.
731 JH/28 Ensure that variables possibly set during message acceptance are marked
732 dead before release of memory in the daemon loop. This stops complaints
733 about them when the debug_store option is enabled. Discovered specifically
734 for sender_rate_period, but applies to a whole set of variables.
735 Do the same for the queue-runner and queue-list loops, for variables set
736 from spool message files. Do the same for the SMTP per-message loop, for
737 certain variables indirectly set in ACL operations.
739 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
740 as a multi-recipient message from a mailinglist manager). The coding had
741 an arbitrary cutoff number of characters while checking for more input;
742 enforced by writing a NUL into the buffer. This corrupted long / fast
743 input. The problem was exposed more widely when more pipelineing of SMTP
744 responses was introduced, and one Exim system was feeding another.
745 The symptom is log complaints of SMTP syntax error (NUL chars) on the
746 receiving system, and refused recipients seen by the sending system
747 (propating to people being dropped from mailing lists).
748 Discovered and pinpointed by David Carter.
750 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
751 replaced by the ${authresults } expansion.
753 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
755 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
756 allows proper process termination in container environments.
758 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
759 Previously the "final dot" had a newline after it; ensure it is CR,LF.
761 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
762 and "err_perm", deprecated since 4.83 when the RFC-defined words
763 "temperror" and "permerror" were introduced.
765 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
766 transport-filters or DKIM-signing. The restriction was lost in the
767 consolidation of verify-callout and delivery SMTP handling.
768 Extend the restriction to also cover ARC-signing.
770 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
771 in defer=pass mode supply a 450 to the initiator. Previously the message
774 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
775 tls_require_ciphers is used as before.
777 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
778 "pass_unscanned". Only tmpfails from the scanner are written to
779 the paniclog, as they may require admin intervention (permission
780 denied, license issues). Other scanner errors (like decompression
781 bombs) do not cause a paniclog entry.
783 JH/36 Fix reinitialisation of DKIM logging variable between messages.
784 Previously it was possible to log spurious information in receive log
787 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
788 triggered odd behaviour from Outlook Express clients.
790 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
793 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
794 since the IETF WG has not yet settled on that versus the original
795 "bare" representation.
797 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
798 Previously the millisecond value corrupted the output.
799 Fix also for syslog_pid=no and log_selector +pid, for which the pid
800 corrupted the output.
806 JH/01 Rework error string handling in TLS interface so that the caller in
807 more cases is responsible for logging. This permits library-sourced
808 string to be attached to addresses during delivery, and collapses
809 pairs of long lines into single ones.
811 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
812 during configuration. Wildcards are allowed and expanded.
814 JH/02 Rework error string handling in DKIM to pass more info back to callers.
815 This permits better logging.
817 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
818 do not close it down and have the child transport start it up again on
819 the passed-on TCP connection. Instead, proxy the child (and any
820 subsequent ones) for TLS via a unix-domain socket channel. Logging is
821 affected: the continued delivery log lines do not have any DNSSEC, TLS
822 Certificate or OCSP information. TLS cipher information is still logged.
824 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
825 identical IP addresses on different listening ports. Will also affect
828 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
829 add noisy ifdef guards to special-case this sillyness.
830 Patch from Bernd Kuhls.
832 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
833 than 255 are no longer allowed.
835 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
836 Disable the session-cache too, which might reduce our load. Since we
837 currrectly use a new context for every connection, both as server and
838 client, there is no benefit for these.
839 GnuTLS appears to not support tickets server-side by default (we don't
840 call gnutls_session_ticket_enable_server()) but client side is enabled
841 by default on recent versions (3.1.3 +) unless the PFS priority string
844 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
845 <https://reproducible-builds.org/specs/source-date-epoch/>.
847 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
848 the check for any unsuccessful recipients did not notice the limit, and
849 erroneously found still-pending ones.
851 JH/08 Pipeline CHUNKING command and data together, on kernels that support
852 MSG_MORE. Only in-clear (not on TLS connections).
854 JH/09 Avoid using a temporary file during transport using dkim. Unless a
855 transport-filter is involved we can buffer the headers in memory for
856 creating the signature, and read the spool data file once for the
857 signature and again for transmission.
859 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
860 4.77 as the kernel support then wasn't solid, having issues in 64bit
861 mode. Now, it's been long enough. Add support for FreeBSD also.
863 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
864 case where the routing stage had gathered several addresses to send to
865 a host before calling the transport for the first, we previously failed
866 to close down TLS in the old transport process before passing the TCP
867 connection to the new process. The new one sent a STARTTLS command
868 which naturally failed, giving a failed delivery and bloating the retry
869 database. Investigation and fix prototype from Wolfgang Breyha.
871 JH/12 Fix check on SMTP command input synchronisation. Previously there were
872 false-negatives in the check that the sender had not preempted a response
873 or prompt from Exim (running as a server), due to that code's lack of
874 awareness of the SMTP input buffering.
876 PP/04 Add commandline_checks_require_admin option.
877 Exim drops privileges sanely, various checks such as -be aren't a
878 security problem, as long as you trust local users with access to their
879 own account. When invoked by services which pass untrusted data to
880 Exim, this might be an issue. Set this option in main configuration
881 AND make fixes to the calling application, such as using `--` to stop
884 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
885 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
886 responses to those, into a single TLS record each way (this usually means
887 a single packet). As a side issue, smtp_enforce_sync now works on TLS
890 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
891 affects you only if you're dancing at the edge of the param size limits.
892 If you are, and this message makes sense to you, then: raise the
893 configured limit or use OpenSSL 1.1. Nothing we can do for older
896 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
897 cmdline element to get the documented default one. Previously it was
900 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
901 are unsuable for non-message-specific reasons - eg. connection timeout,
902 banner-time rejection.
904 JH/16 Fix logging of delivery remote port, when specified by router, under
907 PP/06 Repair manualroute's ability to take options in any order, even if one
908 is the name of a transport.
911 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
913 JH/17 Change the list-building routines interface to use the expanding-string
914 triplet model, for better allocation and copying behaviour.
916 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
917 Previously it was constructed the first time a possibly-matching string
918 was met in the configuration file input during startup; now it is done
921 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
922 compatible one, to avoid the (poorly documented) possibility of a config
923 file in the working directory redirecting the DB files, possibly correpting
924 some existing file. CVE-2017-10140 assigned for BDB.
926 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
927 cache-hot. Previously, although the result was properly cached, the
928 initial verify call returned a defer.
930 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
931 the main verify for receipient in uncached-mode.
933 JH/22 Retire historical build files to an "unsupported" subdir. These are
934 defined as "ones for which we have no current evidence of testing".
936 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
937 if present. Previously it was ignored.
939 JH/24 Start using specified-initialisers in C structure init coding. This is
940 a C99 feature (it's 2017, so now considered safe).
942 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
943 if was a fixed-sized field and bitmask ops via macros; it is now more
946 PP/07 GitHub PR 56: Apply MariaDB build fix.
947 Patch provided by Jaroslav Škarvada.
949 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
950 during Coverity cleanups [4.87 JH/47]
951 Diagnosis and fix provided by Michael Fischer v. Mollard.
953 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
954 the right size to place the terminating semicolon on its own folded
955 line, the header hash was calculated to an incorrect value thanks to
956 the (relaxed) space the fold became.
958 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
959 and confused the parent.
961 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
962 which could crash as a result. This could lead to undeliverable messages.
964 JH/28 Logging: "next input sent too soon" now shows where input was truncated
967 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
968 matters on fast-turnover and PID-randomising systems, which were getting
969 out-of-order delivery.
971 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
972 a possibly-overlapping copy. The symptom was that "Remote host closed
973 connection in response to HELO" was logged instead of the actual 4xx
976 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
977 Previously only that bufferd was discarded, resulting in SYMTP command
980 JH/32 DKIM: when a message has multiple signatures matching an identity given
981 in dkim_verify_signers, run the dkim acl once for each. Previously only
982 one run was done. Bug 2189.
984 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
985 "panic the current process" to "deliberately defer". The panic log is
986 still written with the problem list name; the mail and reject logs now
987 get a temp-reject line for the message that was being handled, saying
988 something like "domains check lookup or other defer". The SMTP 451
989 message is still "Temporary local problem".
991 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
992 A crafted sequence of BDAT commands could result in in-use memory beeing
993 freed. CVE-2017-16943.
995 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
996 from SMTP input. Previously it was always done; now only done for DATA
997 and not BDAT commands. CVE-2017-16944.
999 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1000 to the message (such as an overlong header line). Previously this was
1001 not done and we did not exit BDAT mode. Followon from the previous item
1002 though a different problem.
1008 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1009 than -2003 did; needs libidn2 in addition to libidn.
1011 JH/02 The path option on a pipe transport is now expanded before use.
1013 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1014 Patch provided by "Björn", documentation fix added too.
1016 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1017 missing a wire-to-host endian conversion.
1019 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1020 close after a BDAT command line could be taken as a following command,
1021 giving a synch failure. Fix by only checking for synch immediately
1022 before acknowledging the chunk.
1024 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1025 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1026 macro. Patches provided by Josh Soref.
1028 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1029 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1030 by IANA as a service extension. However, John Klensin suggests that we
1033 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1034 the dkim code may be unix-mode line endings rather than smtp wire-format
1035 CRLF, so prepend a CR to any bare LF.
1037 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1038 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1040 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1041 the first were themselves being wrongly included in the feed into dkim
1042 processing; with most chunk sizes in use this resulted in an incorrect
1043 body hash calculated value.
1045 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1046 DKIM signature block, for verification. Although advised against by
1047 standards it is specifically not ruled illegal.
1049 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1051 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1052 missing a body hash (the bh= tag).
1054 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1055 It seems that HAProxy sends the Proxy Protocol information in clear and
1056 only then does a TLS startup, so do the same.
1058 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1059 TCP connections (such as for Spamd) unless the daemon successfully set
1060 Fast Open mode on its listening sockets. This fixes breakage seen on
1061 too-old kernels or those not configured for Fast Open, at the cost of
1062 requiring both directions being enabled for TFO, and TFO never being used
1063 by non-daemon-related Exim processes.
1065 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1066 endings, at least on the first header line. Try to canonify any that get
1067 past that check, despite the cost.
1069 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1070 now limited to an arbitrary five deep, while parsing addresses with the
1071 strip_excess_angle_brackets option enabled.
1073 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1074 instead leave the unprompted TLS handshake in socket buffer for the
1075 TLS library to consume.
1077 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1079 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1081 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1082 region in one operation, for speed, and this leaves assigned pointers
1083 dangling. Add checks run only under the testsuite which checks all
1084 variables at a store-reset and panics on a dangling pointer; add code
1085 explicitly nulling out all the variables discovered. Fixes one known
1086 bug: a transport crash, where a dangling pointer for $sending_ip_address
1087 originally assigned in a verify callout, is re-used.
1089 PP/06 Drop '.' from @INC in various Perl scripts.
1091 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1093 PP/08 Reduce a number of compilation warnings under clang; building with
1094 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1095 should be warning-free.
1097 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1099 HS/01 Fix portability problems introduced by PP/08 for platforms where
1100 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1102 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1103 chunk. This allows us to accept broken chunked messages. We need a more
1104 general solution here.
1106 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1107 already-broken messages in the queue.
1109 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1111 JH/19 Fix reference counting bug in routing-generated-address tracking.
1117 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1118 supports it and a size is available (ie. the sending peer gave us one).
1120 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1121 years of being deprecated). The replacements are the ACLs
1122 acl_smtp_mime and acl_not_smtp_mime.
1124 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1125 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1126 or even an in-clear connection were permitted. Now, if the host lookup
1127 was dnssec and dane was requested then the host is only used if the
1128 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1129 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1130 if one fails this test.
1131 This means that a poorly-configured remote DNS will make it incommunicado;
1132 but it protects against a DNS-interception attack on it.
1134 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1135 non-noisy when a race steals the message being considered.
1137 JH/05 If main configuration option tls_certificate is unset, generate a
1138 self-signed certificate for inbound TLS connections.
1140 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1141 in rewrites and routers.
1143 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1144 and logged a warning sing 4.83; now they are a configuration file error.
1146 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1147 (lacking @domain). Apply the same qualification processing as RCPT.
1149 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1151 JH/10 Support ${sha256:} applied to a string (as well as the previous
1154 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1155 a cutthrough deliver is pending, as we always want to make a connection.
1156 This also avoids re-routing the message when later placing the cutthrough
1157 connection after a verify cache hit.
1158 Do not update it with the verify result either.
1160 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1161 when routing results in more than one destination address.
1163 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1164 signing (which inhibits the cutthrough capability). Previously only
1165 the presence of an option was tested; now an expansion evaluating as
1166 empty is permissible (obviously it should depend only on data available
1167 when the cutthrough connection is made).
1169 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1170 the relevant preceding SMTP command did not note the pipelining mode.
1172 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1173 Previously they were not counted.
1175 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1176 as one having no matching records. Previously we deferred the message
1177 that needed the lookup.
1179 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1180 distinguished as "(=".
1182 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1183 for missing MX records. Previously it only worked for missing A records.
1185 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1187 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1188 after the data-go-ahead and data-ack. Patch from Jason Betts.
1190 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1191 even for a "none" policy. Patch from Tony Meyer.
1193 JH/22 Fix continued use of a connection for further deliveries. If a port was
1194 specified by a router, it must also match for the delivery to be
1197 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1198 When one of the recipients of a message was unsuitable for the connection
1199 (has no matching addresses), we lost track of needing to mark it
1200 deferred. As a result mail would be lost.
1202 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1204 JH/25 Decoding ACL controls is now done using a binary search; the source code
1205 takes up less space and should be simpler to maintain. Merge the ACL
1206 condition decode tables also, with similar effect.
1208 JH/26 Fix problem with one_time used on a redirect router which returned the
1209 parent address unchanged. A retry would see the parent address marked as
1210 delivered, so not attempt the (identical) child. As a result mail would
1213 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1214 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1215 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1218 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1219 Main config options check_{log,spool}_{inodes,space} are now
1220 100 inodes, 10MB unless set otherwise in the configuration.
1222 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1223 Previously it only applied to the main-section connection policy
1226 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1228 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1229 by me. Added RFC7919 DH primes as an alternative.
1231 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1232 are not in the system include path.
1234 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1235 GnuTLS, when a session startup failed (eg because the client disconnected)
1236 Exim did stdio operations after fclose. This was exposed by a recent
1237 change which nulled out the file handle after the fclose.
1239 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1240 signed directly by the cert-signing cert, rather than an intermediate
1241 OCSP-signing cert. This is the model used by LetsEncrypt.
1243 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1245 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1246 an incoming connection.
1248 HS/02 Bug 1802: Do not half-close the connection after sending a request
1251 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1252 fallback to "prime256v1".
1254 JH/34 SECURITY: Use proper copy of DATA command in error message.
1255 Could leak key material. Remotely exploitable. CVE-2016-9963.
1261 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1262 and 3.4.4 - once the server is enabled to respond to an OCSP request
1263 it does even when not requested, resulting in a stapling non-aware
1264 client dropping the TLS connection.
1266 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1267 support variable-length bit vectors. No functional change.
1269 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1270 The I= interface field on outgoing lines is now after the H= remote
1271 host field, same as incoming lines. There is a separate
1272 outgoing_interface log selector which allows you to disable the
1275 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1276 If not running log_selector +smtp_connection the mainlog would be held
1277 open indefinitely after a "too many connections" event, including to a
1278 deleted file after a log rotate. Leave the per net connection logging
1279 leaving it open for efficiency as that will be quickly detected by the
1280 check on the next write.
1282 HS/01 Bug 1671: Fix post transport crash.
1283 Processing the wait-<transport> messages could crash the delivery
1284 process if the message IDs didn't exist for some reason. When
1285 using 'split_spool_directory=yes' the construction of the spool
1286 file name failed already, exposing the same netto behaviour.
1288 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1289 mime_regex ACL conditions.
1291 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1292 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1293 to HELO, local diagnostic string.
1295 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1296 log line to debug. Even when configured with a tls authenticator many
1297 client connections are expected to not authenticate in this way, so
1298 an authenticate fail is not an error.
1300 HS/02 Add the Exim version string to the process info. This way exiwhat
1301 gives some more detail about the running daemon.
1303 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1304 matter for fast-change records such as DNSBLs.
1306 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1307 retry record, even if constant. There may be multiple transports with
1308 different interface settings and the retry behaviour needs to be kept
1311 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1313 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1315 JH/10 Bug 840: fix log_defer_output option of pipe transport
1317 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1318 hosts_randomize. This matters a lot when combined with mua_wrapper.
1320 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1321 ${quote_pgsql:<string>} operator.
1323 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1325 JH/14 Change header file name for freeradius-client. Relevant if compiling
1326 with Radius support; from the Gentoo tree and checked under Fedora.
1328 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1330 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1331 option transport_filter, meaning no filtering.
1333 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1335 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1336 defaults to "*" (all hosts). The variable is now available when not built
1337 with TLS, default unset, mainly to enable keeping the testsuite sane.
1338 If a server certificate is not supplied (via tls_certificate) an error is
1339 logged, and clients will find TLS connections fail on startup. Presumably
1340 they will retry in-clear.
1341 Packagers of Exim are strongly encouraged to create a server certificate
1342 at installation time.
1344 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1345 with the $config_file variable.
1347 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1348 in transport context, after the attempt, and per-recipient. The latter type
1349 is per host attempted. The event data is the error message, and the errno
1350 information encodes the lookup type (A vs. MX) used for the (first) host,
1351 and the trailing two digits of the smtp 4xx response.
1353 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1354 to write to mainlog (or rejectlog, paniclog) in the window between file
1355 creation and permissions/ownership being changed. Particularly affects
1356 installations where exicyclog is run as root, rather than exim user;
1357 result is that the running daemon panics and dies.
1359 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1361 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1362 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1363 "pri" and "weight". Note that the previous implicit priority given by the
1364 list order is no longer honoured.
1366 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1367 for DKIM processing.
1369 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1370 by defining SUPPORT_SOCKS.
1372 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1373 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1374 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1375 variables are renamed to proxy_{local,external}_{address,port}.
1377 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1378 for a build by defining SUPPORT_I18N
1380 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1381 of the query string, and make ${quote_redis:} do that quoting.
1383 JH/29 Move Events support from Experimental to mainline, enabled by default
1384 and removable for a build by defining DISABLE_EVENT.
1386 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1388 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1389 cached by the daemon.
1391 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1392 by defining LOOKUP_REDIS. The libhiredis library is required.
1394 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1395 keys are given for lookup.
1397 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1398 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1399 only supported when built with TLS support. The PolarSSL SHA routines
1400 are still used when the TLS library is too old for convenient support.
1402 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1403 openssl_options), for security. OpenSSL forces this from version 1.1.0
1404 server-side so match that on older versions.
1406 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1407 allocation for $value could be released as the expansion processing
1408 concluded, but leaving the global pointer active for it.
1410 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1411 and to use the domains and local_parts ACL conditions.
1413 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1414 incorrectly not doubled on cutthrough transmission, hence seen as a
1415 body-termination at the receiving system - resulting in truncated mails.
1416 Commonly the sender saw a TCP-level error, and retransmitted the message
1417 via the normal store-and-forward channel. This could result in duplicates
1418 received - but deduplicating mailstores were liable to retain only the
1419 initial truncated version.
1421 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1423 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1425 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1426 we're in there, support oversigning also; bug 1309.
1428 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1430 HS/04 Add support for keep_environment and add_environment options.
1432 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1433 either intentional arithmetic overflow during PRNG, or testing config-
1436 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1437 delivery resulted in actual delivery. Cancel cutthrough before DATA
1440 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1441 rejects a recipient: pass the reject to the originator.
1443 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1444 Many were false-positives and ignorable, but it's worth fixing the
1447 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1448 for the new environment-manipulation done at startup. Move the routines
1449 from being local to tls.c to being global via the os.c file.
1451 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1452 an extract embedded as result-arg for a map, the first arg for extract
1453 is unavailable so we cannot tell if this is a numbered or keyed
1454 extraction. Accept either.
1460 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1463 JH/02 The smtp transport option "multi_domain" is now expanded.
1465 JH/03 The smtp transport now requests PRDR by default, if the server offers
1468 JH/04 Certificate name checking on server certificates, when exim is a client,
1469 is now done by default. The transport option tls_verify_cert_hostnames
1470 can be used to disable this per-host. The build option
1471 EXPERIMENTAL_CERTNAMES is withdrawn.
1473 JH/05 The value of the tls_verify_certificates smtp transport and main options
1474 default to the word "system" to access the system default CA bundle.
1475 For GnuTLS, only version 3.0.20 or later.
1477 JH/06 Verification of the server certificate for a TLS connection is now tried
1478 (but not required) by default. The verification status is now logged by
1479 default, for both outbound TLS and client-certificate supplying inbound
1482 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1485 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1486 Status Notification (bounce) messages are now MIME format per RFC 3464.
1487 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1488 under the control of the dsn_advertise_hosts option, and routers may
1489 have a dsn_lasthop option.
1491 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1492 default, modifiable by a malware= option. The list separator for
1493 the options can now be changed in the usual way. Bug 68.
1495 JH/10 The smtp_receive_timeout main option is now expanded before use.
1497 JH/11 The incoming_interface log option now also enables logging of the
1498 local interface on delivery outgoing connections.
1500 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1501 if the interface and destination host and port all match.
1503 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1506 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1507 Patch from Andrew Lewis.
1509 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1510 now supports optional time-restrictions, weighting, and priority
1511 modifiers per server. Patch originally by <rommer@active.by>.
1513 JH/16 The spamd_address main option now supports a mixed list of local
1514 and remote servers. Remote servers can be IPv6 addresses, and
1515 specify a port-range.
1517 JH/17 Bug 68: The spamd_address main option now supports an optional
1518 timeout value per server.
1520 JH/18 Bug 1581: Router and transport options headers_add/remove can
1521 now have the list separator specified.
1523 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1526 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1529 JH/21 Support for the A6 type of dns record is withdrawn.
1531 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1532 rather than the verbs used.
1534 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1535 from 255 to 1024 chars.
1537 JH/24 Verification callouts now attempt to use TLS by default.
1539 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1540 are generic router options now. The defaults didn't change.
1542 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1543 Original patch from Alexander Shikoff, worked over by JH.
1545 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1546 files if ZCAT_COMMAND is not executable.
1548 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1550 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1552 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1553 Normally benign, it bites when the pair was led to by a CNAME;
1554 modern usage is to not canonicalize the domain to a CNAME target
1555 (and we were inconsistent anyway for A-only vs AAAA+A).
1557 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1559 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1560 when evaluating $sender_host_dnssec.
1562 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1563 $sender_helo_dnssec variable.
1565 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1567 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1569 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1571 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1572 documented as working, but never had. Support all but $spam_report.
1574 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1575 added for tls authenticator.
1577 HS/03 Add perl_taintmode main config option
1583 TL/01 When running the test suite, the README says that variables such as
1584 no_msglog_check are global and can be placed anywhere in a specific
1585 test's script, however it was observed that placement needed to be near
1586 the beginning for it to behave that way. Changed the runtest perl
1587 script to read through the entire script once to detect and set these
1588 variables, reset to the beginning of the script, and then run through
1589 the script parsing/test process like normal.
1591 TL/02 The BSD's have an arc4random API. One of the functions to induce
1592 adding randomness was arc4random_stir(), but it has been removed in
1593 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1594 function when detected.
1596 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1597 cause callback expansion.
1599 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1600 syntax errors in an expansion can be treated as a string instead of
1601 logging or causing an error, due to the internal use of bool_lax
1602 instead of bool when processing it.
1604 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1605 server certificates when making smtp deliveries.
1607 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1609 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1611 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1613 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1614 Merged patch from Sebastian Wiedenroth.
1616 JH/05 Fix results-pipe from transport process. Several recipients, combined
1617 with certificate use, exposed issues where response data items split
1618 over buffer boundaries were not parsed properly. This eventually
1619 resulted in duplicates being sent. This issue only became common enough
1620 to notice due to the introduction of connection certificate information,
1621 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1623 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1624 size buffer was used, resulting in syntax errors when an expansion
1627 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1628 version 3.3.6 or later.
1630 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1631 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1632 both become "event_action", the variables become $event_name, $event_data
1633 and $event_defer_errno. There is a new variable $verify_mode, usable in
1634 routers, transports and related events. The tls:cert event is now also
1635 raised for inbound connections, if the main configuration event_action
1638 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1639 early OCSP support, but no stapling (appears to be less than 1.0.0).
1641 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1642 server certificate names available under the smtp transport option
1643 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1646 JH/10 Time-related extraction expansions from certificates now use the main
1647 option "timezone" setting for output formatting, and are consistent
1648 between OpenSSL and GnuTLS compilations. Bug 1541.
1650 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1651 encoded parameter in the incoming message. Bug 1558.
1653 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1654 include certificate info, eximon was claiming there were spoolfile
1657 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1659 JH/14 Log delivery-related information more consistently, using the sequence
1660 "H=<name> [<ip>]" wherever possible.
1662 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1663 are problematic for Debian distribution, omit them from the release
1666 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1668 JH/16 Fix string representation of time values on 64bit time_t architectures.
1671 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1672 output list separator was used.
1677 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1678 checkers that were complaining about end of non-void function with no
1681 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1682 This was a regression introduced in 4.83 by another bugfix.
1684 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1686 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1687 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1693 TF/01 Correctly close the server side of TLS when forking for delivery.
1695 When a message was received over SMTP with TLS, Exim failed to clear up
1696 the incoming connection properly after forking off the child process to
1697 deliver the message. In some situations the subsequent outgoing
1698 delivery connection happened to have the same fd number as the incoming
1699 connection previously had. Exim would try to use TLS and fail, logging
1700 a "Bad file descriptor" error.
1702 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1703 utilities have not been installed.
1705 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1706 temporary space as the ACL may create new global variables.
1708 TL/01 LDAP support uses per connection or global context settings, depending
1709 upon the detected version of the libraries at build time.
1711 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1712 to extract and use the src ip:port in logging and expansions as if it
1713 were a direct connection from the outside internet. PPv2 support was
1714 updated based on HAProxy spec change in May 2014.
1716 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1718 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1719 Properly escape header and check for NULL return.
1721 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1724 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1726 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1727 characters in header names, implemented as a verify condition.
1728 Contributed by Michael Fischer v. Mollard.
1730 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1731 results permerror and temperror. Previous values are deprecated but
1732 still accepted. In a future release, err_perm and err_temp will be
1733 completely removed, which will be a backward incompatibility if the
1734 ACL tests for either of these two old results. Patch contributed by
1735 user bes-internal on the mailing list.
1737 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1739 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1740 selectors, in both main and reject logs.
1742 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1745 JH/07 Add malware type "sock" for talking to simple daemon.
1747 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1749 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1750 routers/transports under cutthrough routing.
1752 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1753 numbers. Touch up "bool" conditional to keep the same definition.
1755 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1757 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1759 JH/12 Expand items in router/transport headers_add or headers_remove lists
1760 individually rather than the list as a whole. Bug 1452.
1762 Required for reasonable handling of multiple headers_ options when
1763 they may be empty; requires that headers_remove items with embedded
1764 colons must have them doubled (or the list-separator changed).
1766 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1767 view the policy declared in the DMARC record. Currently, $dmarc_status
1768 is a combined value of both the record presence and the result of the
1771 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1773 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1774 dnslookup router and the smtp transport (applying to the forward
1777 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1778 of ldap servers used for a specific lookup. Patch provided by Heiko
1781 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1782 New variable $lookup_dnssec_authenticated for observability.
1784 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1785 Patch submitted by Lars Timman.
1787 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1789 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1790 Requires trusted mode and valid format message id, aborts otherwise.
1791 Patch contributed by Heiko Schlichting.
1793 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1794 certextract with support for various fields. Bug 1358.
1796 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1797 is requested by default, modifiable by smtp transport option
1800 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1801 operate on certificate variables to give certificate fingerprints
1802 Also new ${sha256:cert_variable}.
1804 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1806 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1807 Christian Aistleitner.
1809 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1811 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1812 file. Patch from Wolfgang Breyha.
1814 JH/25 Expand the coverage of the delivery $host and $host_address to
1815 client authenticators run in verify callout. Bug 1476.
1817 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1818 align with daemon_smtp_ports. Bug 72.
1820 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1821 support and error reporting did not work properly.
1823 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1824 and is readable. Patch from Andrew Colin Kissa.
1826 TL/14 Enhance documentation of ${run expansion and how it parses the
1827 commandline after expansion, particularly in the case when an
1828 unquoted variable expansion results in an empty value.
1830 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1832 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1833 by renaming away from it. Addresses GH issue 15, reported by
1836 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1838 TL/15 SECURITY: prevent double expansion in math comparison functions
1839 (can expand unsanitized data). Not remotely exploitable.
1846 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1848 PP/02 Make -n do something, by making it not do something.
1849 When combined with -bP, the name of an option is not output.
1851 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1854 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1855 $sender_host_name and config options to manage this, and basic check
1858 PP/05 DSCP support for outbound connections and control modifier for inbound.
1860 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1861 (Only plugin which currently uses this is kerberos4, which nobody should
1862 be using, but we should make it available and other future plugins might
1863 conceivably use it, even though it would break NAT; stuff *should* be
1864 using channel bindings instead).
1866 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1867 name; added for Sendmail compatibility; requires admin caller.
1868 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1869 just ignore it); requires trusted caller.
1870 Also parse but ignore: -Ac -Am -X<logfile>
1873 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1875 TL/02 Add +smtp_confirmation as a default logging option.
1877 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1878 Patch by Magnus Holmgren from 2007-02-20.
1880 TL/04 Bugzilla 1281 - Spec typo.
1881 Bugzilla 1283 - Spec typo.
1882 Bugzilla 1290 - Spec grammar fixes.
1884 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1886 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1888 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1889 mailing list by by Dmitry Isaikin.
1891 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1893 JH/02 Support "G" suffix to numbers in ${if comparisons.
1895 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1897 NM/01 Bugzilla 1197 - Spec typo
1898 Bugzilla 1196 - Spec examples corrections
1900 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1902 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1903 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1906 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1907 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1909 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1910 "acl {{name}{arg}...}", and optional args on acl condition
1913 JH/05 Permit multiple router/transport headers_add/remove lines.
1915 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1917 JH/07 Avoid using a waiting database for a single-message-only transport.
1918 Performance patch from Paul Fisher. Bugzilla 1262.
1920 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1923 JH/09 Add $headers_added variable, with content from use of ACL modifier
1924 add_header (but not yet added to the message). Bugzilla 199.
1926 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1927 Pulled from Bugzilla 817 by Wolfgang Breyha.
1929 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1931 (nb: this is the same fix as in Exim 4.80.1)
1933 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1936 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1939 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1941 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1942 advertise SMTP AUTH mechanism to us, instead of a generic
1943 protocol violation error. Also, make Exim more robust to bad
1944 data from the Dovecot auth socket.
1946 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1948 When a queue runner is handling a message, Exim first routes the
1949 recipient addresses, during which it prunes them based on the retry
1950 hints database. After that it attempts to deliver the message to
1951 any remaining recipients. It then updates the hints database using
1954 So if a recipient address works intermittently, it can get repeatedly
1955 deferred at routing time. The retry hints record remains fresh so the
1956 address never reaches the final cutoff time.
1958 This is a fairly common occurrence when a user is bumping up against
1959 their storage quota. Exim had some logic in its local delivery code
1960 to deal with this. However it did not apply to per-recipient defers
1961 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1963 This change adds a proper retry rule check during routing so that the
1964 final cutoff time is checked against the message's age. We only do
1965 this check if there is an address retry record and there is not a
1966 domain retry record; this implies that previous attempts to handle
1967 the address had the retry_use_local_parts option turned on. We use
1968 this as an approximation for the destination being like a local
1969 delivery, as in LMTP.
1971 I suspect this new check makes the old local delivery cutoff check
1972 redundant, but I have not verified this so I left the code in place.
1974 TF/02 Correct gecos expansion when From: is a prefix of the username.
1976 Test 0254 submits a message to Exim with the header
1980 When I ran the test suite under the user fanf2, Exim expanded
1981 the header to contain my full name, whereas it should have added
1982 a Resent-Sender: header. It erroneously treats any prefix of the
1983 username as equal to the username.
1985 This change corrects that bug.
1987 GF/01 DCC debug and logging tidyup
1988 Error conditions log to paniclog rather than rejectlog.
1989 Debug lines prefixed by "DCC: " to remove any ambiguity.
1991 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1993 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1994 Bug spotted by Jeremy Harris; was flawed since initial commit.
1995 Would have resulted in OCSP responses post-SNI triggering an Exim
1996 NULL dereference and crash.
1998 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2000 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2001 Bug detection, analysis and fix by Samuel Thibault.
2002 Bugzilla 1331, Debian bug #698092.
2004 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2006 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2007 Server implementation by Todd Lyons, client by JH.
2008 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2009 config variable "prdr_enable" controls whether the server
2010 advertises the facility. If the client requests PRDR a new
2011 acl_data_smtp_prdr ACL is called once for each recipient, after
2012 the body content is received and before the acl_smtp_data ACL.
2013 The client is controlled by both of: a hosts_try_prdr option
2014 on the smtp transport, and the server advertisement.
2015 Default client logging of deliveries and rejections involving
2016 PRDR are flagged with the string "PRDR".
2018 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2019 fclose(). Diagnosis by Todd Lyons.
2021 PP/17 Update configure.default to handle IPv6 localhost better.
2022 Patch by Alain Williams (plus minor tweaks).
2025 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2026 This is now consistent with GnuTLS, and is now documented: the
2027 previous undocumented portable approach to treating the option as
2028 unset was to force an expansion failure. That still works, and
2029 an empty string is now equivalent.
2031 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2032 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2033 not performing validation itself.
2035 PP/20 Added force_command boolean option to pipe transport.
2036 Patch from Nick Koston, of cPanel Inc.
2038 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2041 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2043 PP/21 Fix eximon continuous updating with timestamped log-files.
2044 Broken in a format-string cleanup in 4.80, missed when I repaired the
2045 other false fix of the same issue.
2046 Report and fix from Heiko Schlichting.
2049 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2050 Report from Prashanth Katuri.
2052 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2053 It's SecureTransport, so affects any MacOS clients which use the
2054 system-integrated TLS libraries, including email clients.
2056 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2057 using a MIME ACL for non-SMTP local injection.
2058 Report and assistance in diagnosis by Warren Baker.
2060 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2062 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2064 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2065 last id that failed so it may be referenced in subsequent ACL's.
2067 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2070 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2071 ldap library initialization, allowing self-signed CA's to be
2072 used. Also properly sets require_cert option later in code by
2073 using NULL (global ldap config) instead of ldap handle (per
2074 session). Bug diagnosis and testing by alxgomz.
2076 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2077 the src/util/ subdirectory.
2079 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2080 renamed to Transport Post Delivery Action by Jeremy Harris, as
2083 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2084 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2085 redis_servers = needs to be configured which will be used by the redis
2086 lookup. Patch from Warren Baker, of The Packet Hub.
2088 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2090 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2091 hostname or reverse DNS when processing a host list. Used suggestions
2092 from multiple comments on this bug.
2094 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2096 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2097 Missed a few lines, added it to make the runtest require no keyboard
2100 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2101 contains upper case chars. Make router use caseful_local_part.
2103 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2104 support when GnuTLS has been built with p11-kit.
2110 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2112 This, or similar/improved, will also be change PP/11 of 4.82.
2118 PP/01 Handle short writes when writing local log-files.
2119 In practice, only affects FreeBSD (8 onwards).
2120 Bugzilla 1053, with thanks to Dmitry Isaikin.
2122 NM/01 Bugzilla 949 - Documentation tweak
2124 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2127 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2129 PP/02 Implemented gsasl authenticator.
2131 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2133 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2134 `pkg-config foo` for cflags/libs.
2136 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2137 with rest of GSASL and with heimdal_gssapi.
2139 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2140 `pkg-config foo` for cflags/libs for the TLS implementation.
2142 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2143 properties get this fed in as external SSF. A number of robustness
2144 and debugging improvements to the cyrus_sasl authenticator.
2146 PP/08 cyrus_sasl server now expands the server_realm option.
2148 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2149 Patch by Jeremy Harris.
2151 PP/10 Added dbmjz lookup type.
2153 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2155 PP/12 MAIL args handles TAB as well as SP, for better interop with
2156 non-compliant senders.
2157 Analysis and variant patch by Todd Lyons.
2159 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2160 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2161 Patch from Dirk Mueller <dmueller@suse.com>
2163 PP/13 tls_peerdn now print-escaped for spool files.
2164 Observed some $tls_peerdn in wild which contained \n, which resulted
2165 in spool file corruption.
2167 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2168 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2169 or write after TLS renegotiation, which otherwise led to messages
2172 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2173 as a tracking header (ie: a signed header comes before the signature).
2174 Patch from Wolfgang Breyha.
2176 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2177 comma-sep list; embedded commas doubled.
2179 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2181 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2183 Report and patch from Dmitry Banschikov.
2185 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2186 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2187 failures. We appear to now support TLS1.1+ with Exim.
2189 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2190 lets Exim select keys and certificates based upon TLS SNI from client.
2191 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2192 before an outbound SMTP session. New log_selector, +tls_sni.
2194 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2195 NULL dereference. Report and patch from Alun Jones.
2197 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2198 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2199 Not seeing resolver debug output on NetBSD, but suspect this is a
2200 resolver implementation change.
2202 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2203 Left warnings. Added "eximon gdb" invocation mode.
2205 PP/21 Defaulting "accept_8bitmime" to true, not false.
2207 PP/22 Added -bw for inetd wait mode support.
2209 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2210 locate the relevant includes and libraries. Made this the default.
2212 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2213 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2215 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2216 This may cause build issues on older platforms.
2218 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2219 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2220 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2221 Added SNI support via GnuTLS too.
2222 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2224 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2226 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2227 Applied second patch from Janne, implementing suggestion to default
2228 multiple-strings-in-record handling to match SPF spec.
2230 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2232 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2233 read-only, out of scope).
2234 Patch from Wolfgang Breyha, report from Stuart Northfield.
2236 PP/29 Fix three issues highlighted by clang analyser static analysis.
2237 Only crash-plausible issue would require the Cambridge-specific
2238 iplookup router and a misconfiguration.
2239 Report from Marcin Mirosław.
2241 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2243 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2244 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2245 As part of this, removing so much warning spew let me fix some minor
2246 real issues in debug logging.
2248 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2249 assignment on my part. Fixed.
2251 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2252 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2253 Janne Snabb (who went above and beyond: thank you).
2255 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2256 string otherwise requires a connection and a bunch more work and it's
2257 relatively easy to get wrong. Should also expose TLS library linkage
2260 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2261 64-bit ${eval} (JH/03).
2263 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2264 GNU libc to support some of the 64-bit stuff, should not lead to
2265 conflicts. Defined before os.h is pulled in, so if a given platform
2266 needs to override this, it can.
2268 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2269 protection layer was required, which is not implemented.
2270 Bugzilla 1254, patch from Wolfgang Breyha.
2272 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2273 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2274 tls_dhparam take prime identifiers. Also unbreak combination of
2275 OpenSSL+DH_params+TLSSNI.
2277 PP/39 Disable SSLv2 by default in OpenSSL support.
2283 PP/01 Solaris build fix for Oracle's LDAP libraries.
2284 Bugzilla 1109, patch from Stephen Usher.
2286 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2288 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2291 TF/02 Fix a couple more cases where we did not log the error message
2292 when unlink() failed. See also change 4.74-TF/03.
2294 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2295 lock up or crash if it happened to be inside a call to libc when it
2296 got a SIGUSR1 from exiwhat.
2298 The SIGUSR1 handler appends the current process status to the process
2299 log which is later printed by exiwhat. It used to use the general
2300 purpose logging code to do this, but several functions it calls are
2301 not safe for signals.
2303 The new output code in the SIGUSR1 handler is specific to the process
2304 log, and simple enough that it's easy to inspect for signal safety.
2305 Removing some special cases also simplifies the general logging code.
2306 Removing the spurious timestamps from the process log simplifies
2309 TF/04 Improved ratelimit ACL condition.
2311 The /noupdate option has been deprecated in favour of /readonly which
2312 has clearer semantics. The /leaky, /strict, and /readonly update modes
2313 are mutually exclusive. The update mode is no longer included in the
2314 database key; it just determines when the database is updated. (This
2315 means that when you upgrade Exim will forget old rate measurements.)
2317 Exim now checks that the per_* options are used with an update mode that
2318 makes sense for the current ACL. For example, when Exim is processing a
2319 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2320 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2321 must specify per_mail/readonly. If you omit the update mode it defaults to
2322 /leaky where that makes sense (as before) or /readonly where required.
2324 The /noupdate option is now undocumented but still supported for
2325 backwards compatibility. It is equivalent to /readonly except that in
2326 ACLs where /readonly is required you may specify /leaky/noupdate or
2327 /strict/noupdate which are treated the same as /readonly.
2329 A useful new feature is the /count= option. This is a generalization
2330 of the per_byte option, so that you can measure the throughput of other
2331 aggregate values. For example, the per_byte option is now equivalent
2332 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2334 The per_rcpt option has been generalized using the /count= mechanism
2335 (though it's more complicated than the per_byte equivalence). When it is
2336 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2337 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2338 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2339 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2340 Note that using per_rcpt with a non-readonly update mode in more than
2341 one ACL will cause the recipients to be double-counted. (The per_mail
2342 and per_byte options don't have this problem.)
2344 The handling of very low rates has changed slightly. If the computed rate
2345 is less than the event's count (usually one) then this event is the first
2346 after a long gap. In this case the rate is set to the same as this event's
2347 count, so that the first message of a spam run is counted properly.
2349 The major new feature is a mechanism for counting the rate of unique
2350 events. The new per_addr option counts the number of different
2351 recipients that someone has sent messages to in the last time period. It
2352 behaves like per_rcpt if all the recipient addresses are different, but
2353 duplicate recipient addresses do not increase the measured rate. Like
2354 the /count= option this is a general mechanism, so the per_addr option
2355 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2356 example, measure the rate that a client uses different sender addresses
2357 with the options per_mail/unique=$sender_address. There are further
2358 details in the main documentation.
2360 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2362 TF/06 Removed a few PCRE remnants.
2364 TF/07 Automatically extract Exim's version number from tags in the git
2365 repository when doing development or release builds.
2367 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2368 Bugzilla 879. Patch from Paul Fisher.
2370 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2371 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2374 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2376 PP/05 Variable $av_failed, true if the AV scanner deferred.
2377 Bugzilla 1078. Patch from John Horne.
2379 PP/06 Stop make process more reliably on build failure.
2380 Bugzilla 1087. Patch from Heiko Schlittermann.
2382 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2383 Bugzilla 1089. Patch from Heiko Schlittermann.
2385 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2386 Bugzilla 1131. Patch from Holger Weiß.
2388 PP/09 Handle IPv6 addresses with SPF.
2389 Bugzilla 860. Patch from Wolfgang Breyha.
2391 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2393 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2396 PP/11 match_* no longer expand right-hand-side by default.
2397 New compile-time build option, EXPAND_LISTMATCH_RHS.
2398 New expansion conditions, "inlist", "inlisti".
2400 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2402 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2404 PP/14 fix log_write() format string regression from TF/03.
2405 Bugzilla 1152. Patch from Dmitry Isaikin.
2411 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2413 PP/02 Harmonised TLS library version reporting; only show if debugging.
2414 Layout now matches that introduced for other libraries in 4.74 PP/03.
2416 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2418 PP/04 New "dns_use_edns0" global option.
2420 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2423 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2424 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2426 TK/01 Updated PolarSSL code to 0.14.2.
2427 Bugzilla 1097. Patch from Andreas Metzler.
2429 PP/07 Catch divide-by-zero in ${eval:...}.
2430 Fixes bugzilla 1102.
2432 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2435 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2436 format-string attack -- SECURITY: remote arbitrary code execution.
2438 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2439 time unintentionally subject to list matching rules, letting the header
2440 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2441 arbitrary string expansion). This allowed for information disclosure.
2443 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2444 INT_MIN/-1 -- value coerced to INT_MAX.
2450 NM/01 Workaround for PCRE version dependency in version reporting
2453 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2454 This fixes portability to compilers other than gcc, notably
2455 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2457 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2458 makefiles for portability to HP-UX and POSIX correctness.
2460 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2461 Also via indented variable definition in the Makefile.
2462 (Debugging by Oliver Heesakkers).
2464 PP/02 Restore caching of spamd results with expanded spamd_address.
2465 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2467 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2468 Improves build reliability. Fix from: Frank Elsner
2470 NM/02 Fix wide character breakage in the rfc2047 coding
2471 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2473 NM/03 Allow underscore in dnslist lookups
2474 Fixes bug 1026. Patch from Graeme Fowler
2476 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2477 Code patches from Adam Ciarcinski of NetBSD.
2479 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2482 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2483 is logged, to avoid truncation. Patch from John Horne.
2485 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2486 Patch from Jakob Hirsch.
2488 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2489 SQL string expansion failure details.
2490 Patch from Andrey Oktyabrski.
2492 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2493 Patch from Simon Arlott.
2495 PP/09 New lookups functionality failed to compile on old gcc which rejects
2496 extern declarations in function scope.
2497 Patch from Oliver Fleischmann
2499 PP/10 Use sig_atomic_t for flags set from signal handlers.
2500 Check getgroups() return and improve debugging.
2501 Fixed developed for diagnosis in bug 927 (which turned out to be
2504 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2505 Patch from Mark Zealey.
2507 PP/12 Bugzilla 1056: Improved spamd server selection.
2508 Patch from Mark Zealey.
2510 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2511 Based on patch from Heiko Schlittermann.
2513 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2514 Patch from Uwe Doering, sign-off by Michael Haardt.
2516 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2517 variable declaration deep within a block. Bug and patch from
2520 PP/15 lookups-Makefile IRIX compatibility coercion.
2522 PP/16 Make DISABLE_DKIM build knob functional.
2524 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2525 Patch by Simon Arlott
2527 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2528 variable argument macros. Our copy now differs from upstream.
2534 TF/01 Failure to get a lock on a hints database can have serious
2535 consequences so log it to the panic log.
2537 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2538 controlled using the smtp_confirmation log selector.
2540 TF/03 Include the error message when we fail to unlink a spool file.
2542 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2543 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2544 for maintaining out-of-tree patches for some time.
2546 PP/01 Bugzilla 139: Documentation and portability issues.
2547 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2548 Handle per-OS dynamic-module compilation flags.
2550 PP/02 Let /dev/null have normal permissions.
2551 The 4.73 fixes were a little too stringent and complained about the
2552 permissions on /dev/null. Exempt it from some checks.
2553 Reported by Andreas M. Kirchwitz.
2555 PP/03 Report version information for many libraries, including
2556 Exim version information for dynamically loaded libraries. Created
2557 version.h, now support a version extension string for distributors
2558 who patch heavily. Dynamic module ABI change.
2560 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2561 privilege escalation vulnerability whereby the Exim run-time user
2562 can cause root to append content of the attacker's choosing to
2565 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2568 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2569 If dropping privileges for untrusted macros, we disabled normal logging
2570 on the basis that it would fail; for the Exim run-time user, this is not
2571 the case, and it resulted in successful deliveries going unlogged.
2572 Fixed. Reported by Andreas Metzler.
2578 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2579 only prepend for the Resent-* case. Fixes regression introduced in
2580 Exim 4.70 by NM/22 for Bugzilla 607.
2582 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2583 increasing numbers of administrators be bitten by this.
2585 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2587 PP/03 Bugzilla 994: added openssl_options main configuration option.
2589 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2591 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2593 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2595 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2596 without a peer certificate, leading to a segfault because of an
2597 assumption that peers always have certificates. Be a little more
2598 paranoid. Problem reported by Martin Tscholak.
2600 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2601 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2602 NB: ClamAV planning to remove STREAM in "middle of 2010".
2603 CL also introduces -bmalware, various -d+acl logging additions and
2604 more caution in buffer sizes.
2606 PP/09 Implemented reverse_ip expansion operator.
2608 PP/10 Bugzilla 937: provide a "debug" ACL control.
2610 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2612 PP/12 Bugzilla 973: Implement --version.
2614 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2616 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2618 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2620 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2621 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2622 ignore trailing whitespace.
2624 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2626 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2629 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2630 Notification from Dr Andrew Aitchison.
2632 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2633 ExtendedDetectionInfo response format.
2634 Notification from John Horne.
2636 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2639 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2640 XSL and documented dependency on system catalogs, with examples of how
2643 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2646 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2647 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2648 configuration file which is writeable by the Exim user or group.
2650 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2651 of configuration files to cover files specified with the -C option if
2652 they are going to be used with root privileges, not just the default
2655 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2656 option (effectively making it always true).
2658 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2659 files to be used while preserving root privileges.
2661 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2662 that rogue child processes cannot use them.
2664 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2665 run-time user, instead of root.
2667 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2668 Exim run-time user without dropping privileges.
2670 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2671 result string, instead of calling string_vformat() twice with the same
2674 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2675 for other users. Others should always drop root privileges if they use
2676 -C on the command line, even for a whitelisted configure file.
2678 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2680 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2686 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2687 $data_path, and $header_path variables; fixed documentation bugs and
2690 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2691 exipick to access non-standard spools, including the "frozen" queue
2694 NM/01 Bugzilla 965: Support mysql stored procedures.
2695 Patch from Alain Williams
2697 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2699 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2700 Patch from Andreas Metzler
2702 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2703 Patch from Kirill Miazine
2705 NM/05 Bugzilla 671: Added umask to procmail example.
2707 JJ/03 installed exipick 20100323.0, fixing doc bug
2709 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2710 directory. Notification and patch from Dan Rosenberg.
2712 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2714 TK/02 Improve log output when DKIM signing operation fails.
2716 MH/01 Treat the transport option dkim_domain as a colon separated
2717 list, not as a single string, and sign the message with each element,
2718 omitting multiple occurences of the same signer.
2720 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2721 Bugzilla 985, 986. Patch by Simon Arlott
2723 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2724 Patch by Simon Arlott
2726 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2727 MBX locking. Notification from Dan Rosenberg.
2733 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2735 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2737 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2739 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2741 NM/04 Bugzilla 915: Flag broken perl installation during build.
2747 TK/01 Added patch by Johannes Berg that expands the main option
2748 "spamd_address" if it starts with a dollar sign.
2750 TK/02 Write list of recipients to X-Envelope-Sender header when building
2751 the mbox-format spool file for content scanning (suggested by Jakob
2754 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2755 (http://www.dcc-servers.net/) support via dccifd. Activated by
2756 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2758 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2759 by Mark Daniel Reidel <mr@df.eu>.
2761 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2762 When building exim an external PCRE library is now needed -
2763 PCRE is a system library on the majority of modern systems.
2764 See entry on PCRE_LIBS in EDITME file.
2766 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2767 conversation. Added nologin parameter to request.
2768 Patch contributed by Kirill Miazine.
2770 TF/01 Do not log submission mode rewrites if they do not change the address.
2772 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2774 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2775 log files in place. Contributed by Roberto Lima.
2777 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2779 TF/03 Bugzilla 615: When checking the local_parts router precondition
2780 after a local_part_suffix or local_part_prefix option, Exim now
2781 does not use the address's named list lookup cache, since this
2782 contains cached lookups for the whole local part.
2784 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2785 Robert Millan. Documentation is in experimental-spec.txt.
2787 TF/04 Bugzilla 668: Fix parallel build (make -j).
2789 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2791 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2792 Patch provided by Jan Srzednicki.
2794 TF/05 Leading white space used to be stripped from $spam_report which
2795 wrecked the formatting. Now it is preserved.
2797 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2798 that they are available at delivery time.
2800 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2802 TF/08 TLS error reporting now respects the incoming_interface and
2803 incoming_port log selectors.
2805 TF/09 Produce a more useful error message if an SMTP transport's hosts
2806 setting expands to an empty string.
2808 NM/06 Bugzilla 744: EXPN did not work under TLS.
2809 Patch provided by Phil Pennock.
2811 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2812 Patch provided by Richard Godbee.
2814 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2815 acl_smtp_notquit, added index entry.
2817 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2818 Patch provided by Eugene Bujak.
2820 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2821 accept(). Patch provided by Maxim Dounin.
2823 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2824 Patch provided by Phil Pennock.
2826 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2828 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2829 Patch provided by Brad "anomie" Jorsch.
2831 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2832 Patch provided by Dean Brooks.
2834 TK/05 Add native DKIM support (does not depend on external libraries).
2836 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2837 Patch provided by Graeme Fowler.
2839 NM/16 Bugzilla 851: Documentation example syntax fix.
2841 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2843 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2846 NM/19 Bugzilla 745: TLS version reporting.
2847 Patch provided by Phil Pennock.
2849 NM/20 Bugzilla 167: bool: condition support.
2850 Patch provided by Phil Pennock.
2852 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2853 clients. Patch provided by Phil Pennock.
2855 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2856 Patch provided by Brad "anomie" Jorsch.
2858 NM/23 Bugzilla 687: Fix misparses in eximstats.
2859 Patch provided by Heiko Schlittermann.
2861 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2862 Patch provided by Heiko Schlittermann.
2864 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2865 plus update to original patch.
2867 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2869 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2870 Patch provided by David Brownlee.
2872 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2874 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2876 NM/30 Bugzilla 888: TLS documentation bugfixes.
2878 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2880 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2881 Unlike the original bugzilla I have changed all shell scripts in src tree.
2883 NM/33 Bugzilla 898: Transport filter timeout fix.
2884 Patch by Todd Rinaldo.
2886 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2887 Patch by Serge Demonchaux.
2889 NM/35 Bugzilla 39: Base64 decode bug fixes.
2890 Patch by Jakob Hirsch.
2892 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2894 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2896 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2898 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2904 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2905 ALT-N's libdkim that I have put here:
2906 http://duncanthrax.net/exim-experimental/
2908 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2909 were called 'true' and it seems that C99 defines that as a reserved
2910 keyword to be used with 'bool' variable types. That means you could
2911 not include C99-style headers which use bools without triggering
2912 build errors in sieve.c.
2914 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2915 as mailq or other aliases. Changed the --help handling significantly
2916 to do whats expected. exim_usage() emits usage/help information.
2918 SC/01 Added the -bylocaldomain option to eximstats.
2920 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2922 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2924 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2930 PH/01 Another patch from the Sieve maintainer.
2932 PH/02 When an IPv6 address is converted to a string for single-key lookup
2933 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2934 dots are used instead of colons so that keys in lsearch files need not
2935 contain colons. This was done some time before quoting was made available
2936 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2937 (notated using the quote facility) so as to distinguish them from IPv4
2938 keys. This meant that lookups for IP addresses in host lists did not work
2939 for iplsearch lookups.
2941 This has been fixed by arranging for IPv6 addresses to be expressed with
2942 colons if the lookup type is iplsearch. This is not incompatible, because
2943 previously such lookups could never work.
2945 The situation is now rather anomalous, since one *can* have colons in
2946 ordinary lsearch keys. However, making the change in all cases is
2947 incompatible and would probably break a number of configurations.
2949 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2952 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2953 conversion specification without a maximum field width, thereby enabling
2954 a rogue spamd server to cause a buffer overflow. While nobody in their
2955 right mind would setup Exim to query an untrusted spamd server, an
2956 attacker that gains access to a server running spamd could potentially
2957 exploit this vulnerability to run arbitrary code as the Exim user.
2959 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2960 $primary_hostname instead of what libspf2 thinks the hosts name is.
2962 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2963 a directory entry by the name of the lookup key. Previously, if a
2964 symlink pointed to a non-existing file or a file in a directory that
2965 Exim lacked permissions to read, a lookup for a key matching that
2966 symlink would fail. Now it is enough that a matching directory entry
2967 exists, symlink or not. (Bugzilla 503.)
2969 PH/03 The body_linecount and body_zerocount variables are now exported in the
2972 PH/04 Added the $dnslist_matched variable.
2974 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2975 This means they are set thereafter only if the connection becomes
2978 PH/06 Added the client_condition to authenticators so that some can be skipped
2979 by clients under certain conditions.
2981 PH/07 The error message for a badly-placed control=no_multiline_responses left
2982 "_responses" off the end of the name.
2984 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2986 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2987 (without spaces) instead of just copying the configuration text.
2989 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2991 PH/11 Added $max_received_linelength.
2993 PH/12 Added +ignore_defer and +include_defer to host lists.
2995 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2996 way in which PCRE > 7.0 is built.
2998 PH/14 Implemented queue_only_load_latch.
3000 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3001 MAIL command. The effect was to mangle the value on 64-bit systems.
3003 PH/16 Another patch from the Sieve maintainer.
3005 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3007 PH/18 If a system quota error occurred while trying to create the file for
3008 a maildir delivery, the message "Mailbox is full" was not appended to the
3009 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3010 only to a quota excession during the actual writing of the file.
3012 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3013 characters?) which causes problems in log lines. The DN values are now
3014 passed through string_printing() before being added to log lines.
3016 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3017 and InterBase are left for another time.)
3019 PH/21 Added message_body_newlines option.
3021 PH/22 Guard against possible overflow in moan_check_errorcopy().
3023 PH/23 POSIX allows open() to be a macro; guard against that.
3025 PH/24 If the recipient of an error message contained an @ in the local part
3026 (suitably quoted, of course), incorrect values were put in $domain and
3027 $local_part during the evaluation of errors_copy.
3033 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3034 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3037 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3038 issue a MAIL command.
3040 PH/02 In an ACL statement such as
3042 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3044 if a client was not listed at all, or was listed with a value other than
3045 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3046 the condition was not true (as it should be), so access was not denied.
3047 The bug was that the ! inversion was incorrectly passed on to the second
3048 item. This has been fixed.
3050 PH/03 Added additional dnslists conditions == and =& which are different from
3051 = and & when the dns lookup returns more than one IP address.
3053 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3054 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3056 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3057 FSYNC, which compiles an option called disable_fsync that allows for
3058 bypassing fsync(). The documentation is heavily laced with warnings.
3060 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3062 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3063 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3064 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3065 including adding "make clean"; (3) Added -fPIC when compiling the test
3066 dynamically loaded module, to get rid of a warning.
3068 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3069 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3070 = 0s. The bug is otherwise harmless.
3072 PH/07 There was a bug in the dovecot authenticator such that the value of
3073 $auth1 could be overwritten, and so not correctly preserved, after a
3074 successful authentication. This usually meant that the value preserved by
3075 the server_setid option was incorrect.
3077 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3079 PH/09 Installed PCRE release 7.0.
3081 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3082 run for batched SMTP input. It is now run at the start of every message
3083 in the batch. While fixing this I discovered that the process information
3084 (output by running exiwhat) was not always getting set for -bs and -bS
3085 input. This is fixed, and it now also says "batched" for BSMTP.
3087 PH/11 Added control=no_pipelining.
3089 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3090 patch, slightly modified), and move the expansion of helo_data till after
3091 the connection is made in the smtp transport (so it can use these
3094 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3096 PH/14 Added log_selector = +pid.
3098 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3100 PH/16 Add ${if forany and ${if forall.
3102 PH/17 Added dsn_from option to vary the From: line in DSNs.
3104 PH/18 Flush SMTP output before performing a callout, unless control =
3105 no_callout_flush is set.
3107 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3108 was true (the default) a successful delivery failed to delete the retry
3109 item, thus causing premature timeout of the address. The bug is now
3112 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3114 PH/21 Long custom messages for fakedefer and fakereject are now split up
3115 into multiline responses in the same way that messages for "deny" and
3116 other ACL rejections are.
3118 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3119 with slight modification.
3121 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3122 draft, changing the syntax and factoring some duplicate code".
3124 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3125 for deliveries of the second and subsequent messages over the same SMTP
3128 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3129 ${reduce, with only minor "tidies".
3131 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3133 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3134 expansion side effects.
3136 PH/27 When a message times out after an over-quota error from an Exim-imposed
3137 quota, the bounce message says "mailbox is full". This message was not
3138 being given when it was a system quota that was exceeded. It now should
3141 MH/03 Made $recipients available in local_scan(). local_scan() already has
3142 better access to the recipient list through recipients_list[], but
3143 $recipients can be useful in postmaster-provided expansion strings.
3145 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3146 in the case of a MAIL command with additional options following the
3147 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3148 were accidentally chopped off.
3150 PH/29 SMTP synchronization checks are implemented when a command is read -
3151 there is a check that no more input is waiting when there shouldn't be
3152 any. However, for some commands, a delay in an ACL can mean that it is
3153 some time before the response is written. In this time, more input might
3154 arrive, invalidly. So now there are extra checks after an ACL has run for
3155 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3156 pipelining has not been advertised.
3158 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3160 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3161 returned in the SMTP response when smtp_return_error_details was false.
3162 This has been fixed.
3164 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3165 instead of the C I/O that was originally supplied, because problems were
3166 reported on Solaris.
3168 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3169 Exim which did not show up earlier: it was assuming that a call to
3170 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3171 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3172 was a macro that became an assignment, so it seemed to work. This has
3173 changed to a proper function call with a void return, hence the compile
3174 error. Exim's code has been fixed.
3176 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3179 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3181 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3183 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3184 because it turns out to be too big or there is a timeout) there is no
3185 "Completed" line in the log. When some messages of this type were
3186 selected by exigrep, they were listed as "not completed". Others were
3187 picked up by some special patterns. I have improved the selection
3188 criteria to be more general.
3190 PH/38 The host_find_failed option in the manualroute router can now be set
3191 to "ignore", to completely ignore a host whose IP address cannot be
3192 found. If all hosts are ignored, the behaviour is controlled by the new
3193 host_all_ignored option.
3195 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3196 homing or because of multiple MX records with /mx) generated more than
3197 one IP address, and the following item turned out to be the local host,
3198 all the secondary addresses of the first item were incorrectly removed
3199 from the list, along with the local host and any following hosts (which
3200 is what is supposed to happen).
3202 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3203 whoever called Exim into the -H file. In the case of the daemon it was
3204 behaving confusingly. When first started, it used values for whoever
3205 started the daemon, but after a SIGHUP it used the Exim user (because it
3206 calls itself on a restart). I have changed the code so that it now always
3209 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3210 message are rejected with the same error (e.g. no authentication or bad
3211 sender address), and a DATA command is nevertheless sent (as can happen
3212 with PIPELINING or a stupid MUA), the error message that was given to the
3213 RCPT commands is included in the rejection of the DATA command. This is
3214 intended to be helpful for MUAs that show only the final error to their
3217 PH/42 Another patch from the Sieve maintainer.
3219 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3220 Eximstats - Fixed some broken HTML links and added missing column headers
3222 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3223 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3225 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3227 PH/43 Yet another patch from the Sieve maintainer.
3229 PH/44 I found a way to check for a TCP/IP connection going away before sending
3230 the response to the final '.' that terminates a message, but only in the
3231 case where the client has not sent further data following the '.'
3232 (unfortunately, this is allowed). However, in many cases there won't be
3233 any further data because there won't be any more messages to send. A call
3234 to select() can be used: if it shows that the input is "ready", there is
3235 either input waiting, or the socket has been closed. An attempt to read
3236 the next input character can distinguish the two cases. Previously, Exim
3237 would have sent an OK response which the client would never have see.
3238 This could lead to message repetition. This fix should cure that, at
3239 least in a lot of common cases.
3241 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3242 advertised in response to EHLO.
3248 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3249 fixed by 4.65/MH/01 (is this a record?) are fixed:
3251 (i) An empty string was always treated as zero by the numeric comparison
3252 operators. This behaviour has been restored.
3254 (ii) It is documented that the numeric comparison operators always treat
3255 their arguments as decimal numbers. This was broken in that numbers
3256 starting with 0 were being interpreted as octal.
3258 While fixing these problems I realized that there was another issue that
3259 hadn't been noticed. Values of message_size_limit (both the global option
3260 and the transport option) were treated as octal if they started with 0.
3261 The documentation was vague. These values are now always treated as
3262 decimal, and I will make that clear in the documentation.
3268 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3269 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3272 MH/01 Don't check that the operands of numeric comparison operators are
3273 integers when their expansion is in "skipping" mode (fixes bug
3274 introduced by 4.64-PH/07).
3276 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3277 child addresses, Exim now panics and dies. Previously, because the count
3278 is held in a short int, deliveries were likely to be lost. As such a
3279 large number of recipients for a single message is ridiculous
3280 (performance will be very, very poor), I have chosen to impose a limit
3281 rather than extend the field.
3287 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3288 leftover -K file (the existence of which was triggered by #402).
3289 While we were at it, introduced process PID as part of the -K
3290 filename. This should rule out race conditions when creating
3293 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3294 processing considerably. Previous code took too long for large mails,
3295 triggering a timeout which in turn triggers #401.
3297 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3298 in the DK code in transports.c. sendfile() is not really portable,
3299 hence the _LINUX specificness.
3301 TF/01 In the add_headers option to the mail command in an Exim filter,
3302 there was a bug that Exim would claim a syntax error in any
3303 header after the first one which had an odd number of characters
3306 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3307 callout verification, Exim cached a "reject" for the entire domain. This
3308 is correct for most verifications, but it is not correct for a recipient
3309 verification with use_sender or use_postmaster set, because in that case
3310 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3311 case of MAIL FROM:<> rejection from other early rejections (e.g.
3312 rejection of HELO). When verifying a recipient using a non-null MAIL
3313 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3314 Whatever the result of the callout, the value of the domain cache is
3315 left unchanged (for any other kind of callout, getting as far as trying
3316 RCPT means that the domain itself is ok).
3318 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3321 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3322 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3323 session, the daemon ignores ECONNECT errors and logs others; it now
3324 ignores EPIPE as well.
3326 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3327 (quoted-printable decoding).
3329 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3330 later the small subsequent patch to fix an introduced bug.
3332 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3334 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3336 PH/08 An error is now given if message_size_limit is specified negative.
3338 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3339 to be given (somewhat) arbitrary names.
3341 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3344 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3345 miscellaneous code fixes
3347 PH/10 Added the log_reject_target ACL modifier to specify where to log
3350 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3351 hostname. This is wrong, because it relates to the incoming message (and
3352 probably the interface on which it is arriving) and not to the outgoing
3353 callout (which could be using a different interface). This has been
3354 changed to use the value of the helo_data option from the smtp transport
3355 instead - this is what is used when a message is actually being sent. If
3356 there is no remote transport (possible with a router that sets up host
3357 addresses), $smtp_active_hostname is used.
3359 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3360 tweaks were necessary in order to get it to work (see also 21 below):
3361 (a) The code assumed that strncpy() returns a negative number on buffer
3362 overflow, which isn't the case. Replaced with Exim's string_format()
3364 (b) There were several signed/unsigned issues. I just did the minimum
3365 hacking in of casts. There is scope for a larger refactoring.
3366 (c) The code used strcasecmp() which is not a standard C function.
3367 Replaced with Exim's strcmpic() function.
3368 (d) The code set only $1; it now sets $auth1 as well.
3369 (e) A simple test gave the error "authentication client didn't specify
3370 service in request". It would seem that Dovecot has changed its
3371 interface. Fortunately there's a specification; I followed it and
3372 changed what the client sends and it appears to be working now.
3374 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3377 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3378 address A is aliased to B and C, where B exists and C does not. Without
3379 -v the output is "A verified" because verification stops after a
3380 successful redirection if more than one address is generated. However,
3381 with -v the child addresses are also verified. Exim was outputting "A
3382 failed to verify" and then showing the successful verification for C,
3383 with its parentage. It now outputs "B failed to verify", showing B's
3384 parentage before showing the successful verification of C.
3386 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3387 look up a TXT record in a specific list after matching in a combined
3390 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3391 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3392 they consult the DNS. I had assumed they would set it the way they
3393 wanted; and indeed my experiments on Linux seem to show that in some
3394 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3395 To be on the safe side, however, I have now made the interface to
3396 host_find_byname() similar to host_find_bydns(), with an argument
3397 containing the DNS resolver options. The host_find_byname() function now
3398 sets these options at its start, just as host_find_bydns() does. The smtp
3399 transport options dns_qualify_single and dns_search_parents are passed to
3400 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3401 of host_find_byname() use the default settings of RES_DEFNAMES
3402 (qualify_single) but not RES_DNSRCH (search_parents).
3404 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3405 spool_read_header() do less string testing, by means of a preliminary
3406 switch on the second character of optional "-foo" lines. (This is
3407 overdue, caused by the large number of possibilities that now exist.
3408 Originally there were few.) While I was there, I also converted the
3409 str(n)cmp tests so they don't re-test the leading "-" and the first
3410 character, in the hope this might squeeze out yet more improvement.
3412 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3413 flag allowing group syntax was set by the header_syntax check but not
3414 turned off, possible causing trouble later; (2) The flag was not being
3415 set at all for the header_verify test, causing "group"-style headers to
3416 be rejected. I have now set it in this case, and also caused header_
3417 verify to ignore an empty address taken from a group. While doing this, I
3418 came across some other cases where the code for allowing group syntax
3419 while scanning a header line wasn't quite right (mostly, not resetting
3420 the flag correctly in the right place). These bugs could have caused
3421 trouble for malformed header lines. I hope it is now all correct.
3423 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3424 with the "reply" argument non-NULL. The code, however (which originally
3425 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3426 but it didn't always do it. This confused somebody who was copying the
3427 code for some other use. I have removed all the tests.
3429 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3430 feature that was used to support insecure browsers during the U.S. crypto
3431 embargo. It requires special client support, and Exim is probably the
3432 only MTA that supported it -- and would never use it because real RSA is
3433 always available. This code has been removed, because it had the bad
3434 effect of slowing Exim down by computing (never used) parameters for the
3435 RSA_EXPORT functionality.
3437 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3438 authenticator to fail if there's a tab character in the incoming data
3439 (there should never be unless someone is messing about, as it's supposed
3440 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3441 if the connection is using TLS or if the remote IP is the same as the
3442 local IP, and the "valid-client-cert option" if a client certificate has
3445 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3446 authenticators. This can be used for authorization after authentication
3447 succeeds. (In the case of plaintext, it servers for both authentication
3450 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3451 if any retry times were supplied.
3453 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3454 connection, where there is no client IP address to check. In this
3455 situation, the verify now always succeeds.
3457 PH/25 Applied John Jetmore's -Mset patch.
3459 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3461 PH/27 In a string expansion for a processed (not raw) header when multiple
3462 headers of the same name were present, leading whitespace was being
3463 removed from all of them, but trailing whitespace was being removed only
3464 from the last one. Now trailing whitespace is removed from each header
3465 before concatenation. Completely empty headers in a concatenation (as
3466 before) are ignored.
3468 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3469 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3471 PH/29 [Removed. This was a change that I later backed out, and forgot to
3472 correct the ChangeLog entry (that I had efficiently created) before
3473 committing the later change.]
3475 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3476 address errors (4xx response to RCPT) over the same connection as other
3477 messages routed to the same hosts. Such deliveries are always "forced",
3478 so retry times are not inspected. This resulted in far too many retries
3479 for the affected addresses. The effect occurred only when there were more
3480 hosts than the hosts_max_try setting in the smtp transport when it had
3481 the 4xx errors. Those hosts that it had tried were not added to the list
3482 of hosts for which the message was waiting, so if all were tried, there
3483 was no problem. Two fixes have been applied:
3485 (i) If there are any address or message errors in an SMTP delivery, none
3486 of the hosts (tried or untried) are now added to the list of hosts
3487 for which the message is waiting, so the message should not be a
3488 candidate for sending over the same connection that was used for a
3489 successful delivery of some other message. This seems entirely
3490 reasonable: after all the message is NOT "waiting for some host".
3491 This is so "obvious" that I'm not sure why it wasn't done
3492 previously. Hope I haven't missed anything, but it can't do any
3493 harm, as the worst effect is to miss an optimization.
3495 (ii) If, despite (i), such a delivery is accidentally attempted, the
3496 routing retry time is respected, so at least it doesn't keep
3497 hammering the server.
3499 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3500 in ${readsocket because some servers need this prod.
3502 PH/32 Added some extra debug output when updating a wait-xxx database.
3504 PH/33 The hint "could be header name not terminated by colon", which has been
3505 given for certain expansion errors for a long time, was not being given
3506 for the ${if def:h_colon_omitted{... case.
3508 PH/34 The spec says: "With one important exception, whenever a domain list is
3509 being scanned, $domain contains the subject domain." There was at least
3510 one case where this was not true.
3512 PH/35 The error "getsockname() failed: connection reset by peer" was being
3513 written to the panic log as well as the main log, but it isn't really
3514 panic-worthy as it just means the connection died rather early on. I have
3515 removed the panic log writing for the ECONNRESET error when getsockname()
3518 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3519 runs only) independently of the message's sender address. This meant
3520 that, if the 4xx error was in fact related to the sender, a different
3521 message to the same recipient with a different sender could confuse
3522 things. In particular, this can happen when sending to a greylisting
3523 server, but other circumstances could also provoke similar problems.
3524 I have changed the default so that the retry time for these errors is now
3525 based a combination of the sender and recipient addresses. This change
3526 can be overridden by setting address_retry_include_sender=false in the
3529 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3530 remote server are returned as part of bounce messages. This was not
3531 happening for LMTP over a pipe (the lmtp transport), but now it is the
3532 same for both kinds of LMTP.
3534 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3535 in header lines that were in fact CNAMEs. This is no longer the case.
3537 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3538 and queue runs started by the daemon processed all messages. This has
3539 been fixed so that -R and -S can now usefully be given with -q<time>.
3541 PH/40 Import PCRE release 6.7 (fixes some bugs).
3543 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3545 PH/42 Give an error if -q is specified more than once.
3547 PH/43 Renamed the variables $interface_address and $interface_port as
3548 $received_ip_address and $received_port, to make it clear that these
3549 values apply to message reception, and not to the outgoing interface when
3550 a message is delivered. (The old names remain recognized, of course.)
3552 PH/44 There was no timeout on the connect() call when using a Unix domain
3553 socket in the ${readsocket expansion. There now is.
3555 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3556 be meaningful with "accept".
3558 SC/01 Eximstats V1.43
3559 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3561 SC/02 Eximstats V1.44
3562 Use a glob alias rather than an array ref in the generated
3563 parser. This improves both readability and performance.
3565 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3566 Collect SpamAssassin and rejection statistics.
3567 Don't display local sender or destination tables unless
3568 there is data to show.
3569 Added average volumes into the top table text output.
3571 SC/04 Eximstats V1.46
3572 Collect data on the number of addresses (recipients)
3573 as well as the number of messages.
3575 SC/05 Eximstats V1.47
3576 Added 'Message too big' to the list of mail rejection
3577 reasons (thanks to Marco Gaiarin).
3579 SC/06 Eximstats V1.48
3580 Mainlog lines which have GMT offsets and are too short to
3581 have a flag are now skipped.
3583 SC/07 Eximstats V1.49 (Alain Williams)
3584 Added the -emptyok flag.
3586 SC/08 Eximstats V1.50
3587 Fixes for obtaining the IP address from reject messages.
3589 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3590 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3591 whitespace changes from 4.64-PH/27
3593 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3596 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3597 are found, allow negative numbers in numeric criteria)
3599 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3601 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3602 to match changes made in 4.64-PH/43
3604 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3606 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3607 "domain" is actually an IP address, and force a failure. This locks out
3608 those revolvers/nameservers that support "A-for-A" lookups, in
3609 contravention of the specifications.
3611 PH/48 When a host name was looked up from an IP address, and the subsequent
3612 forward lookup of the name timed out, the host name was left in
3613 $sender_host_name, contrary to the specification.
3615 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3616 restricted to single-key lookups, Exim was not diagnosing an error if
3617 * or *@ was used with a query-style lookup.
3619 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3621 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3622 long ago, but noone interested enough thought of it. Let's just say that
3623 the "1.1" means that there are some new functions that weren't there at
3624 some point in the past.
3626 PH/51 Error processing for expansion failure of helo_data from an smtp
3627 transport during callout processing was broken.
3629 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3630 tested/used via the -bh/-bhc/-bs options.
3632 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3633 bug, fixed in subsequent PCRE releases).
3635 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3636 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3638 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3644 SC/01 Use a glob alias rather than an array ref in eximstats generated
3645 parser. This improves both readability and performance.
3647 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3648 Don't display local sender or destination tables in eximstats unless
3649 there is data to show.
3650 Added average volumes into the eximstats top table text output.
3652 SC/03 Collect data on the number of addresses (recipients) as well
3653 as the number of messages in eximstats.
3655 TF/01 Correct an error in the documentation for the redirect router. Exim
3656 does (usually) call initgroups() when daemonizing.
3658 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3659 with consistent privilege compared to when running as a daemon.
3661 TF/03 Note in the spec that $authenticated_id is not set for local
3662 submissions from trusted users.
3664 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3665 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3667 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3668 by adding some example configuration directives to the default
3669 configuration file. A little bit of work is required to uncomment the
3670 directives and define how usernames and passwords are checked, but
3671 there is now a framework to start from.
3673 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3674 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3675 without this. I don't know how relevant this is to other LDAP libraries.
3677 PH/02 Add the verb name to the "unknown ACL verb" error.
3679 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3681 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3683 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3684 directory not expanded when it should be if an expanded home directory
3685 was set for the address (which is overridden by the transport).
3687 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3690 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3691 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3692 because it is too late at that time, and has no effect.
3694 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3695 security issue with \' (bugzilla #107). I could not use the
3696 PQescapeStringConn() function, because it needs a PGconn value as one of
3699 PH/08 When testing addresses using -bt, indicate those final addresses that
3700 are duplicates that would not cause an additional delivery. At least one
3701 person was confused, thinking that -bt output corresponded to deliveries.
3702 (Suppressing duplicates isn't a good idea as you lose the information
3703 about possibly different redirections that led to the duplicates.)
3705 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3706 systems where poll() doesn't work, in particular OS X.
3708 PH/10 Added more information to debugging output for retry time not reached.
3710 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3711 operations in malware.c.
3713 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3716 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3717 syslog_duplication set false, log lines that would normally be written
3718 both the the main log and to the reject log were not written to syslog at
3721 PH/14 In the default configuration, change the use of "message" in ACL warn
3722 statements to "add_header".
3724 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3725 not followed by a command (e.g. "seen endif").
3727 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3728 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3731 PH/17 Added extra conditions to the default value of delay_warning_condition
3735 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3736 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3737 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3740 The Auto-Submitted: and various List- headers are standardised, whereas I
3741 don't think Precedence: ever was.
3743 PH/18 Refactored debugging code in route_finduser() to show more information,
3744 in particular, the error code if getpwnam() issues one.
3746 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3747 This is apparently needed in addition to the PH/07 change above to avoid
3748 any possible encoding problems.
3750 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3751 but not after initializing Perl.
3753 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3754 output them only if debugging. By default they are written stderr,
3755 apparently, which is not desirable.
3757 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3760 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3763 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3765 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3766 authenticated or an ident call has been made. Suppress the default
3767 values for $authenticated_id and $authenticated_sender (but permit -oMai
3768 and -oMas) when testing with -bh.
3770 PH/24 Re-jigged the order of the tests in the default configuration so that the
3771 tests for valid domains and recipients precede the DNS black list and CSA
3772 tests, on the grounds that those ones are more expensive.
3774 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3775 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3776 This bug exists in every version of Exim that I still have, right back to
3779 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3780 However, an attempt to turn on case-sensitivity in a regex key by
3781 including (?-i) didn't work because the subject string was already
3782 lowercased, and the effects were non-intuitive. It turns out that a
3783 one-line patch can be used to allow (?-i) to work as expected.
3789 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3790 other effects) broke the use of negated acl sub-conditions.
3792 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3795 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3796 "Deny" causes Exim to reject the incoming connection with a 554 error.
3797 Unfortunately, if there is a major crisis, such as a disk failure,
3798 tcp-wrappers gives "deny", whereas what one would like would be some
3799 kind of temporary error. A kludge has been added to help with this.
3800 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3801 554 error is used if errno is still zero or contains ENOENT (which occurs
3802 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3805 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3807 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3808 errors. Otherwise a message that provokes a temporary error (when other
3809 messages do not) can cause a whole host to time out.
3811 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3812 addresses were routed directly to files or pipes from a redirect router.
3813 File deliveries just didn't batch; pipe deliveries might have suffered
3816 PH/06 A failure to get a lock for a hints database would erroneously always say
3817 "Failed to get write lock", even when it was really a read lock.
3819 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3820 of 0600. This has been changed to use the value of the lockfile_mode
3821 option (which defaults to 0600).
3823 PH/08 Applied small patch from the Sieve maintainer.
3825 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3826 folder from quota calculations, a direct delivery into this folder messed
3827 up the contents of the maildirsize file. This was because the regex was
3828 used only to exclude .Trash (or whatever) when the size of the mailbox
3829 was calculated. There was no check that a delivery was happening into an
3830 excluded directory. This bug has been fixed by ignoring all quota
3831 processing for deliveries into excluded directories.
3833 PH/10 Added the maildirfolder_create_regex option to appendfile.
3839 PH/01 The code for finding all the local interface addresses on a FreeBSD
3840 system running IPv6 was broken. This may well have applied to all BSD
3841 systems, as well as to others that have similar system calls. The broken
3842 code found IPv4 interfaces correctly, but gave incorrect values for the
3843 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3844 that it would not match correctly against @[] and not recognize the IPv6
3847 PH/02 The ipliteral router was not recognizing addresses of the form user@
3848 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3850 PH/03 Added disable_ipv6.
3852 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3853 decoded form, because it is most often used to construct To: headers
3854 lines in autoreplies, and the decoded form may well be syntactically
3855 invalid. However, $reply_address has leading white space removed, and all
3856 newlines turned into spaces so that the autoreply transport does not
3859 PH/05 If group was specified without a user on a router, and no group or user
3860 was specified on a transport, the group from the router was ignored.
3862 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3863 for visible compile-time settings that can be used to change these
3864 numbers, for those that want even more. Backwards compatibility with old
3865 spool files has been maintained. However, going back to a previous Exim
3866 release will lost any variables that are in spool files.
3868 PH/07 Two small changes when running in the test harness: increase delay when
3869 passing a TCP/IP connection to a new process, in case the original
3870 process has to generate a bounce, and remove special handling of
3871 127.0.0.2 (sic), which is no longer necessary.
3873 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3874 be the same on different OS.
3876 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3879 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3880 whether --show-vars was specified or not
3882 JJ/02 exipick: Added support for new ACL variable spool format introduced
3885 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3886 syntactically invalid From: or Reply-to: line, and a filter used this to
3887 generate an autoreply, and therefore failed to obtain an address for the
3888 autoreply, Exim could try to deliver to a non-existent relative file
3889 name, causing unrelated and misleading errors. What now happens is that
3890 it logs this as a hard delivery error, but does not attempt to create a
3893 PH/11 The exinext utility has a -C option for testing purposes, but although
3894 the given file was scanned by exinext itself; it wasn't being passed on
3895 when Exim was called.
3897 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3898 an end-of-file indication when reading a command response.
3900 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3901 compiled. In many other places in Exim, IPv6 addresses are always
3902 recognized, so I have changed this. It also means that IPv4 domain
3903 literals of the form [IPV4:n.n.n.n] are now always recognized.
3905 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3906 used if the router is not running as root, for example, when verifying at
3907 ACL time, or when using -bh. The debugging output from this situation was
3908 non-existent - all you got was a failure to exec. I have made two
3911 (a) Failures to set uid/gid, the current directory, or a process leader
3912 in a subprocess such as that created by queryprogram now generate
3913 suitable debugging output when -d is set.
3915 (b) The queryprogram router detects when it is not running as root,
3916 outputs suitable debugging information if -d is set, and then runs
3917 the subprocess without attempting to change uid/gid.
3919 PH/15 Minor change to Makefile for building test_host (undocumented testing
3922 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3923 additional section of a DNS packet that returns MX or SRV records.
3924 Instead, it always explicitly searches for A/AAAA records. This avoids
3925 major problems that occur when a DNS server includes only records of one
3926 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3927 fixed another bug: if SRV records were looked up and the corresponding
3928 address records were *not* found in the additional section, the port
3929 values from the SRV records were lost.
3931 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3932 using the correct key (the original address) when searching the retry
3933 rules in order to find which one to use for generating the retry hint.
3935 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3936 adding the default one. Similarly, if it contains a Reply-To: header, the
3937 errors_reply_to option, if set, is not used.
3939 PH/19 When calculating a retry time, Exim used to measure the "time since
3940 failure" by looking at the "first failed" field in the retry record. Now
3941 it does not use this if it is later than than the arrival time of the
3942 message. Instead it uses the arrival time. This makes for better
3943 behaviour in cases where some deliveries succeed, thus re-setting the
3944 "first failed" field. An example is a quota failure for a huge message
3945 when small messages continue to be delivered. Without this change, the
3946 "time since failure" will always be short, possible causing more frequent
3947 delivery attempts for the huge message than are intended.
3948 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3950 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3951 $1, $2, $3) because the numerical variables can be reset during some
3952 expansion items (e.g. "match"), thereby losing the authentication data.
3954 PH/21 Make -bV show the size of off_t variables so that the test suite can
3955 decide whether to run tests for quotas > 2G.
3957 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3958 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3959 filecount value is greater than 2G or if a quota value is greater than 2G
3960 on a system where the size of off_t is not greater than 4, a panic error
3963 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3964 never match. The debug and -bh output now contains an explicit error
3965 message indicating a malformed IPv4 address or mask.
3967 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3968 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3969 PH/23 above applies.
3971 PH/25 Do not write to syslog when running in the test harness. The only
3972 occasion when this arises is a failure to open the main or panic logs
3973 (for which there is an explicit test).
3975 PH/26 Added the /no_tell option to "control=freeze".
3977 PH/27 If a host name lookup failed very early in a connection, for example, if
3978 the IP address matched host_lookup and the reverse lookup yielded a name
3979 that did not have a forward lookup, an error message of the form "no IP
3980 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3981 could be logged. Now it outputs the IP address instead of "NULL".
3983 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3984 allows the sender and the authenticated sender to be set when
3985 submitting a message from within Exim. Since child_open_exim() is
3986 documented for local_scan(), the new function should be too.
3988 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3989 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3990 results in an empty string is now treated as unset.
3992 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3994 PH/31 Added sender_verify_fail logging option.
3996 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3997 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3998 tidied the source and removed it altogether.
4000 PH/33 When a queue run was abandoned because the load average was too high, a
4001 log line was always written; now it is written only if the queue_run log
4002 selector is set. In addition, the log line for abandonment now contains
4003 information about the queue run such as the pid. This is always present
4004 in "start" and "stop" lines but was omitted from the "abandon" line.
4006 PH/34 Omit spaces between a header name and the colon in the error message that
4007 is given when verify = headers_syntax fails (if there are lots of them,
4008 the message gets confusing).
4010 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4011 names, as there are now some PTR records that contain slashes. This check
4012 is only to protect against broken name servers that fall over on strange
4013 characters, so the fact that it applies to all lookups doesn't matter.
4015 PH/36 Now that the new test suite is complete, we can remove some of the
4016 special code in Exim that was needed for the old test suite. For example,
4017 sorting DNS records because real resolvers return them in an arbitrary
4018 order. The new test suite's fake resolver always returns records in the
4021 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4022 bounces) except when queue_only is set, to avoid logging races between
4023 the different processes.
4025 PH/38 Panic-die if .include specifies a non-absolute path.
4027 PH/39 A tweak to the "H" retry rule from its user.
4029 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4030 a label. They prevented compilation on older perls.
4032 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4033 a warning to be raised on newish perls.
4035 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4036 on queue. Changes to match documented behaviour of showing count of
4037 messages matching specified criteria.
4039 PH/40 Changed the default ident timeout from 30s to 5s.
4041 PH/41 Added support for the use of login_cap features, on those BSD systems
4042 that have them, for controlling the resources used by pipe deliveries.
4044 PH/42 The content-scanning code uses fopen() to create files in which to put
4045 message data. Previously it was not paying any attention to the mode of
4046 the files. Exim runs with umask(0) because the rest of the code creates
4047 files with open(), and sets the required mode explicitly. Thus, these
4048 files were ending up world-writeable. This was not a big issue, because,
4049 being within the spool directory, they were not world-accessible. I have
4050 created a function called modefopen, which takes an additional mode
4051 argument. It sets umask(777), creates the file, chmods it to the required
4052 mode, then resets the umask. All the relevant calls to fopen() in the
4053 content scanning code have been changed to use this function.
4055 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4056 to 24 hours. This avoids potential overflow problems when processing G
4057 and H retry rules. I suspect nobody ever tinkers with this value.
4059 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4061 PH/45 When the plaintext authenticator is running as a client, the server's
4062 challenges are checked to ensure they are valid base64 strings. By
4063 default, the authentication attempt is cancelled if an invalid string is
4064 received. Setting client_ignore_invalid_base64 true ignores these errors.
4065 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4066 they are received. Thus, the responses can be made to depend on the
4067 challenges. If an invalid string is ignored, an empty string is placed in
4070 PH/46 Messages that are created by the autoreply transport now contains a
4071 References: header, in accordance with RFCs 2822 and 3834.
4073 PH/47 Added authenticated_sender_force to the smtp transport.
4075 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4077 PH/49 Installed latest patch from the Sieve maintainer.
4079 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4080 was also set, the appendfile transport was unnecessarily scanning a
4081 directory of message files (e.g. for maildir delivery) to find the count
4082 of files (along with the size), even though it did not need this
4083 information. It now does the scan only if it needs to find either the
4084 size of the count of files.
4086 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4088 PH/52 Two bugs concerned with error handling when the smtp transport is
4091 (i) Exim was not creating retry information for temporary errors given
4092 for individual recipients after the DATA command when the smtp transport
4093 was used in LMTP mode. This meant that they could be retried too
4094 frequently, and not timed out correctly.
4096 (ii) Exim was setting the flag that allows error details to be returned
4097 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4098 recipients that were returned after the DATA command.
4100 PH/53 This is related to PH/52, but is more general: for any failing address,
4101 when detailed error information was permitted to be returned to the
4102 sender, but the error was temporary, then after the final timeout, only
4103 "retry timeout exceeded" was returned. Now it returns the full error as
4104 well as "retry timeout exceeded".
4106 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4107 do this, and (what is worse) MTAs that accept it.
4109 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4110 will now be deprecated.
4112 PH/56 New os.c-cygwin from the Cygwin maintainer.
4114 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4115 formats (previously only available in exim formats via -bpr, -bpru,
4116 and -bpra. Now also available in native and exiqgrep formats)
4118 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4119 with very large, slow to parse queues
4121 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4123 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4125 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4126 responses to authentication challenges, though it was showing the
4127 challenges; (ii) I've removed the CR characters from the debug output for
4130 PH/58 Allow for the insertion of a newline as well as a space when a string
4131 is turned into more than one encoded-word during RFC 2047 encoding. The
4132 Sieve code now uses this.
4134 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4135 data_4xx, lost_connection, tls_required.
4137 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4138 message was being sent as an SMTP response.
4140 PH/61 Add -l and -k options to exicyclog.
4142 PH/62 When verifying, if an address was redirected to one new address, so that
4143 verification continued, and the new address failed or deferred after
4144 having set something in $address_data, the value of $address_data was not
4145 passed back to the ACL. This was different to the case when no
4146 redirection occurred. The value is now passed back in both cases.
4148 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4149 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4150 use login_cap.h, so on its own it isn't the distinguishing feature. The
4151 new name refers directly to the setclassresources() function.
4153 PH/65 Added configuration files for NetBSD3.
4155 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4157 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4158 is preferred over IPv4.
4160 PH/68 The bounce_return_message and bounce_return_body options were not being
4161 honoured for bounces generated during the reception of non-SMTP messages.
4162 In particular, this applied to messages rejected by the ACL. This bug has
4163 been fixed. However, if bounce_return_message is true and bounce_return_
4164 body is false, the headers that are returned for a non-SMTP message
4165 include only those that have been read before the error was detected.
4166 (In the case of an ACL rejection, they have all been read.)
4168 PH/69 The HTML version of the specification is now built in a directory called
4169 spec_html instead of spec.html, because the latter looks like a path with
4170 a MIME-type, and this confuses some software.
4172 PH/70 Catch two compiler warnings in sieve.c.
4174 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4175 function verify_get_ident() calls ip_connect() to connect a socket, but
4176 if the "connect()" function timed out, ip_connect() used to close the
4177 socket. However, verify_get_ident() also closes the socket later, and in
4178 between Exim writes to the log, which may get opened at this point. When
4179 the socket was closed in ip_connect(), the log could get the same file
4180 descriptor number as the socket. This naturally causes chaos. The fix is
4181 not to close the socket in ip_connect(); the socket should be closed by
4182 the function that creates it. There was only one place in the code where
4183 this was missing, in the iplookup router, which I don't think anybody now
4184 uses, but I've fixed it anyway.
4186 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4187 well as to direct DNS lookups. Otherwise the handling of names in host
4188 lists is inconsistent and therefore confusing.
4194 PH/01 Two changes to the default runtime configuration:
4196 (1) Move the checks for relay_from_hosts and authenticated clients from
4197 after to before the (commented out) DNS black list checks.
4199 (2) Add control=submission to the relay_from_hosts and authenticated
4200 clients checks, on the grounds that messages accepted by these
4201 statements are most likely to be submissions.
4203 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4205 (1) Generate an error if the third argument for the ${prvs expansion is
4208 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4211 (3) Reset the variables that are obtained from the first argument of
4212 ${prvscheck and used in the second argument before leaving the code,
4213 because their memory is reclaimed, so using them afterwards may do
4216 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4217 one (it's much easier than Tom thought :-).
4219 (5) Because of (4), we can now allow for the use of $prvscheck_result
4220 inside the third argument.
4222 PH/03 For some reason, the default setting of PATH when running a command from
4223 a pipe transport was just "/usr/bin". I have changed it to
4226 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4227 anything to be listed in the output from -bV.
4229 PH/05 When a filter generated an autoreply, the entire To: header line was
4230 quoted in the delivery log line, like this:
4232 => >A.N.Other <ano@some.domain> <original@ddress> ...
4234 This has been changed so that it extracts the operative address. There
4235 may be more than one such address. If so, they are comma-separated, like
4238 => >ano@some.domain,ona@other.domain <original@ddress> ...
4240 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4241 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4242 not being looked up in the rDNS to get a host name, Exim was showing the
4243 IP address twice in Received: lines, even though the IP addresses were
4244 identical. For example:
4246 Received: from [1.2.3.4] (helo=[1.2.3.4])
4248 However, if the real host name was known, it was omitting the HELO data
4249 if it matched the actual IP address. This has been tidied up so that it
4250 doesn't show the same IP address twice.
4252 PH/07 When both +timestamp and +memory debugging was on, the value given by
4253 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4254 called by the debug printing, thereby overwriting the timestamp buffer.
4255 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4257 PH/08 When the original message was included in an autoreply transport, it
4258 always said "this is a copy of the message, including all the headers",
4259 even if body_only or headers_only was set. It now gives an appropriate
4262 PH/09 Applied a patch from the Sieve maintainer which:
4264 o fixes some comments
4265 o adds the (disabled) notify extension core
4266 o adds some debug output for the result of if/elsif tests
4267 o points to the current vacation draft in the documentation
4268 and documents the missing references header update
4272 o fixes a bug in processing the envelope test (when testing
4273 multiple envelope elements, the last element determined the
4276 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4277 Electronic Mail") by including:
4279 Auto-submitted: auto-generated
4281 in the messages that it generates (bounce messages and others, such as
4282 warnings). In the case of bounce messages for non-SMTP messages, there was
4283 also a typo: it was using "Auto_submitted" (underscore instead of
4284 hyphen). Since every message generated by Exim is necessarily in response
4285 to another message, thes have all been changed to:
4287 Auto-Submitted: auto-replied
4289 in accordance with these statements in the RFC:
4291 The auto-replied keyword:
4293 - SHOULD be used on messages sent in direct response to another
4294 message by an automatic process,
4296 - MUST NOT be used on manually-generated messages,
4298 - MAY be used on Delivery Status Notifications (DSNs) and Message
4299 Disposition Notifications (MDNs),
4301 - MUST NOT be used on messages generated by automatic or periodic
4302 processes, except for messages which are automatic responses to
4305 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4306 to the default Received: header definition.
4308 PH/12 Added log selector acl_warn_skipped (default on).
4310 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4311 variables because (a) they are in the wrong storage pool and (b) even if
4312 they were copied, it wouldn't work properly because of the caching.
4314 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4315 checking when decoding. Apparently there are clients that generate
4316 overlong encoded strings. Why am I not surprised?
4318 PH/15 If the first argument of "${if match_address" was not empty, but did not
4319 contain an "@" character, Exim crashed. Now it writes a panic log message
4320 and treats the condition as false.
4322 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4324 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4325 extension "envelope-auth". The code is finished and in agreement with
4326 other implementations, but there is no documentation so far and in fact,
4327 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4328 not changing the active code.
4330 Print executed "if" and "elsif" statements when debugging is used. This
4331 helps a great deal to understand what a filter does.
4333 Document more things not specified clearly in RFC3028. I had all this
4334 sorted out, when out of a sudden new issues came to my mind. Oops."
4336 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4339 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4340 canonical form (as documented). However, after a host name lookup from
4341 the IP address, check_host() was doing a simple string comparison with
4342 addresses acquired from the DNS when checking that the found name did
4343 have the original IP as one of its addresses. Since any found IPv6
4344 addresses are likely to be in abbreviated form, the comparison could
4345 fail. Luckily, there already exists a function for doing the comparison
4346 by converting both addresses to binary, so now that is used instead of
4347 the text comparison.
4349 PH/20 There was another similar case to PH/19, when a complete host name was
4350 given in a host list; looking up its IP address could give an abbreviated
4351 form, whereas the current host's name might or might not be abbreviated.
4352 The same fix has been applied.
4358 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4359 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4362 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4363 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4365 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4367 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4368 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4369 Also, for "Auto-Submitted", treat anything other than "no" as
4370 non-personal, in accordance with RFC 3834. (Previously it treated
4371 anything starting "auto-" as non-personal.)
4373 TF/01 The control=submission/name=... option had a problem with syntax
4374 errors if the name included a slash character. The /name= option
4375 now slurps the rest of the string, so it can include any characters
4376 but it must come last in the list of options (after /sender_retain
4379 PH/05 Some modifications to the interface to the fake nameserver for the new
4387 TK/01 Added the "success_on_redirect" address verification option. See
4388 NewStuff for rationale and an example.
4390 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4392 PH/02 Patch to exigrep to allow it to work on syslog lines.
4394 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4395 fread() to skip over the body file's header line, because in Cygwin the
4396 header line is locked and is inaccessible.
4398 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4399 co-exist for some time) to make it clear that it is the Exim ID that is
4400 referenced, not the Message-ID: header line.
4402 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4403 string_format() function, because snprintf() does not exist on all
4406 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4407 ${stat: expansion item.
4409 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4410 protocol synchronization error", to keep the pedants happy.
4412 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4413 well as for IRIX systems, when gcc is being used. See the host.c source
4416 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4418 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4421 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4422 transport and to the smtp transport in LMTP mode.
4424 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4426 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4427 run a filter in a subprocess. This could lead to confusion in subsequent
4428 lookups in the parent process. There should also be a search_tidyup() at
4429 the end of the subprocess.
4431 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4432 only if the host matched helo_try_verify_hosts, which caused the
4433 verification to occur when the EHLO/HELO command was issued. The ACL just
4434 tested the remembered result. Now, if a previous verification attempt has
4435 not happened, "verify = helo" does it there and then.
4437 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4439 TK/03 Fix log output including CR from clamd.
4441 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4442 exist provoked a memory error which could cause a segfault.
4444 PH/15 Installed PCRE 6.2
4446 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4448 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4449 of the problem. Specifically, suggested +O2 rather than +O1 for the
4452 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4454 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4455 record, the port was not correctly logged when the outgoing_port log
4456 selector was set (it logged the transort's default port).
4458 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4459 fallback_hosts, and "hosts" in the smtp transport.
4461 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4462 host errors such as "Connection refused".
4464 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4465 authentication with radiusclient 0.4.9:
4467 - Error returned from rc_read_config was caught wrongly
4468 - Username/password not passed on to radius server due to wrong length.
4470 The presumption is that some radiusclient API changes for 4.51/PH/17
4471 were not taken care of correctly. The code is still untested by me (my
4472 Linux distribution still has 0.3.2 of radiusclient), but it was
4473 contributed by a Radius user.
4475 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4476 expanding the "port" option of the smtp transport.
4478 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4479 while reading a MIME header. Thanks to Tom Hughes for a patch.
4481 PH/24 Include config.h inside local_scan.h so that configuration settings are
4484 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4485 that in an ACL for RCPT (for example), you can examine exactly what was
4488 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4489 commands, but it was not correctly comparing the address with the actual
4490 client host address. Thus, it would show the EHLO address in Received:
4491 header lines when this was not necessary.
4493 PH/27 Added the % operator to ${eval:}.
4495 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4496 it should be ignoring failures (because with -C, for example, it has lost
4497 privilege). It wasn't ignoring creation failures other than "already
4500 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4503 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4504 because an input line was too long, either on its own, or by virtue of
4505 too many continuations, the temporary file was not being removed, and the
4506 return code was incorrect.
4508 PH/31 Missing "BOOL" in function definition in filtertest.c.
4510 PH/32 Applied Sieve patches from the maintainer.
4512 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4514 PH/33 Added "verify = not_blind".
4516 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4517 Local/Makefile (with some defaults set). These are used in built scripts
4518 such as exicyclog, but they have never been used in the exim_install
4519 script (though there are many overriding facilities there). I have
4520 arranged that the exim_install script now takes note of these two
4523 PH/35 Installed configuration files for Dragonfly.
4525 PH/36 When a locally submitted message by a trusted user did not contain a
4526 From: header, and the sender address was obtained from -f or from an SMTP
4527 MAIL command, and the trusted user did not use -F to supply a sender
4528 name, $originator_name was incorrectly used when constructing a From:
4529 header. Furthermore, $originator_name was used for submission mode
4530 messages from external hosts without From: headers in a similar way,
4531 which is clearly wrong.
4533 PH/37 Added control=suppress_local_fixups.
4535 PH/38 When log_selector = +received_sender was set, and the addition of the
4536 sender made the log line's construction buffer exactly full, or one byte
4537 less than full, an overflow happened when the terminating "\n" was
4540 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4541 when the result of a list match is failure because a DNS lookup failed.
4543 PH/40 RM_COMMAND is now used in the building process.
4545 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4546 the "build-* directories that it finds.
4548 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4549 address in a domain literal was a prefix of an interface address.
4551 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4552 when verifying a sender address, unless rewrite_headers is false.
4554 PH/44 Wrote a long comment about why errors_to addresses are verified as
4555 recipients, not senders.
4557 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4558 the ratelimit ACL was added.
4560 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4562 PH/46 Added extra information about PostgreSQL errors to the error string.
4564 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4565 avoiding the need to install special zones in a real server. This is
4566 backwards compatible; if it can't find the fake resolver, it drops back.
4567 Thus, both old and new test suites can be run.
4569 TF/02 Added util/ratelimit.pl
4571 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4572 clock is set back in time.
4574 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4575 Candler <B.Candler@pobox.com>.
4577 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4578 OK for addresses that are the result of redirections.
4580 PH/48 A number of further additions for the benefit of the new test suite,
4581 including a fake gethostbyname() that interfaces to the fake DNS resolver
4584 TF/06 The fix for widen_domains has also been applied to qualify_single and
4585 search_parents which are the other dnslookup options that can cause
4588 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4591 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4593 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4594 service" if it encountered an MX record with an empty target hostname.
4595 The message is now "an MX or SRV record indicated no SMTP service".
4597 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4598 if the DNS of the sending site is misconfigured. This is quite a
4599 common situation. This change restores the behaviour of treating a
4600 helo verification defer as a failure.
4602 PH/51 If self=fail was set on a router, the bounce message did not include the
4603 actual error message.
4609 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4611 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4612 command itself ended in error, the underlying message about the transport
4613 filter timeout was being overwritten with the pipe command error. Now the
4614 underlying error message should be appended to the second error message.
4616 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4618 PH/02 Reduce the amount of output that "make" produces by default. Full output
4619 can still be requested.
4621 PH/03 The warning log line about a condition test deferring for a "warn" verb
4622 was being output only once per connection, rather than after each
4623 occurrence (because it was using the same function as for successful
4624 "warn" verbs). This seems wrong, so I have changed it.
4626 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4627 it should not have, which might have caused a crash in the right
4628 circumstances, but probably never did.
4630 PH/04 Installed a modified version of Tony Finch's patch to make submission
4631 mode fix the return path as well as the Sender: header line, and to
4632 add a /name= option so that you can make the user's friendly name appear
4635 TF/03 Added the control = fakedefer ACL modifier.
4637 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4638 Mark Lowes for thorough testing.
4640 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4642 TK/03 Merged latest SRS patch from Miles Wilton.
4644 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4645 with the definition in sysexits.h (which is #included earlier).
4646 Fortunately, Exim does not actually use EX_OK. The code used to try to
4647 preserve the sysexits.h value, by assuming that macro definitions were
4648 scanned for macro replacements. I have been disabused of this notion,
4649 so now the code just undefines EX_OK before #including unistd.h.
4651 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4652 in the smtp transport. When a block could not be written in a single
4653 write() function, the timeout was being re-applied to each part-write.
4654 This seems wrong - if the receiver was accepting one byte at a time it
4655 would take for ever. The timeout is now adjusted when this happens. It
4656 doesn't have to be particularly precise.
4658 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4659 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4661 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4662 without a domain if the check to <postmaster@domain> fails.
4664 SC/01 Eximstats: added -xls and the ability to specify output files
4665 (patch written by Frank Heydlauf).
4667 SC/02 Eximstats: use FileHandles for outputting results.
4669 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4671 SC/04 Eximstats: fixed display of large numbers with -nvr option
4673 SC/05 Eximstats: fixed merging of reports with empty tables.
4675 SC/06 Eximstats: added the -include_original_destination flag
4677 SC/07 Eximstats: removed tabs and trailing whitespace.
4679 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4681 TK/06 MBOX spool code: Add real "From " MBOX separator line
4682 so the .eml file is really in mbox format (even though
4683 most programs do not really care). Patch from Alex Miller.
4685 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4686 The latter is generated from $received_to and is only set if the
4687 message has one envelope recipient. SA can use these headers,
4688 obviously out-of-the-box. Patch from Alex Miller.
4690 PH/08 The ${def test on a variable was returning false if the variable's
4691 value was "0", contrary to what the specification has always said!
4692 The result should be true unless the variable is empty.
4694 PH/09 The syntax error of a character other than { following "${if
4695 def:variable_name" (after optional whitespace) was not being diagnosed.
4696 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4697 accidental colon was present, for example, could give incorrect results.
4699 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4700 result is used (not including appendfile, where other changes are about
4703 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4704 This involved changing a lot of size variables from int to off_t. It
4705 should work with maildirs and everything.
4707 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4708 spamd dying while we are connected to it.
4710 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4713 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4714 function for BDB 4.3.
4716 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4718 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4719 the sources for PCRE have been re-arranged and more files are now
4722 PH/15 The code I had for printing potentially long long variables in PH/11
4723 above was not the best (it lost precision). The length of off_t variables
4724 is now inspected at build time, and an appropriate printing format (%ld
4725 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4726 to be "long long int" or "long int". This is needed for the internal
4727 formatting function string_vformat().
4729 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4730 the configuration file to be ":syslog", then the script "guesses" where
4731 the logs files are, rather than using the compiled in default. In our
4732 case the guess is not the same as the compiled default, so the script
4733 suddenly stopped working when I started to use syslog. The patch checks
4734 to see if log_file_path is "". If so, it attempts to read it from exim
4735 with no configuration file to get the compiled in version, before it
4736 falls back to the previous guessing code."
4738 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4739 implementing BATV in an Exim configuration. See NewStuff for the gory
4742 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4743 Makefile that are specific to HP-UX.
4745 PH/18 If the "use_postmaster" option was set for a recipient callout together
4746 with the "random" option, the postmaster address was used as the MAIL
4747 FROM address for the random test, but not for the subsequent recipient
4748 test. It is now used for both.
4750 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4751 patch removes a few documentation additions to RFC 3028, because the
4752 latest draft now contains them. It adds the new en;ascii-case comparator
4753 and a new error check for 8bit text in MIME parts. Comparator and
4754 require names are now matched exactly. I enabled the subaddress
4755 extension, but it is not well tested yet (read: it works for me)."
4757 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4758 rework some of the code of TK/09 above to avoid the hardwired use of
4759 "%lld" and "long long". Replaced the call to snprintf() with a call to
4762 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4763 records point to non-existent hosts", "retry timeout exceeded", and
4764 "retry time not reached for any host after a long failure period".
4766 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4767 experimental DomainKeys support:
4769 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4770 (2) On an error such as an illegally used "control", the wrong name for
4771 the control was given.
4773 These problems did NOT occur unless DomainKeys support was compiled.
4775 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4777 PH/24 Added ${if match_ip condition.
4779 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4780 databases so that it will be absolutely obvious if a crash occurs in the
4781 DB library. This is a regular occurrence (often caused by mis-matched
4784 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4785 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4786 picked up on a user's system that detects such things. There doesn't seem
4787 to be a gcc warning option for this - only an attribute that has to be
4788 put on the function's prototype. It seems that in Fedora Core 4 they have
4789 set this on a number of new functions. No doubt there will be more in due
4792 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4793 specify a transport. However, if an address that was verified by such a
4794 router was the subject of a callout, Exim crashed because it tried to
4795 read the rcpt_include_affixes from the non-existent transport. Now it
4796 just assumes that the setting of that option is false. This bug was
4797 introduced by 4.51/PH/31.
4799 PH/28 Changed -d+all to exclude +memory, because that information is very
4800 rarely of interest, but it makes the output a lot bigger. People tend to
4801 do -d+all out of habit.
4803 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4804 code in os-type was giving problems when libc.so lives in lib64, like on
4807 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4808 aren't the modern standard, and it seems that some systems' include files
4809 don't always have them. Exim was already checking for some of the newer
4810 ones like T_AAAA, and defining it itself. I've added checks for all the
4811 record types that Exim uses.
4813 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4814 not automatically generating a new one, as it is supposed to. This
4815 prevented TLS from working. If the file did exist, but contained invalid
4816 data, a new version was generated, as expected. It was only the case of a
4817 non-existent file that was broken.
4819 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4820 with a change in libdomainkeys > 0.64.
4822 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4823 from DNS. If the selector record carries the flag, it now has
4824 precedence over the domain-wide flag.
4826 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4828 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4829 the use of a port name that isn't defined in /etc/services) occurred, the
4830 message was deferred as in a normal delivery, and thus remained on the
4831 spool, instead of being failed because of the mua_wrapper setting. This
4832 is now fixed, and I tidied up some of the mua_wrapper messages at the
4835 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4836 the messages in a hash of arrays rather than using individual hashes.
4837 This is a bit cleaner and results in dramatic memory savings, albeit
4838 at a slight CPU cost.
4840 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4841 as requested by Marc Sherman.
4843 SC/10 Eximstats: added histograms for user specified patterns as requested
4846 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4848 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4849 fopen() in the content-scanning modules that did not already have it.
4855 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4856 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4858 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4860 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4862 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4863 to test Sieve filters that use "vacation".
4865 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4866 that changes the way the GnuTLS parameters are stored in the cache file.
4867 The new format can be generated externally. For backward compatibility,
4868 if the data in the cache doesn't make sense, Exim assumes it has read an
4869 old-format file, and it generates new data and writes a new file. This
4870 means that you can't go back to an older release without removing the
4873 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4874 work if there are any delivery delays because "one_time" forces the
4875 parent to be marked "delivered", so its unseen clone is never tried
4876 again. For this reason, Exim now forbids the simultaneous setting of
4879 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4880 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4881 entry if you want to know the details. The fix, however, neglected to
4882 consider the case when local delivery batching is involved. The test for
4883 "previously delivered" was not happening when checking to see if an
4884 address could be batched with a previous (undelivered) one; under
4885 certain circumstances this could lead to multiple deliveries to the same
4888 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4889 in its include files, and this causes problems building Exim.
4891 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4892 header_syntax) but Exim was just ignoring anything given after a slash.
4893 In particular, this caused confusion with an attempt to use "verify =
4894 reverse_host_lookup/defer_ok". An error is now given when options are
4895 supplied for verify items that do not have them. (Maybe reverse_host_
4896 lookup should have a defer_ok option, but that's a different point.)
4898 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4899 defined by RFC 821) to 2048, because there were problems with some AUTH
4900 commands, and RFC 1869 says the size should be increased for extended
4901 SMTP commands that take arguments.
4903 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4906 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4907 "unknown" error; now it says that the functionality isn't in the binary.
4909 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4910 an address' error message when a string expansion fails (syntax or
4911 whatever). Otherwise the password may appear in the log. Following change
4912 PH/42 below, there is no longer a chance of it appearing in a bounce
4915 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4917 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4918 of its addresses was ever tried. (Bugzilla bug #2.)
4920 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4921 the result incorrectly in the debug output. (It correctly added a newline
4922 to what was transported.)
4924 TF/01 Added $received_time.
4926 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4927 commented out examples of how to interface to a virus scanner and to
4928 SpamAssassin. Also added commented examples of av_scanner and
4929 spamd_address settings.
4931 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4932 and controls are allowed in which ACLs. There were a couple of minor
4933 errors. Some of the entries in the conditions table (which is a table of
4934 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4935 negation of where the condition IS allowed.
4937 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4939 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4940 header file does not have a version number, so I've had to invent a new
4941 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4942 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4943 radiusclient), but it was contributed by a Radius user.
4945 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4946 files or directories, for OpenSSL.
4948 PH/19 When an Exim process that is running as root has to create an Exim log
4949 file, it does so in a subprocess that runs as exim:exim so as to get the
4950 ownership right at creation (otherwise, other Exim processes might see
4951 the file with the wrong ownership). There was no test for failure of this
4952 fork() call, which would lead to the process getting stuck as it waited
4953 for a non-existent subprocess. Forks do occasionally fail when resources
4954 run out. I reviewed all the other calls to fork(); they all seem to check
4957 PH/20 When checking for unexpected SMTP input at connect time (before writing
4958 the banner), Exim was not dealing correctly with a non-positive return
4959 from the read() function. If the client had disconnected by this time,
4960 the result was a log entry for a synchronization error with an empty
4961 string after "input=" when read() returned zero. If read() returned -1
4962 (an event I could not check), uninitialized data bytes were printed.
4963 There were reports of junk text (parts of files, etc) appearing after
4966 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4968 PH/22 Added support for macro redefinition, and (re)definition in between
4969 driver and ACL definitions.
4971 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4972 forgetting to use the resulting value; it was using the unexpanded value.
4974 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4975 hadn't been configured. The fix is from Juergen Kreileder, who
4976 understands it better than I do:
4978 "Here's what I see happening with three configured cyrus_sasl
4979 authenticators configured (plain, login, cram-md5):
4981 On startup auth_cyrus_sasl_init() gets called for each of these.
4982 This means three calls to sasl_listmech() without a specified mech_list.
4983 => SASL tests which mechs of all available mechs actually work
4984 => three warnings about OTP not working
4985 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4987 With the patch, sasl_listmech() also gets called three times. But now
4988 SASL's mech_list option is set to the server_mech specified in the the
4989 authenticator. Or in other words, the answer from sasl_listmech()
4990 gets limited to just the mech you're testing for (which is different
4992 => the return list contains just 'plain' or 'login', 'cram-md5' or
4993 nothing depending on the value of ob->server_mech.
4995 I've just tested the patch: Authentication still works fine,
4996 unavailable mechs specified in the exim configuration are still
4997 caught, and the auth.log warnings about OTP are gone."
4999 PH/25 When debugging is enabled, the contents of the command line are added
5000 to the debugging output, even when log_selector=+arguments is not
5003 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5004 answer is "GNU", and only if the return is "GNU/something" is the answer
5007 PH/27 $acl_verify_message is now set immediately after the failure of a
5008 verification in an ACL, and so is available in subsequent modifiers. In
5009 particular, the message can be preserved by coding like this:
5011 warn !verify = sender
5012 set acl_m0 = $acl_verify_message
5014 Previously, $acl_verify_message was set only while expanding "message"
5015 and "log_message" when a very denied access.
5017 PH/28 Modified OS/os.c-Linux with
5019 -#ifndef OS_LOAD_AVERAGE
5020 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5022 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5023 nomenclature these days.)
5025 PH/29 Installed patch from the Sieve maintainer that adds the options
5026 sieve_useraddress and sieve_subaddress to the redirect router.
5028 PH/30 In these circumstances:
5029 . Two addresses routed to the same list of hosts;
5030 . First host does not offer TLS;
5031 . First host accepts first address;
5032 . First host gives temporary error to second address;
5033 . Second host offers TLS and a TLS session is established;
5034 . Second host accepts second address.
5035 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5036 and peerdn, if requested) that were in fact used only for the second
5039 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5040 attention to any local part prefix or suffix that was matched by the
5041 router that accepted the address. It now behaves in the same way as it
5042 does for delivery: the affixes are removed from the local part unless
5043 rcpt_include_affixes is set on the transport.
5045 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5046 timeout during the DATA phase of an incoming message.
5048 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5049 applied a patch sanctioned by the Sieve maintainer.
5051 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5052 the uid or gid is negative. A case of a negative gid caused this to be
5053 noticed. The fix allows for either to be negative.
5055 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5056 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5059 PH/36 The change PH/12 above was broken. Fixed it.
5061 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5062 the grounds that routing the same address twice would always produce the
5063 same answer. This might have been true once, but it is certainly no
5064 longer true now. Routing a child address may depend on the previous
5065 routing that produced that child. Some complicated redirection strategies
5066 went wrong when messages had multiple recipients, and made Exim's
5067 behaviour dependent on the order in which the addresses were given.
5069 I have moved the duplicate checking until after the routing is complete.
5070 Exim scans the addresses that are assigned to local and remote
5071 transports, and removes any duplicates. This means that more work will be
5072 done, as duplicates will always all be routed, but duplicates are
5073 presumably rare, so I don't expect this is of any significance.
5075 For deliveries to pipes, files, and autoreplies, the duplicate checking
5076 still happens during the routing process, since they are not going to be
5079 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5080 It corrects a timeout issue with spamd. This is Ian's comment: "The
5081 background is that sometimes spamd either never reads data from a
5082 connection it has accepted, or it never writes response data. The exiscan
5083 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5084 blindly assumes that writes won't block so it may never time out."
5086 PH/39 Allow G after quota size as well as K and M.
5088 PH/40 The value set for $authenticated_id in an authenticator may not contain
5089 binary zeroes or newlines because the value is written to log lines and
5090 to spool files. There was no check on this. Now the value is run through
5091 the string_printing() function so that such characters are converted to
5092 printable escape sequences.
5094 PH/41 $message_linecount is a new variable that contains the total number of
5095 lines in the message. Compare $body_linecount, which is the count for the
5098 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5099 bounce and delay warning messages, except in certain special cases, which
5102 (a) An SMTP error message from a remote host;
5103 (b) A message specified in a :fail: redirection;
5104 (c) A message specified in a "fail" command in a system filter;
5105 (d) A message specified in a FAIL return from the queryprogram router;
5106 (e) A message specified by the cannot_route_message router option.
5108 In these cases only, Exim does include the error details in bounce and
5109 warning messages. There are also a few cases where bland messages such
5110 as "unrouteable address" or "local delivery error" are given.
5112 PH/43 $value is now also set for the "else" part of a ${run expansion.
5114 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5115 being worked on, but at least Exim now implements the latest version to
5118 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5119 process to complete was treated as a delivery failure, a timeout while
5120 writing the message to the pipe was logged, but erroneously treated as a
5121 successful delivery. Such timeouts include transport filter timeouts. For
5122 consistency with the overall process timeout, these timeouts are now
5123 treated as errors, giving rise to delivery failures by default. However,
5124 there is now a new Boolean option for the pipe transport called
5125 timeout_defer, which, if set TRUE, converts the failures into defers for
5126 both kinds of timeout. A transport filter timeout is now identified in
5129 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5130 systems where "make" and "gmake" are different, calling "gmake" at top
5131 level broke things. I've arranged for the value of $(MAKE) to be passed
5132 from the Makefile to this script so that it can call the same version of
5136 A note about Exim versions 4.44 and 4.50
5137 ----------------------------------------
5139 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5140 changes of various kinds. As a consequence, a big documentation update was
5141 needed. This delayed the release for rather longer than seemed good, especially
5142 in the light of a couple of (minor) security issues. Therefore, the changes
5143 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5144 release. So 4.44 and 4.50 are in effect two different branches that both start
5147 I have left the 4.50 change log unchanged; it contains all the changes since
5148 4.43. The change log for 4.44 is below; many of its items are identical to
5149 those for 4.50. This seems to be the most sensible way to preserve the
5150 historical information.
5156 1. Minor wording change to the doc/README.SIEVE file.
5158 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5159 computation of the current number of files was incorrect.
5161 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5162 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5165 4. Give more explanation in the error message when the command for a transport
5166 filter fails to execute.
5168 5. There are several places where Exim runs a non-Exim command in a
5169 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5170 was being done only for the command run by the queryprogram router. It is
5171 now done for all such subprocesses. The other cases are: ${run, transport
5172 filters, and the commands run by the lmtp and pipe transports.
5174 6. Added CONFIGURE_GROUP build-time option.
5176 7. Some older OS have a limit of 256 on the maximum number of file
5177 descriptors. Exim was using setrlimit() to set 1000 as a large value
5178 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5179 systems. I've change it so that if it can't get 1000, it tries for 256.
5181 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5182 was an oversight, and furthermore, ever since the addition of extra
5183 controls (e.g. 4.43/32), the checks on when to allow different forms of
5184 "control" were broken. There should now be diagnostics for all cases when a
5185 control that does not make sense is encountered.
5187 9. Added the /retain_sender option to "control=submission".
5189 10. $recipients is now available in the predata ACL (oversight).
5191 11. Tidy the search cache before the fork to do a delivery from a message
5192 received from the command line. Otherwise the child will trigger a lookup
5193 failure and thereby defer the delivery if it tries to use (for example) a
5194 cached ldap connection that the parent has called unbind on.
5196 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5197 of $address_data from the recipient verification was clobbered by the
5198 sender verification.
5200 13. The value of address_data from a sender verification is now available in
5201 $sender_address_data in subsequent conditions in the ACL statement.
5203 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5205 15. Added a new option "connect=<time>" to callout options, to set a different
5208 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5209 was its contents. (It was OK if the option was not defined at all.)
5211 17. A "Completed" log line is now written for messages that are removed from
5212 the spool by the -Mrm option.
5214 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5215 information about exactly what failed.
5217 19. Added -dd to debug only the daemon process.
5219 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5220 handles timeouts, both on the server side and network timeouts. Renamed the
5221 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5223 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5224 It is now set to "smtps".
5226 22. $host_address is now set to the target address during the checking of
5227 ignore_target_hosts.
5229 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5230 being passed; this would have caused $sender_host_name to have been used if
5231 matching the list had actually called for a host name (not very likely,
5232 since this list is usually IP addresses). A host name is now passed as
5235 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5236 code that specifies a non-restarting handler (typically sigaction() in
5237 modern systems) in an attempt to fix a rare and obscure crash bug.
5239 25. Narrowed the window for a race in the daemon that could cause it to ignore
5240 SIGCHLD signals. This is not a major problem, because they are used only to
5241 wake it up if nothing else does.
5243 26. A malformed maildirsize file could cause Exim to calculate negative values
5244 for the mailbox size or file count. Odd effects could occur as a result.
5245 The maildirsize information is now recalculated if the size or filecount
5248 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5249 support for a long time. Removed HAVE_SYS_VFS_H.
5251 28. Installed the latest version of exipick from John Jetmore.
5253 29. In an address list, if the pattern was not a regular expression, an empty
5254 subject address (from a bounce message) matched only if the pattern was an
5255 empty string. Non-empty patterns were not even tested. This was the wrong
5256 because it is perfectly reasonable to use an empty address as part of a
5257 database query. An empty address is now tested by patterns that are
5258 lookups. However, all the other forms of pattern expect the subject to
5259 contain a local part and a domain, and therefore, for them, an empty
5260 address still always fails if the pattern is not itself empty.
5262 30. Exim went into a mad DNS loop when attempting to do a callout where the
5263 host was specified on an smtp transport, and looking it up yielded more
5264 than one IP address.
5266 31. Re-factored the code for checking spool and log partition space into a
5267 function that finds that data and another that does the check. The former
5268 is then used to implement four new variables: $spool_space, $log_space,
5269 $spool_inodes, and $log_inodes.
5271 32. The RFC2047 encoding function was originally intended for short strings
5272 such as real names; it was not keeping to the 75-character limit for
5273 encoded words that the RFC imposes. It now respects the limit, and
5274 generates multiple encoded words if necessary. To be on the safe side, I
5275 have increased the buffer size for the ${rfc2047: expansion operator from
5278 33. It is now permitted to omit both strings after an "if" condition; if the
5279 condition is true, the result is "true". As before, when the second string
5280 is omitted, a false condition yields an empty string. This makes it less
5281 cumbersome to write custom ACL and router conditions.
5283 34. Failure to deliver a bounce message always caused it to be frozen, even if
5284 there was an errors_to setting on the router. The errors_to setting is now
5287 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5288 canonical form (fully expanded) before being placed in
5289 $sender_host_address.
5291 36. The table in the code that translates DNS record types into text (T_A to
5292 "A" for instance) was missing entries for NS and CNAME. It is just possible
5293 that this could have caused confusion if both these types were looked up
5294 for the same domain, because the text type is used as part of Exim's
5295 per-process caching. But the chance of anyone hitting this buglet seems
5298 37. The dnsdb lookup has been extended in a number of ways.
5300 (1) There is a new type, "zns", which walks up the domain tree until it
5301 finds some nameserver records. It should be used with care.
5303 (2) There is a new type, "mxh", which is like "mx" except that it returns
5304 just the host names, not the priorities.
5306 (3) It is now possible to give a list of domains (or IP addresses) to be
5307 looked up. The behaviour when one of the lookups defers can be
5308 controlled by a keyword.
5310 (4) It is now possible to specify the separator character for use when
5311 multiple records are returned.
5313 38. The dnslists ACL condition has been extended: it is now possible to supply
5314 a list of IP addresses and/or domains to be looked up in a particular DNS
5317 39. Added log_selector=+queue_time_overall.
5319 40. When running the queue in the test harness, wait just a tad after forking a
5320 delivery process, to get repeatability of debugging output.
5322 41. Include certificate and key file names in error message when GnuTLS fails
5323 to set them up, because the GnuTLS error message doesn't include the name
5324 of the failing file when there is a problem reading it.
5326 42. Allow both -bf and -bF in the same test run.
5328 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5330 44. The "Exiscan patch" is now merged into the mainline Exim source.
5332 45. Sometimes the final signoff response after QUIT could fail to get
5333 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5334 < 0 before doing a fflush(). This bug looks as though it goes back to the
5335 introduction of TLS in release 3.20, but "sometimes" must have been rare
5336 because the tests only now provoked it.
5338 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5339 (this can affect the format of dates).
5341 47. exim_tidydb, when checking for the continued existence of a message for
5342 which it has found a message-specific retry record, was not finding
5343 messages that were in split spool directories. Consequently, it was
5344 deleting retry records that should have stayed in existence.
5346 48. Steve fixed some bugs in eximstats.
5348 49. The SPA authentication driver was not abandoning authentication and moving
5349 on to the next authenticator when an expansion was forced to fail,
5350 contradicting the general specification for all authenticators. Instead it
5351 was generating a temporary error. It now behaves as specified.
5353 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5354 (the order specifies the preference for clients). The order is now AES256,
5355 AES128, 3DES, ARCFOUR128.
5357 51. Small patch to Sieve code - explicitly set From: when generating an
5360 52. Exim crashed if a remote delivery caused a very long error message to be
5361 recorded - for instance if somebody sent an entire SpamAssassin report back
5362 as a large number of 550 error lines. This bug was coincidentally fixed by
5363 increasing the size of one of Exim's internal buffers (big_buffer) that
5364 happened as part of the Exiscan merge. However, to be on the safe side, I
5365 have made the code more robust (and fixed the comments that describe what
5368 53. Now that there can be additional text after "Completed" in log lines (if
5369 the queue_time_overall log selector is set), a one-byte patch to exigrep
5370 was needed to allow it to recognize "Completed" as not the last thing in
5373 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5374 patch that reportedly fixes this has been added. I am not expert enough to
5375 create a test for it. This is what the patch creator wrote:
5377 "I found a little strange behaviour of ldap code when working with
5378 Windows 2003 AD Domain, where users was placed in more than one
5379 Organization Units. When I tried to give exim partial DN, the exit code
5380 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5381 But simultaneously result of request was absolutely normal ldap result,
5382 so I produce this patch..."
5384 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5385 REFERENCE, so I have modified the code to exclude the patch when that macro
5388 55. Some experimental protocols are using DNS PTR records for new purposes. The
5389 keys for these records are domain names, not reversed IP addresses. The
5390 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5391 leaves it alone. Component reversal etc. now happens only for IP addresses.
5394 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5396 57. Double the size of the debug message buffer (to 2048) so that more of very
5397 long debug lines gets shown.
5399 58. The exicyclog utility now does better if the number of log files to keep
5400 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5402 59. Two changes related to the smtp_active_hostname option:
5404 (1) $smtp_active_hostname is now available as a variable.
5405 (2) The default for smtp_banner uses $smtp_active_hostname instead
5406 of $primary_hostname.
5408 60. The host_aton() function is supposed to be passed a string that is known
5409 to be a valid IP address. However, in the case of IPv6 addresses, it was
5410 not checking this. This is a hostage to fortune. Exim now panics and dies
5411 if the condition is not met. A case was found where this could be provoked
5412 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5413 components; fortuitously, this particular loophole had already been fixed
5414 by change 4.50/55 above.
5416 If there are any other similar loopholes, the new check in host_aton()
5417 itself should stop them being exploited. The report I received stated that
5418 data on the command line could provoke the exploit when Exim was running as
5419 exim, but did not say which command line option was involved. All I could
5420 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5421 running as the user.
5424 61. There was a buffer overflow vulnerability in the SPA authentication code
5425 (which came originally from the Samba project). I have added a test to the
5426 spa_base64_to_bits() function which I hope fixes it.
5429 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5430 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5432 63. The daemon start-up calls getloadavg() while still root for those OS that
5433 need the first call to be done as root, but it missed one case: when
5434 deliver_queue_load_max is set with deliver_drop_privilege. This is
5435 necessary for the benefit of the queue runner, because there is no re-exec
5436 when deliver_drop_privilege is set.
5438 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5439 This has been fixed.
5441 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5442 was in use, was not putting the data itself into the right store pool;
5443 consequently, it could be overwritten for a subsequent message in the same
5444 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5447 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5449 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5450 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5451 the function were treating the return as a boolean value, which happened to
5452 work because 0=false and not-0=true, but is not correct code.
5454 68. The host_aton() function was not handling scoped IPv6 addresses (those
5455 with, for example, "%eth0" on the end) correctly.
5457 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5458 negated items (that is, ~something) in unsigned ints. Some compilers
5459 apparently mutter when there is no cast.
5461 70. If an address verification called from an ACL failed, and did not produce a
5462 user-specific message (i.e. there was only a "system" message), nothing was
5463 put in $acl_verify_message. In this situation, it now puts the system
5466 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5467 session; change 4.31/43 added the unwanted input to the log line - except
5468 that it did not do this in the start of session case. It now does.
5470 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5471 This is wrong and can cause the other end to generate a synchronization
5472 error if it is another Exim or anything else that does the synchronization
5473 check. A QUIT command is no longer sent after a timeout.
5475 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5476 during host lookups.
5478 74. The defer_ok option of callout verification was not working if it was used
5479 when verifying addresses in header lines, that is, for this case:
5481 verify = header_sender/callout=defer_ok
5483 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5484 those file descriptors could be used for SMTP connections. If anything
5485 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5486 could be sent to the SMTP client, causing chaos. The daemon now opens
5487 stdin, stdout, and stderr to /dev/null when it puts itself into the
5490 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5491 log by default. The user can override this with suitable Perl magic.
5493 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5494 the log message when discard triggers, was not working for the DATA ACL or
5495 for the non-SMTP ACL.
5497 78. Error message wording change in sieve.c.
5499 79. If smtp_accept_max_per_host was set, the number of connections could be
5500 restricted to fewer than expected, because the daemon was trying to set up
5501 a new connection before checking whether the processes handling previous
5502 connections had finished. The check for completed processes is now done
5503 earlier. On busy systems, this bug wouldn't be noticed because something
5504 else would have woken the daemon, and it would have reaped the completed
5507 80. If a message was submitted locally by a user whose login name contained one
5508 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5509 It caused a spool format error. I have fixed the spool reading code. A
5510 related problem was that the "from" clause in the Received: line became
5511 illegal because of the space(s). It is now covered by ${quote_local_part.
5513 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5516 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5518 83. Patch from Sieve maintainer to fix unterminated string problem in
5519 "vacation" handling.
5521 84. Some minor changes to the Linux configuration files to help with other
5522 OS variants using glibc.
5524 85. One more patch for Sieve to update vacation handling to latest spec.
5527 ----------------------------------------------------
5528 See the note above about the 4.44 and 4.50 releases.
5529 ----------------------------------------------------
5535 1. Change 4.43/35 introduced a bug that caused file counts to be
5536 incorrectly computed when quota_filecount was set in an appendfile
5539 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5540 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5543 3. Give more explanation in the error message when the command for a transport
5544 filter fails to execute.
5546 4. There are several places where Exim runs a non-Exim command in a
5547 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5548 was being done only for the command run by the queryprogram router. It is
5549 now done for all such subprocesses. The other cases are: ${run, transport
5550 filters, and the commands run by the lmtp and pipe transports.
5552 5. Some older OS have a limit of 256 on the maximum number of file
5553 descriptors. Exim was using setrlimit() to set 1000 as a large value
5554 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5555 systems. I've change it so that if it can't get 1000, it tries for 256.
5557 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5558 was an oversight, and furthermore, ever since the addition of extra
5559 controls (e.g. 4.43/32), the checks on when to allow different forms of
5560 "control" were broken. There should now be diagnostics for all cases when a
5561 control that does not make sense is encountered.
5563 7. $recipients is now available in the predata ACL (oversight).
5565 8. Tidy the search cache before the fork to do a delivery from a message
5566 received from the command line. Otherwise the child will trigger a lookup
5567 failure and thereby defer the delivery if it tries to use (for example) a
5568 cached ldap connection that the parent has called unbind on.
5570 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5571 of $address_data from the recipient verification was clobbered by the
5572 sender verification.
5574 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5575 was its contents. (It was OK if the option was not defined at all.)
5577 11. A "Completed" log line is now written for messages that are removed from
5578 the spool by the -Mrm option.
5580 12. $host_address is now set to the target address during the checking of
5581 ignore_target_hosts.
5583 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5584 being passed; this would have caused $sender_host_name to have been used if
5585 matching the list had actually called for a host name (not very likely,
5586 since this list is usually IP addresses). A host name is now passed as
5589 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5590 code that specifies a non-restarting handler (typically sigaction() in
5591 modern systems) in an attempt to fix a rare and obscure crash bug.
5593 15. Narrowed the window for a race in the daemon that could cause it to ignore
5594 SIGCHLD signals. This is not a major problem, because they are used only to
5595 wake it up if nothing else does.
5597 16. A malformed maildirsize file could cause Exim to calculate negative values
5598 for the mailbox size or file count. Odd effects could occur as a result.
5599 The maildirsize information is now recalculated if the size or filecount
5602 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5603 support for a long time. Removed HAVE_SYS_VFS_H.
5605 18. Updated exipick to current release from John Jetmore.
5607 19. Allow an empty sender to be matched against a lookup in an address list.
5608 Previously the only cases considered were a regular expression, or an
5611 20. Exim went into a mad DNS lookup loop when doing a callout where the
5612 host was specified on the transport, if the DNS lookup yielded more than
5615 21. The RFC2047 encoding function was originally intended for short strings
5616 such as real names; it was not keeping to the 75-character limit for
5617 encoded words that the RFC imposes. It now respects the limit, and
5618 generates multiple encoded words if necessary. To be on the safe side, I
5619 have increased the buffer size for the ${rfc2047: expansion operator from
5622 22. Failure to deliver a bounce message always caused it to be frozen, even if
5623 there was an errors_to setting on the router. The errors_to setting is now
5626 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5627 canonical form (fully expanded) before being placed in
5628 $sender_host_address.
5630 24. Updated eximstats to version 1.33
5632 25. Include certificate and key file names in error message when GnuTLS fails
5633 to set them up, because the GnuTLS error message doesn't include the name
5634 of the failing file when there is a problem reading it.
5636 26. Expand error message when OpenSSL has problems setting up cert/key files.
5639 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5640 (this can affect the format of dates).
5642 28. exim_tidydb, when checking for the continued existence of a message for
5643 which it has found a message-specific retry record, was not finding
5644 messages that were in split spool directories. Consequently, it was
5645 deleting retry records that should have stayed in existence.
5647 29. eximstats updated to version 1.35
5648 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5649 1.35 - bugfix such that pie charts by volume are generated correctly
5651 30. The SPA authentication driver was not abandoning authentication and moving
5652 on to the next authenticator when an expansion was forced to fail,
5653 contradicting the general specification for all authenticators. Instead it
5654 was generating a temporary error. It now behaves as specified.
5656 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5657 (the order specifies the preference for clients). The order is now AES256,
5658 AES128, 3DES, ARCFOUR128.
5660 31. Small patch to Sieve code - explicitly set From: when generating an
5663 32. Exim crashed if a remote delivery caused a very long error message to be
5664 recorded - for instance if somebody sent an entire SpamAssassin report back
5665 as a large number of 550 error lines. This bug was coincidentally fixed by
5666 increasing the size of one of Exim's internal buffers (big_buffer) that
5667 happened as part of the Exiscan merge. However, to be on the safe side, I
5668 have made the code more robust (and fixed the comments that describe what
5671 33. Some experimental protocols are using DNS PTR records for new purposes. The
5672 keys for these records are domain names, not reversed IP addresses. The
5673 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5674 leaves it alone. Component reversal etc. now happens only for IP addresses.
5677 34. The host_aton() function is supposed to be passed a string that is known
5678 to be a valid IP address. However, in the case of IPv6 addresses, it was
5679 not checking this. This is a hostage to fortune. Exim now panics and dies
5680 if the condition is not met. A case was found where this could be provoked
5681 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5682 components; fortuitously, this particular loophole had already been fixed
5683 by change 4.50/55 or 4.44/33 above.
5685 If there are any other similar loopholes, the new check in host_aton()
5686 itself should stop them being exploited. The report I received stated that
5687 data on the command line could provoke the exploit when Exim was running as
5688 exim, but did not say which command line option was involved. All I could
5689 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5690 running as the user.
5693 35. There was a buffer overflow vulnerability in the SPA authentication code
5694 (which came originally from the Samba project). I have added a test to the
5695 spa_base64_to_bits() function which I hope fixes it.
5698 36. The daemon start-up calls getloadavg() while still root for those OS that
5699 need the first call to be done as root, but it missed one case: when
5700 deliver_queue_load_max is set with deliver_drop_privilege. This is
5701 necessary for the benefit of the queue runner, because there is no re-exec
5702 when deliver_drop_privilege is set.
5704 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5705 was in use, was not putting the data itself into the right store pool;
5706 consequently, it could be overwritten for a subsequent message in the same
5707 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5710 38. Sometimes the final signoff response after QUIT could fail to get
5711 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5712 < 0 before doing a fflush(). This bug looks as though it goes back to the
5713 introduction of TLS in release 3.20, but "sometimes" must have been rare
5714 because the tests only now provoked it.
5720 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5721 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5722 Now it returns an integer. A number of calls were still expecting a T/F
5723 return. Fortuitously, in all cases, the tests worked in OK situations,
5724 which is the norm. However, things would have gone wrong on any write
5725 failures on the smtp file descriptor. This function is used when sending
5726 messages over SMTP and also when doing verify callouts.
5728 2. When Exim is called to do synchronous delivery of a locally submitted
5729 message (the -odf or -odi options), it no longer closes stderr before doing
5732 3. Implemented the mua_wrapper option.
5734 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5736 5. Implemented the functions header_remove(), header_testname(),
5737 header_add_at_position(), and receive_remove_recipient(), and exported them
5740 6. If an ACL "warn" statement specified the addition of headers, Exim already
5741 inserted X-ACL-Warn: at the start if there was no header name. However, it
5742 was not making this test for the second and subsequent header lines if
5743 there were newlines in the string. This meant that an invalid header could
5744 be inserted if Exim was badly configured.
5746 7. Allow an ACL "warn" statement to add header lines at the start or after all
5747 the Received: headers, as well as at the end.
5749 8. Added the rcpt_4xx retry error code.
5751 9. Added postmaster_mailfrom=xxx to callout verification option.
5753 10. Added mailfrom=xxxx to the callout verification option, for verify=
5756 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5757 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5759 12. Inserted some casts to stop certain compilers warning when using pointer
5760 differences as field lengths or precisions in printf-type calls (mostly
5761 affecting debugging statements).
5763 13. Added optional readline() support for -be (dynamically loaded).
5765 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5766 same clock tick as a message's arrival, so that its received time was the
5767 same as the "first fail" time on the retry record, and that message
5768 remained on the queue past the ultimate address timeout, every queue runner
5769 would try a delivery (because it was past the ultimate address timeout) but
5770 after another failure, the ultimate address timeout, which should have then
5771 bounced the address, did not kick in. This was a "< instead of <=" error;
5772 in most cases the first failure would have been in the next clock tick
5773 after the received time, and all would be well.
5775 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5776 being recognized when the domain list was tested by the match_domain
5777 condition in an expansion string.
5779 16. Added the ${str2b64: operator.
5781 17. Exim was always calling setrlimit() to set a large limit for the number of
5782 processes, without checking whether the existing limit was already
5783 adequate. (It did check for the limit on file descriptors.) Furthermore,
5784 errors from getrlimit() and setrlimit() were being ignored. Now they are
5785 logged to the main and panic logs, but Exim does carry on, to try to do its
5786 job under whatever limits there are.
5788 18. Imported PCRE 5.0.
5790 19. Trivial typo in log message " temporarily refused connection" (the leading
5793 20. If the log selector return_path_on_delivery was set and an address was
5794 redirected to /dev/null, the delivery process crashed because it assumed
5795 that a return path would always be set for a "successful" delivery. In this
5796 case, the whole delivery is bypassed as an optimization, and therefore no
5799 21. Internal re-arrangement: the function for sending a challenge and reading
5800 a response while authentication was assuming a zero-terminated challenge
5801 string. It's now changed to take a pointer and a length, to allow for
5802 binary data in such strings.
5804 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5806 23. Exim was not respecting finduser_retries when seeking the login of the
5807 uid under which it was called; it was always trying 10 times. (The default
5808 setting of finduser_retries is zero.) Also, it was sleeping after the final
5809 failure, which is pointless.
5811 24. Implemented tls_on_connect_ports.
5813 25. Implemented acl_smtp_predata.
5815 26. If the domain in control=submission is set empty, Exim assumes that the
5816 authenticated id is a complete email address when it generates From: or
5817 Sender: header lines.
5819 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5820 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5821 chown and chgrp in /bin and hostname in /usr/bin.
5823 28. Exim was keeping the "process log" file open after each use, just as it
5824 does for the main log. This opens the possibility of it remaining open for
5825 long periods when the USR1 signal hits a daemon. Occasional processlog
5826 errors were reported, that could have been caused by this. Anyway, it seems
5827 much more sensible not to leave this file open at all, so that is what now
5830 29. The long-running daemon process does not normally write to the log once it
5831 has entered its main loop, and it closes the log before doing so. This is
5832 so that log files can straightforwardly be renamed and moved. However,
5833 there are a couple of unusual error situations where the daemon does write
5834 log entries, and I had neglected to close the log afterwards.
5836 30. The text of an SMTP error response that was received during a remote
5837 delivery was being truncated at 512 bytes. This is too short for some of
5838 the long messages that one sometimes sees. I've increased the limit to
5841 31. It is now possible to make retry rules that apply only when a message has a
5842 specific sender, in particular, an empty sender.
5844 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5845 it possible to be selective about when SMTP synchronization is enforced.
5847 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5849 32. Implemented hosts_connection_nolog.
5851 33. Added an ACL for QUIT.
5853 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5856 35. Added mailbox_size and mailbox_filecount to appendfile.
5858 36. Added control = no_multiline_responses to ACLs.
5860 37. There was a bug in the logic of the code that waits for the clock to tick
5861 in the case where the clock went backwards by a substantial amount such
5862 that the microsecond fraction of "now" was more than the microsecond
5863 fraction of "then" (but the whole seconds number was less).
5865 38. Added support for the libradius Radius client library this is found on
5866 FreeBSD (previously only the radiusclient library was supported).
5872 1. When certain lookups returned multiple values in the form name=value, the
5873 quoting of the values was not always being done properly. Specifically:
5874 (a) If the value started with a double quote, but contained no whitespace,
5876 (b) If the value contained whitespace other than a space character (i.e.
5877 tabs or newlines or carriage returns) it was not quoted.
5878 This fix has been applied to the mysql and pgsql lookups by writing a
5879 separate quoting function and calling it from the lookup code. The fix
5880 should probably also be applied to nisplus, ibase and oracle lookups, but
5881 since I cannot test any of those, I have not disturbed their existing code.
5883 2. A hit in the callout cache for a specific address caused a log line with no
5884 reason for rejecting RCPT. Now it says "Previous (cached) callout
5885 verification failure".
5887 3. There was an off-by-one bug in the queryprogram router. An over-long
5888 return line was truncated at 256 instead of 255 characters, thereby
5889 overflowing its buffer with the terminating zero. As well as fixing this, I
5890 have increased the buffer size to 1024 (and made a note to document this).
5892 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5893 when Exim is waiting for an SMTP response from a remote server, Exim
5894 restarts its select() call on the socket, thereby resetting its timeout.
5895 This is not a problem when such interrupts are rare. Somebody set up a cron
5896 job to run exiwhat every 2 minutes, which is less than the normal select()
5897 timeout (5 or 10 minutes). This meant that the select() timeout never
5898 kicked in because it was always reset. I have fixed this by comparing the
5899 time when an interrupt arrives with the time at the start of the first call
5900 to select(). If more time than the timeout has elapsed, the interrupt is
5901 treated as a timeout.
5903 5. Some internal re-factoring in preparation for the addition of Sieve
5904 extensions (by MH). In particular, the "personal" test is moved to a
5905 separate function, and given an option for scanning Cc: and Bcc: (which is
5906 not set for Exim filters).
5908 6. When Exim created an email address using the login of the caller as the
5909 local part (e.g. when creating a From: or Sender: header line), it was not
5910 quoting the local part when it contained special characters such as @.
5912 7. Installed new OpenBSD configuration files.
5914 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5915 try to make them clearer.
5917 9. Callout options, other than the timeout value, were being ignored when
5918 verifying sender addresses in header lines. For example, when using
5920 verify = header_sender/callout=no_cache
5922 the cache was (incorrectly) being used.
5924 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5925 only the Cygwin environment.
5927 11. When return_path_on_delivery was set as a log selector, if different remote
5928 addresses in the same message used different return paths and parallel
5929 remote delivery occurred, the wrong values would sometimes be logged.
5930 (Whenever a remote delivery process finished, the return path value from
5931 the most recently started remote delivery process was logged.)
5933 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5934 lines when AUTH and/or TLS are in use. This is the "received protocol"
5935 field. Exim used to use "asmtp" for authenticated SMTP, without any
5936 indication (in the protocol name) for TLS use. Now it follows the RFC and
5937 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5938 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5939 names appear in log lines as well as in Received: header lines.
5941 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5942 extensions, and comparison tests, and to fix some bugs.
5944 14. Changes to the "personal" filter test:
5946 (1) The test was buggy in that it was just doing the equivalent of
5947 "contains" tests on header lines. For example, if a user's address was
5948 anne@some.where, the "personal" test would incorrectly be true for
5950 To: susanne@some.where
5952 This test is now done by extracting each address from the header in turn,
5953 and checking the entire address. Other tests that are part of "personal"
5954 are now done using regular expressions (for example, to check local parts
5955 of addresses in From: header lines).
5957 (2) The list of non-personal local parts in From: addresses has been
5958 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5959 taken from the Sieve specification recommendations.
5961 (3) If the message contains any header line starting with "List-" it is
5962 treated as non-personal.
5964 (4) The test for "circular" in the Subject: header line has been removed
5965 because it now seems ill-conceived.
5967 15. Minor typos in src/EDITME comments corrected.
5969 16. Installed latest exipick from John Jetmore.
5971 17. If headers_add on a router specified a text string that was too long for
5972 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5973 of string_sprintf() is now avoided.
5975 18. $message_body_size was not set (it was always zero) when running the DATA
5976 ACL and the local_scan() function.
5978 19. For the "mail" command in an Exim filter, no default was being set for
5979 the once_repeat time, causing a random time value to be used if "once" was
5980 specified. (If the value happened to be <= 0, no repeat happened.) The
5981 default is now 0s, meaning "never repeat". The "vacation" command was OK
5982 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5983 (I found it when inspecting the code).
5985 20. There is now an overall timeout for performing a callout verification. It
5986 defaults to 4 times the callout timeout, which applies to individual SMTP
5987 commands during the callout. The overall timeout applies when there is more
5988 than one host that can be tried. The timeout is checked before trying the
5989 next host. This prevents very long delays if there are a large number of
5990 hosts and all are timing out (e.g. when the network connections are timing
5991 out). The value of the overall timeout can be changed by specifying an
5992 additional sub-option for "callout", called "maxwait". For example:
5994 verify = sender/callout=5s,maxwait=20s
5996 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5997 to the end before writing, but this should make it even safer).
5999 22. Exim was forgetting that it had advertised PIPELINING for the second and
6000 subsequent messages on an SMTP connection. It was also not resetting its
6001 memory on STARTTLS and an internal HELO.
6003 23. When Exim logs an SMTP synchronization error within a session, it now
6004 records whether PIPELINING has been advertised or not.
6006 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6007 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6008 rather than long int.
6010 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6012 26. Added the never_mail option to autoreply.
6018 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6019 crash if the getsockname() call failed; this can happen if a connection is
6020 closed very soon after it is established. The problem was simply in the
6021 order in which certain operations were done, causing Exim to try to write
6022 to the SMTP stream before it had set up the file descriptor. The bug has
6023 been fixed by making things happen in the correct order.
6029 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6030 before the connection was closed, thus losing the rejection response.
6032 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6033 some early Solaris releases, but causes trouble in current releases where
6034 socklen_t is defined.
6036 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6039 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6042 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6043 that releasing the top of it at the end releases what was used for sub-
6044 expansions (unless the block got too big). However, discard this block if
6045 the first thing is a variable or header, so that we can use its block when
6046 it is dynamic (useful for very large $message_headers, for example).
6048 6. Lookups now cache *every* query, not just the most recent. A new, separate
6049 store pool is used for this. It can be recovered when all lookup caches are
6050 flushed. Lookups now release memory at the end of their result strings.
6051 This has involved some general refactoring of the lookup sources.
6053 7. Some code has been added to the store_xxx() functions to reduce the amount
6054 of flapping under certain conditions.
6056 8. log_incoming_interface used to affect only the <= reception log lines. Now
6057 it causes the local interface and port to be added to several more SMTP log
6058 lines, for example "SMTP connection from", and rejection lines.
6060 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6062 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6064 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6065 could be overwritten at the end of the current message (or the start of a
6066 new message if it was set in a HELO ACL). The value is now preserved for
6067 the duration of the SMTP connection.
6069 12. If a transport had a headers_rewrite setting, and a matching header line
6070 contained an unqualified address, that address was qualified, even if it
6071 did not match any rewriting rules. The underlying bug was that the values
6072 of the flags that permit the existence of unqualified sender and recipient
6073 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6074 non-local messages, and by -bnq for local messages) were not being
6075 preserved with the message after it was received.
6077 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6078 "next input=" as part of the text comprising the host identity instead of
6079 the correct text. The code was using the same buffer for two different
6080 strings. However, depending on which order the printing function evaluated
6081 its arguments, the bug did not always show up. Under Linux, for example, my
6082 test suite worked just fine.
6084 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6085 doesn't work with some older versions of Perl. It has been changed to "my",
6086 which in any case is probably the better facility to use.
6088 15. A really picky compiler found some instances of statements for creating
6089 error messages that either had too many or two few arguments for the format
6092 16. The size of the buffer for calls to the DNS resolver has been increased
6093 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6094 for addresses that have a lot of PTR records. This alleviates a problem; it
6095 does not fully solve it.
6097 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6098 buffer now truncates the list and logs the incident, which is the same
6099 action as happens when Exim is looking up a host name and its aliases.
6100 Previously in this situation something unpredictable would happen;
6101 sometimes it was "internal error: store_reset failed".
6103 18. If a server dropped the connection unexpectedly when an Exim client was
6104 using GnuTLS and trying to read a response, the client delivery process
6105 crashed while trying to generate an error log message.
6107 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6108 string, for example:
6110 warn message = H1: something\nH2: something
6112 the text was added as a single header line from Exim's point of view
6113 though it ended up OK in the delivered message. However, searching for the
6114 second and subsequent header lines using $h_h2: did not work. This has been
6115 fixed. Similarly, if a system filter added multiple headers in this way,
6116 the routers could not see them.
6118 20. Expanded the error message when iplsearch is called with an invalid key to
6119 suggest using net-iplsearch in a host list.
6121 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6122 ACLs are no longer actually imposed (and a message to that effect is
6125 22. If a "gecos" field in a passwd entry contained escaped characters, in
6126 particular, if it contained a \" sequence, Exim got it wrong when building
6127 a From: or a Sender: header from that name. A second bug also caused
6128 incorrect handling when an unquoted " was present following a character
6129 that needed quoting.
6131 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6132 was not being matched caselessly.
6134 24. Arranged for all hyphens in the exim.8 source to be escaped with
6137 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6138 properly. Recipient callout cache records were still being keyed to include
6139 the sender, even when use_sender was set false. This led to far more
6140 callouts that were necessary. The sender is no longer included in the key
6141 when use_sender is false.
6143 26. Added "control = submission" modifier to ACLs.
6145 27. Added the ${base62d: operator to decode base 62 numbers.
6147 28. dnsdb lookups can now access SRV records.
6149 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6150 the configuration file.
6152 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6153 (-v) mode. This makes the output for a verbose queue run more intelligible.
6155 31. Added a use_postmaster feature to recipient callouts.
6157 32. Added the $body_zerocount variable, containing the number of binary zero
6158 bytes in the message body.
6160 33. The time of last modification of the "new" subdirectory is now used as the
6161 "mailbox time last read" when there is a quota error for a maildir
6164 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6166 35. Added +ignore_unknown as a special item in host lists.
6168 36. Code for decoding IPv6 addresses in host lists is now included, even if
6169 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6170 address was recognized as an IP address, but was then not correctly decoded
6171 into binary, causing unexpected and incorrect effects when compared with
6178 1. Very minor rewording of debugging text in manualroute to say "list of
6179 hosts" instead of "hostlist".
6181 2. If verify=header_syntax was set, and a header line with an unqualified
6182 address (no domain) and a large number of spaces between the end of the
6183 name and the colon was received, the reception process suffered a buffer
6184 overflow, and (when I tested it) crashed. This was caused by some obsolete
6185 code that should have been removed. The fix is to remove it!
6187 3. When running in the test harness, delay a bit after writing a bounce
6188 message to get a bit more predictability in the log output.
6190 4. Added a call to search_tidyup() just before forking a reception process. In
6191 theory, someone could use a lookup in the expansion of smtp_accept_max_
6192 per_host which, without the tidyup, could leave open a database connection.
6194 5. Added the variables $recipient_data and $sender_data which get set from a
6195 lookup success in an ACL "recipients" or "senders" condition, or a router
6196 "senders" option, similar to $domain_data and $local_part_data.
6198 6. Moved the writing of debug_print from before to after the "senders" test
6201 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6202 problems for message scanning, either using a data ACL, or using
6203 local_scan() because the Received: header was not generated till after they
6204 were called (in order to set the time as the time of reception completion).
6205 I have revised the way this works. The header is now generated after the
6206 body is received, but before the ACL or local_scan() are called. After they
6207 are run, the timestamp in the header is updated.
6213 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6214 before starting a queue runner without re-exec. This happened only when
6215 deliver_drop_privilege was set or when the Exim user was set to root. The
6216 effect of the bug was that timeouts during subsequent deliveries caused
6217 crashes instead of being properly handled. The handler is now left at its
6218 default (and expected) setting.
6220 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6221 message, again when deliver_drop_privilege is set or Exim is run as root.
6222 The bug described in (1) was not present in this case, but the tidying up
6223 of the other signals was missing. I have made the two cases consistent.
6225 3. The ignore_target_hosts setting on a manualroute router was being ignored
6226 for hosts that were looked up using the /MX notation.
6228 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6231 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6232 operated on the sender address. After changing the $sender_address to <>
6233 for the sender address verify, Exim was re-instated it as the original
6234 (before rewriting) address, but remembering that it had rewritten it, so it
6235 wasn't rewriting it again. This bug also had the effect of breaking the
6236 sender address verification caching when the sender address was rewritten.
6238 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6239 This has been changed so that if the ip literal address matches
6240 ignore_target_hosts, the router declines.
6242 7. Added expansion conditions match_domain, match_address, and match_local_
6243 part (NOT match_host).
6245 8. The placeholder for the Received: header didn't have a length field set.
6247 9. Added code to Exim itself and to exim_lock to test for a specific race
6248 condition that could lead to file corruption when using MBX delivery. The
6249 issue is with the lockfile that is created in /tmp. If this file is removed
6250 after a process has opened it but before that process has acquired a lock,
6251 there is the potential for a second process to recreate the file and also
6252 acquire a lock. This could lead to two Exim processes writing to the file
6253 at the same time. The added code performs the same test as UW imapd; it
6254 checks after acquiring the lock that its file descriptor still refers to
6255 the same named file.
6257 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6258 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6261 11. Added the smtp_active_hostname option. If used, this will typically be made
6262 to depend on the incoming interface address. Because $interface_address is
6263 not set up until the daemon has forked a reception process, error responses
6264 that can happen earlier (such as "too many connections") no longer contain
6267 12. If an expansion in a condition on a "warn" statement fails because a lookup
6268 defers, the "warn" statement is abandoned, and the next ACL statement is
6269 processed. Previously this caused the whole ACL to be aborted.
6271 13. Added the iplsearch lookup type.
6273 14. Added ident_timeout as a log selector.
6275 15. Added tls_certificate_verified as a log selector.
6277 16. Added a global option tls_require_ciphers (compare the smtp transport
6278 option of the same name). This controls incoming TLS connections.
6280 17. I finally figured out how to make tls_require_ciphers do a similar thing
6281 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6282 before starting the TLS session.
6284 18. Tabs are now shown as \t in -bP output.
6286 19. If the log selector return_path_on_delivery was set, Exim crashed when
6287 bouncing a message because it had too many Received: header lines.
6289 20. If two routers both had headers_remove settings, and the first one included
6290 a superfluous trailing colon, the final name in the first list and the
6291 first name in the second list were incorrectly joined into one item (with a
6292 colon in the middle).
6298 1. Added -C and -D options to the exinext utility, mainly to make it easier
6299 to include in the automated testing, but these could be helpful when
6300 multiple configurations are in use.
6302 2. The exinext utility was not formatting the output nicely when there was
6303 an alternate port involved in the retry record key, nor when there was a
6304 message id as well (for retries that were specific to a specific message
6305 and a specific host). It was also confused by IPv6 addresses, because of
6306 the additional colons they contain. I have fixed the IPv4 problem, and
6307 patched it up to do a reasonable job for IPv6.
6309 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6310 delivery, the log line now contains "pipelined" if PIPELINING was used.
6312 4. An SMTP transport process used to panic and die if the bind() call to set
6313 an explicit outgoing interface failed. This has been changed; it is now
6314 treated in the same way as a connect() failure.
6316 5. A reference to $sender_host_name in the part of a conditional expansion
6317 that was being skipped was still causing a DNS lookup. This no longer
6320 6. The def: expansion condition was not recognizing references to header lines
6321 that used bh_ and bheader_.
6323 7. Added the _cache feature to named lists.
6325 8. The code for checking quota_filecount in the appendfile transport was
6326 allowing one more file than it should have been.
6328 9. For compatibility with Sendmail, the command line option
6336 and sets the incoming protocol and host name (for trusted callers). The
6337 host name and its colon can be omitted when only the protocol is to be set.
6338 Note the Exim already has two private options, -pd and -ps, that refer to
6339 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6340 "s", but I don't think that's a major issue.
6342 10. A number of refactoring changes to the code, none of which should affect
6345 (a) The number of logging options was getting close to filling up the
6346 32-bit word that was used as a bit map. I have split them into two classes:
6347 those that are passed in the argument to log_write(), and those that are
6348 only ever tested independently outside of that function. These are now in
6349 separate 32-bit words, so there is plenty of room for expansion again.
6350 There is no change in the user interface or the logging behaviour.
6352 (b) When building, for example, log lines, the code previously used a
6353 macro that called string_cat() twice, in order to add two strings. This is
6354 not really sufficiently general. Furthermore, there was one instance where
6355 it was actually wrong because one of the argument was used twice, and in
6356 one call a function was used. (As it happened, calling the function twice
6357 did not affect the overall behaviour.) The macro has been replaced by a
6358 function that can join an arbitrary number of extra strings onto a growing
6361 (c) The code for expansion conditions now uses a table and a binary chop
6362 instead of a serial search (which was left over from when there were very
6363 few conditions). Also, it now recognizes conditions like "pam" even when
6364 the relevant support is not compiled in: a suitably worded error message is
6365 given if an attempt is made to use such a condition.
6367 11. Added ${time_interval:xxxxx}.
6369 12. A bug was causing one of the ddress fields not to be passed back correctly
6370 from remote delivery subprocesses. The field in question was not being
6371 subsequently used, so this caused to problems in practice.
6373 13. Added new log selectors queue_time and deliver_time.
6375 14. Might have fixed a bug in maildirsizefile handling that threw up
6376 "unexpected character" debug warnings, and recalculated the data
6377 unnecessarily. In any case, I expanded the warning message to give more
6380 15. Added the message "Restricted characters in address" to the statements in
6381 the default ACL that block characters like @ and % in local parts.
6383 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6384 Three changes have been made:
6386 (a) There was a serious bug; a negative response to MAIL caused the whole
6387 recipient domain to be cached as invalid, thereby blocking all messages
6388 to all local parts at the same domain, from all senders. This bug has
6389 been fixed. The domain is no longer cached after a negative response to
6390 MAIL if the sender used is not empty.
6392 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6395 (c) A new callout option, "use_sender" has been added for people who want
6396 the modified behaviour.
6402 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6405 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6406 indeed breaks things for older releases.
6408 3. Added additional logging to the case where there is a problem reading data
6409 from a filter that is running in a subprocess using a pipe, in order to
6410 try to track down a specific problem.
6412 4. Testing facility fudge: when running in the test harness and attempting
6413 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6414 getting "No route to host". Convert this to a timeout.
6416 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6419 6. Some OS don't have socklen_t but use size_t instead. This affects the
6420 fifth argument of getsockopt() amongst other things. This is now
6421 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6422 can be set for individual OS. I have set it for SunOS5, OSF1, and
6423 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6424 some earlier ones do not.
6426 7. Change 4.30/15 was not doing the test caselessly.
6428 8. The standard form for an IPv6 address literal was being rejected by address
6429 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6430 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6431 this, as well as the form without the "IPv6" on the front (but only when
6432 address literals are enabled, of course).
6434 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6436 10. Exim crashed if a message with an empty sender address specified by -f
6437 encountered a router with an errors_to setting. This could be provoked only
6438 by a command such as
6442 where an empty string was supplied; "<>" did not hit this bug.
6444 11. Installed PCRE release 4.5.
6446 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6447 remained set. It is now erased.
6449 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6450 times from message ids (which are base 36 rather than the normal 62).
6452 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6453 were being counted as actual protocol errors, and logged if the log
6454 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6455 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6456 and DATA following a set of rejected RCPTs do not count as protocol errors.
6457 In other words, Exim assumes they were pipelined, though this may not
6458 actually be the case. Of course, in all cases the client gets an
6459 appropriate error code.
6461 15. If a lookup fails in an ACL condition, a message about the failure may
6462 be available; it is used if testing the ACL cannot continue, because most
6463 such messages specify what the cause of the deferral is. However, some
6464 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6465 that caused an old message to be retained and used if a later statement
6466 caused a defer, replacing the real cause of the deferral.
6468 16. If an IP address had so many PTR records that the DNS lookup buffer
6469 was not large enough to hold them, Exim could crash while trying to process
6470 the truncated data. It now detects and logs this case.
6472 17. Further to 4.21/58, another change has been made: if (and only if) the
6473 first line of a message (the first header line) ends with CRLF, a bare LF
6474 in a subsequent header line has a space inserted after it, so as not to
6475 terminate the header.
6477 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6478 unnecessarily, used atoi() instead of strtol(), and didn't check the
6479 termination when getting file sizes from file names by regex.
6481 19. Completely re-implemented the support for maildirsize files, in the light
6482 of a number of problems with the previous contributed implementation
6483 (4.30/29). In particular:
6485 . If the quota is zero, the maildirsize file is maintained, but no quota is
6488 . If the maildir directory does not exist, it is created before any attempt
6489 to write a maildirsize file.
6491 . The quota value in the file is just a cache; if the quota is changed in
6492 the transport, the new value overrides.
6494 . A regular expression is available for excluding directories from the
6497 20. The autoreply transport checks the characters in options that define the
6498 message's headers; it allows continued headers, but it was checking with
6499 isspace() after an embedded newline instead of explicitly looking for a
6502 21. If all the "regular" hosts to which an address was routed had passed their
6503 expiry times, and had not reached their retry times, the address was
6504 bounced, even if fallback hosts were defined. Now Exim should go on to try
6507 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6508 equivalent code in the SMTP transport. Some hosts send humungous responses
6509 to HELO/EHLO, more than 1024 it seems.
6511 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6512 gives compiler warnings in some environments. I've now done it "properly",
6515 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6516 (because of problems with the built-in one) was declared to return uschar *
6517 instead of char *, causing compiler failure.
6519 25. Fixed a file descriptor leak when processing alias/forward files.
6521 26. Fixed a minor format string issue in dbfn.c.
6523 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6525 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6526 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6529 29. When a sender address is verified, it is cached, to save repeating the test
6530 when there is more than one recipient in a message. However, when the
6531 verification involves a callout, it is possible for different callout
6532 options to be set for different recipients. It is too complicated to keep
6533 track of this in the cache, so now Exim always runs a verification when a
6534 callout is required, relying on the callout cache for the optimization.
6535 The overhead is duplication of the address routing, but this should not be
6538 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6539 to be verified with callout=postmaster, and the main callout worked but the
6540 postmaster check failed, the verification correctly failed. However, if a
6541 subsequent RCPT command asked for sender verification *without* the
6542 postmaster check, incorrect caching caused this verification also to fail,
6545 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6546 it was not caching the DNS options (qualify_single, search_parents) that
6547 were used when the lookup failed. A subsequent lookup with different
6548 options therefore always gave the same answer, though there were cases
6549 where it should not have. (Example: a "domains = !$mx_any" option on a
6550 dnslookup router: the "domains" option is always processed without any
6551 widening, but the router might have qualify_single set.) Now Exim uses the
6552 cached value only when the same options are set.
6554 32. Added John Jetmore's "exipick" utility to the distribution.
6556 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6557 than a timeout (e.g. a certificate is required, and is not provided), an
6558 Exim server now closes the connection immediately. Previously it waited for
6559 the client to close - but if the client is SSL, it seems that they each
6560 wait for each other, leading to a delay before one of them times out.
6562 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6563 maintained 0.8.x compatibility because I don't think many are using it, and
6564 it is clearly obsolete.
6566 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6569 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6570 client certificate was expired. A simple patch fixes this, though I don't
6571 understand the full logic of why the verify callback is called multiple
6574 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6575 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6576 which causes problems with some clients (such as the Certicom SSL Plus
6577 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6578 disables the coutermeasure allowing Eudora to connect."
6580 38. Exim was not checking that a write() to a log file succeeded. This could
6581 lead to Bad Things if a log got too big, in particular if it hit a file
6582 size limit. Exim now panics and dies if it cannot write to a log file, just
6583 as it does if it cannot open a log file.
6585 39. Modified OS/Makefile-Linux so that it now contains
6587 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6589 The two -D definitions ensure that Exim is compiled with large file
6590 support, which makes it possible to handle log files that are bigger than
6593 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6594 instance) a domain was checked against a named list that involved a lookup,
6595 causing $domain_data to be set, then another domain was checked against the
6596 same list, then the first domain was re-checked, the value of $domain_data
6597 after the final check could be wrong. In particular, if the second check
6598 failed, it could be set empty. This bug probably also applied to
6601 41. The strip_trailing_dot option was not being applied to the address given
6602 with the -f command-line option.
6604 42. The code for reading a message's header from the spool was incrementing
6605 $received_count, but never initializing it. This meant that the value was
6606 incorrect (doubled) while delivering a message in the same process in which
6607 it was received. In the most common configuration of Exim, this never
6608 happens - a fresh exec is done - but it can happen when
6609 deliver_drop_privilege is set.
6611 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6612 it now includes up to 150 characters of the unexpected data in the log
6615 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6616 and building data strings. The size of both of these buffers was 10 000
6617 bytes - far larger than anybody would *ever* want, thought I. Needless to
6618 say, somebody hit the limit. I have increased the maximum line length to
6619 20 000 and the maximum data length of concatenated lines to 100 000. I have
6620 also fixed two bugs, because there was no checking on these buffers. Tsk,
6621 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6622 buffer is too small.
6624 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6625 lsearch lookups. Now it does.
6627 46. When parsing a route_list item in a manualroute router, a fixed-length
6628 buffer was used for the list of hosts. I made this 1024 bytes long,
6629 thinking that nobody would ever have a list of hosts that long. Wrong.
6630 Somebody had a whole pile of complicated expansion conditions, and the
6631 string was silently truncated, leading to an expansion error. It turns out
6632 that it is easier to change to an unlimited length (owing to other changes
6633 that have happened since this code was originally written) than to build
6634 structure for giving a limitation error. The length of the item that
6635 expands into the list of hosts is now unlimited.
6637 47. The lsearch lookup could not handle data where the length of text line was
6638 more than 4095 characters. Such lines were truncated, leading to shortened
6639 data being returned. It should now handle lines of any length.
6641 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6642 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6645 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6646 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6647 no longer happens in comment lines. A list of replacements is now placed
6648 at the head of all of the source files, except those whose only change is
6649 to replace PERL_COMMAND in the very first #! line.
6651 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6652 messages on the queue, with a bottom-up merge sort, using code contributed
6653 by Michael Haardt. This should make operations like -bp somewhat faster on
6654 large queues. It won't affect queue runners, except when queue_run_in_order
6657 51. Installed eximstats 1.31 in the distribution.
6659 52. Added support for SRV lookups to the dnslookup router.
6661 53. If an ACL referred to $message_body or $message_body_end, the value was not
6662 reset for any messages that followed in the same SMTP session.
6664 54. The store-handling optimization for building very long strings was not
6665 differentiating between the different store pools. I don't think this
6666 actually made any difference in practice, but I've tidied it.
6668 55. While running the routers to verify a sender address, $sender_address
6669 was still set to the sender address. This is wrong, because when routing to
6670 send a bounce to the sender, it would be empty. Therefore, I have changed
6671 it so that, while verifying a sender address, $sender_address is set to <>.
6672 (There is no change to what happens when verifying a recipient address.)
6674 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6675 target A or AAAA records (if not already returned) without resetting the
6676 qualify_single or search_parents options of the DNS resolver. These are
6677 inappropriate in this case because the targets of MX and SRV records must
6678 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6679 target that, when qualified, matched something in the local domain. These
6680 two options are now turned off when doing these lookups.
6682 57. It seems that at least some releases of Reiserfs (which does not have the
6683 concept of a fixed number of inodes) returns zero and not -1 for the
6684 number of available inodes. This interacted badly with check_spool_inodes,
6685 which assumed that -1 was the "no such thing" setting. What I have done is
6686 to check that the total number of inodes is greater than zero before doing
6687 the test of how many are available.
6689 58. When a "warn" ACL statement has a log_message modifier, the message is
6690 remembered, and not repeated. This is to avoid a lot of repetition when a
6691 message has many recipients that cause the same warning to be written.
6692 However, Exim was preserving the list of already written lines for an
6693 entire SMTP session, which doesn't seem right. The memory is now reset if a
6694 new message is started.
6696 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6697 debugging output unless log_rewrite was also set.
6699 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6700 of (int)(handle) when we know that handle contains (void *)(-1).
6702 61. The Exim daemon panic-logs an error return when it closes the incoming
6703 connection. However "connection reset by peer" seems to be common, and
6704 isn't really an error worthy of noting specially, so that particular error
6707 62. When Exim is trying to find all the local interfaces, it used to panic and
6708 die if the ioctl to get the interface flags failed. However, it seems that
6709 on at least one OS (Solaris 9) it is possible to have an interface that is
6710 included in the list of interfaces, but for which you get a failure error
6711 for this call. This happens when the interface is not "plumbed" into a
6712 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6713 failure of the "get flags" call assumes that the interface is down.
6715 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6716 makes life easier for people who are doing arithmetic on fields extracted
6717 from dates, where you often get leading zeros that should not be
6718 interpreted as octal.
6720 64. Added qualify_domain to the redirect router, to override the global
6723 65. If a pathologically long header line contained very many addresses (the
6724 report of this problem mentioned 10 000) and each of them was rewritten,
6725 Exim could use up a very large amount of memory. (It kept on making new
6726 copies of the header line as it rewrote, and never released the old ones.)
6727 At the expense of a bit more processing, the header rewriting function has
6728 been changed so that it no longer eats memory in this way.
6730 66. The generation of the Received: header has been moved from the time that a
6731 message starts to be received, to the time that it finishes. The timestamp
6732 in the Received: header should now be very close to that of the <= log
6733 line. There are two side-effects of this change:
6735 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6736 logged header lines no longer include the local Received: line, because
6737 it has not yet been created. The same applies to a copy of the message
6738 that is returned to a non-SMTP sender when a message is rejected.
6740 (b) When a filter file is tested using -bf, no additional Received: header
6741 is added to the test message. After some thought, I decided that this
6744 This change does not affect the value of $received_for. It is still set
6745 after address rewriting, but before local_scan() is called.
6747 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6749 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6750 gave an unhelpful panic error message, and a defer error. I have managed to
6751 change this behaviour so that it now rejects any supplied certificate,
6752 which seems right, as the list of acceptable certificates is empty.
6754 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6755 gave an unhelpful defer error. I have not managed to make this reject any
6756 supplied certificates, but the error message it gives is "no certificate
6757 supplied", which is not helpful.
6759 70. exigrep's output now also includes lines that are not associated with any
6760 message, but which match the given pattern. Implemented by a patch from
6761 Martin Sluka, which also tidied up the Perl a bit.
6763 71. Recipient callout verification, like sender verification, was using <> in
6764 the MAIL FROM command. This isn't really the right thing, since the actual
6765 sender may affect whether the remote host accepts the recipient or not. I
6766 have changed it to use the actual sender in the callout; this means that
6767 the cache record is now keyed on a recipient/sender pair, not just the
6768 recipient address. There doesn't seem to be a real danger of callout loops,
6769 since a callout by the remote host to check the sender would use <>.
6770 [SEE ABOVE: changed after hitting problems.]
6772 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6773 temporary errors. However, in the case of such a code being given after
6774 the end of a data transmission (i.e. after ".") Exim was failing to write
6775 a retry record for the message. (Yes, there was some broken host that was
6776 actually sending 8xx at this point.)
6778 73. An unknown lookup type in a host list could cause Exim to panic-die when
6779 the list was checked. (An example that provoked this was putting <; in the
6780 middle of a list instead of at the start.) If this happened during a DATA
6781 ACL check, a -D file could be left lying around. This kind of configuration
6782 error no longer causes Exim to die; instead it causes a defer error. The
6783 incident is still logged to the main and panic logs.
6785 74. Buglet left over from Exim 3 conversion. The message "too many messages
6786 in one connection" was written to the rejectlog but not the mainlog, except
6787 when address rewriting (yes!) was being logged.
6789 75. Added write_rejectlog option.
6791 76. When a system filter was run not as root (that is, when system_filter_user
6792 was set), the values of the $n variables were not being returned to the
6793 main process; thus, they were not subsequently available in the $sn
6796 77. Added +return_path_on_delivery log selector.
6798 78. A connection timeout was being treated differently from recipients deferred
6799 when testing hosts_max_try with a message that was older than the host's
6800 retry timeout. (The host should not be counted, thus allowing all hosts to
6801 be tried at least once before bouncing.) This may have been the cause of an
6802 occasionally reported bug whereby a message would remain on the queue
6803 longer than the retry timeout, but would be bounced if a delivery was
6804 forced. I say "may" because I never totally pinned down the problem;
6805 setting up timeout/retry tests is difficult. See also the next item.
6807 79. The ultimate address timeout was not being applied to errors that involved
6808 a combination of host plus message (for example, a timeout on a MAIL
6809 command). When an address resolved to a number of possible hosts, and they
6810 were not all tried for each delivery (e.g. because of hosts_max_try), a
6811 message could remain on the queue longer than the retry timeout.
6813 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6816 81. Fixed an obscure SMTP outgoing bug which required at least the following
6817 conditions: (a) there was another message waiting for the same server;
6818 (b) the server returned 5xx to all RCPT commands in the first message so
6819 that the message was not completed; (c) the server dropped the connection
6820 or gave a negative response to the RSET that Exim sends to abort the
6821 transaction. The observed case was a dropped connection after DATA that had
6822 been sent in pipelining mode. That is, the server had advertised PIPELINING
6823 but was not implementing it correctly. The effect of the bug was incorrect
6824 behaviour, such as trying another host, and this could lead to a crash.
6830 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6831 and daemon.c were passed as pointers to ints; they should have been
6832 pointers to socklen_t variables (which are typically unsigned ints).
6834 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6837 3. Fixed a really odd bug that affected only the testing scheme; patching a
6838 certain fixed string in the binary changed the value of another string that
6839 happened to be identical to the end of the original first string.
6841 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6842 name", it returns that address as the IP address. On some operating
6843 systems (e.g. Solaris), it also passes back the IP address string as the
6844 "host name". However, on others (e.g. Linux), it passes back an empty
6845 string. Exim wasn't checking for this, and was changing the host name to an
6846 empty string, assuming it had been canonicalized.
6848 5. Although rare, it is permitted to have more than one PTR record for a given
6849 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6850 all the names associated with an address, because they do in Solaris.
6851 However, it seems that they do not in Linux for data that comes from the
6852 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6853 I found this out when I moved to a new Linux workstation and tried to run
6854 the Exim test suite.
6856 To get round this problem I have changed the code so that it now does its
6857 own call to the DNS to look up PTR records when searching for a host name.
6858 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6859 addresses that are only in /etc/hosts are still found.
6861 This behaviour is, however, controlled by an option called host_lookup_
6862 order, which defaults to "bydns:byaddr". If people want to use the other
6863 order, or indeed, just use one or the other means of lookup, they can
6864 specify it in this variable.
6866 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6867 some operating systems, this comes back from gethostbyaddr() as an empty
6868 string, and this is what Exim used to test for. However, it seems that in
6869 other systems, "." is yielded. Exim now tests for this case too.
6871 7. The values of check_spool_space and check_log_space are now held internally
6872 as a number of kilobytes instead of an absolute number of bytes. If a
6873 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6874 kilobyte. This means that much larger values can be stored.
6876 8. Exim monitor: an attempt to get the action menu when not actually pointing
6877 at a message produces an empty menu entitled "No message selected". This
6878 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6879 no entries in it ("Shell widget menu has zero width and/or height"). So I
6880 have added a single, blank menu entry in this case.
6882 9. Added ${quote_local_part.
6884 10. MIME decoding is now applied to the contents of Subject: header lines when
6887 11. Now that a reference to $sender_host_address automatically causes a reverse
6888 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6889 host lookup before query-style lookups in lists that might use this
6890 variable. This has therefore been abolished, and the "net-" prefix is no
6891 longer necessary for query-style lookups.
6893 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6894 have been a typo for LFLAGS, so it has been changed.
6896 13. The install script calls Exim with "-C /dev/null" in order to find the
6897 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6898 to be output. However, since Exim outputs its version number before the
6899 error, it didn't break the script. It just looked ugly. I fixed this by
6900 always allowing "-C /dev/null" if the caller is root.
6902 14. Ignore overlarge ACL variable number when reading spool file - insurance
6903 against a later release with more variables having written the file.
6905 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6906 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6907 this, as well as the form without the "IPv6" on the front.
6909 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6910 OS/Makefile-Darwin file.
6912 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6913 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6915 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6916 variable. However, some OS use ints for inodes. I've added cast to long int
6917 to get rid of the compiler warning.
6919 19. I had forgotten to lock out "/../" in configuration file names when
6920 ALT_CONFIG_PREFIX was set.
6922 20. Routers used for verification do not need to specify transports. However,
6923 if such a router generated a host list, and callout was configured, Exim
6924 crashed, because it could not find a port number from the (non-existent)
6925 transport. It now assumes port 25 in this circumstance.
6927 21. Added the -t option to exigrep.
6929 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6930 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6931 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6932 them is set, LOOKUP_LSEARCH is forced.
6934 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6936 24. Added sender and host information to the "rejected by local_scan()" log
6937 line; previously there was no indication of these.
6939 25. Added .include_if_exists.
6941 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6942 Linux. It turns out that not all file systems support this. Apparently some
6943 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6944 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6945 not supported on the file descriptor, is now ignored when Exim is trying to
6946 sync a directory. This applies only to Linux.
6948 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6950 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6951 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6954 29. Added support for maildirsize files from supplied patch (modified a bit).
6956 30. The use of :fail: followed by an empty string could lead Exim to respond to
6957 sender verification failures with (e.g.):
6959 550 Verification failed for <xxx>
6960 550 Sender verify failed
6962 where the first response line was missing the '-' that indicates it is not
6963 the final line of the response.
6965 31. The loop for finding the name of the user that called Exim had a hardwired
6966 limit of 10; it now uses the value of finduser_retries, which is used for
6967 all other user lookups.
6969 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6972 33. Exim was neglecting to zero errno before one call of strtol() when
6973 expanding a string and expecting an integer value. On some systems this
6974 resulted in spurious "integer overflow" errors. Also, it was casting the
6975 result into an int without checking.
6977 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6978 did not work. The code looks as if it has *never* worked, though it appears
6979 to have been documented since at least release 1.62. I have made it work.
6981 35. The "timeout_DNS" error in retry rules, also documented since at least
6982 1.62, also never worked. As it isn't clear exactly what this means, and
6983 clearly it isn't a major issue, I have abolished the feature by treating it
6984 as "timeout", and writing a warning to the main and panic logs.
6986 36. The display of retry rules for -brt wasn't always showing the error code
6989 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6990 timeout_connect_A, timeout_connect_MX.
6992 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6993 to the empty sender.
6995 39. The daemon was not analysing the content of -oX till after it had closed
6996 stderr and disconnected from the controlling terminal. This meant that any
6997 syntax errors were only noted on the panic log, and the return code from
6998 the command was 0. By re-arranging the code a little, I've made the
6999 decoding happen first, so such errors now appear on stderr, and the return
7000 code is 1. However, the actual setting up of the sockets still happens in
7001 the disconnected process, so errors there are still only recorded on the
7004 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7005 connections (as happens on some IP stacks) was logged at start up time as
7006 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7007 it from "IPv6 and IPv4", which means that two separate sockets are being
7010 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7011 says whether AF_INET or AF_INET6 was passed as an argument.
7013 42. Exiwhat output was messed up when time zones were included in log
7016 43. Exiwhat now gives more information about the daemon's listening ports,
7017 and whether -tls-on-connect was used.
7019 44. The "port" option of the smtp transport is now expanded.
7021 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7022 silently ignored. Now an error message is written to the main and panic
7025 46. There's a new ACL modifier called "logwrite" which writes to a log file
7026 as soon as it is encountered.
7028 47. Added $local_user_uid and $local_user_gid at routing time.
7030 48. Exim crashed when trying to verify a sender address that was being
7033 49. Exim was recognizing only a space character after ".include". It now also
7034 recognizes a tab character.
7036 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7037 extracting the relevant information from the specification. The man page no
7038 longer contains scrambled data for the -d option, and I've added a section
7039 at the front about calling Exim under different names.
7041 51. Added "extra_headers" argument to the "mail" command in filter files.
7043 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7046 53. Installed eximstats 1.29.
7048 54. Added transport_filter_timeout as a generic transport option.
7050 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7051 Cc: header lines. This was required by RFC 822, but it not required by RFC
7054 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7055 incoming messages that did not have them. Now it does so only if the
7056 message originates locally, that is, if there is no associated remote host
7057 address. When Resent- header lines are present, this applies to the Resent-
7058 lines rather than the non-Resent- lines.
7060 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7061 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7063 58. Messages for configuration errors now include the name of the main
7064 configuration file - useful now that there may be more than one file in a
7065 list (.included file names were always shown).
7067 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7068 for those rare installations that do not start the daemon as root or run it
7069 setuid root. I've cut out the call to initgroups() if the daemon is not
7072 60. The Exim user and group can now be bound into the binary as text strings
7073 that are looked up at the start of Exim's processing.
7075 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7077 62. Added $mailstore_basename variable.
7079 63. Installed patch to sieve.c from Michael Haardt.
7081 64. When Exim failed to open the panic log after failing to open the main log,
7082 the original message it was trying to log was written to stderr and debug
7083 output, but if they were not available (the usual case in production), it
7084 was lost. Now it is written to syslog before the two lines that record the
7085 failures to open the logs.
7087 65. Users' Exim filters run in subprocesses under the user's uid. It is
7088 possible for a "deliver" command or an alias in a "personal" command to
7089 provoke an address rewrite. If logging of address rewriting is configured,
7090 this fails because the process is not running as root or exim. There may be
7091 a better way of dealing with this, but for the moment (because 4.30 needs
7092 to be released), I have disabled address rewrite logging when running a
7093 filter in a non-root, non-exim process.
7099 1. The buildconfig auxiliary program wasn't quoting the value set for
7100 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7101 not defined. This bug was masked in 4.22 by the effect that was fixed in
7104 2. Some messages that were rejected after a message id was allocated were
7105 shown as "incomplete" by exigrep. It no longer does this for messages that
7106 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7108 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7109 have allow_domain_literals set, the ID did not get logged in the <= line.
7110 Domain literals are now always recognized in Message-ID: header lines.
7112 4. The first argument for a ${extract expansion item is the key name or field
7113 number. Leading and trailing spaces in this item were not being ignored,
7114 causing some misleading effects.
7116 5. When deliver_drop_privilege was set, single queue runner processes started
7117 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7118 same command in the process it spins off) were not dropping privilege.
7120 6. When the daemon running as "exim" started a queue runner, it always
7121 re-executed Exim in the spun-off process. This is a waste of effort when
7122 deliver_drop_privilege is set. The new process now just calls the
7123 queue-runner function directly.
7129 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7132 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7133 ignored. Though the use of -oP was forcing the writing of a pid file, it
7134 was always written to the default place.
7136 3. If the message "no IP address found for host xxxx" is generated during
7137 incoming verification, it is now followed by identification of the incoming
7138 connection (so you can more easily find what provoked it).
7140 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7142 5. Added some features to "harden" Exim a bit more against certain attacks:
7144 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7145 be put in Local/Makefile. This is like the never_users runtime option,
7146 but it cannot be overridden. The default setting is "root".
7148 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7149 prefix string with which any file named in a -C command line option
7152 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7153 is retained for -C and -D only if the caller of Exim is root. Without
7154 it, the exim user may also use -C and -D and retain privilege.
7156 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7157 command line option is disabled.
7159 6. Macro names set by the -D option must start with an upper case letter, just
7160 like macro names defined in the configuration file.
7162 7. Added "dereference=" facility to LDAP.
7164 8. Two instances of the typo "uknown" in the source files are fixed.
7166 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7167 the Configure-Makefile script screwed up while processing it.
7169 10. Incorporated PCRE 4.4.
7171 11. The SMTP synchronization check was not operating right at the start of an
7172 SMTP session. For example, it could not catch a HELO sent before the client
7173 waited for the greeting. There is now a check for outstanding input at the
7174 point when the greeting is written. Because of the duplex, asynchronous
7175 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7176 way, but not yet received, when the check is performed.
7178 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7179 on TCP/IP sockets, because this apparently causes some broken clients to
7182 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7183 unchanged) from the Cygwin maintainer.
7185 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7186 was supported instead of testing for mbx. Effectively a typo.
7188 15. The spa authenticator server code was not checking that the input it
7189 received was valid base64.
7191 16. The debug output line for the "set" modifier in ACLs was not showing the
7192 name of the variable that was being set.
7194 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7196 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7197 lookup on its own. Something else was needed to trigger the lookup. For
7198 example, a match in host_lookup or the need for a host name in a host list.
7199 Now, if $sender_host_name is referenced and the host name has not yet been
7200 looked up, a lookup is performed. If the lookup fails, the variable remains
7201 empty, and $host_lookup_failed is set to "1".
7203 19. Added "eqi" as a case-independent comparison operator.
7205 20. The saslauthd authentication condition could segfault if neither service
7206 nor realm was specified.
7208 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7209 error message that was logged was misleading, and incoming SMTP
7210 connections were dropped. The message is now more accurate, and temporary
7211 errors are given to SMTP connections.
7213 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7214 (except RSET) with a 421 error, until QUIT is received. However, it was
7215 failing to send a response to QUIT.
7217 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7218 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7219 ACL is now run after the helo_try_verify_hosts code.
7221 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7222 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7223 case-independent, so other case variants are also recognized). Apparently
7224 some systems use these upper case variants.
7226 25. If more than two messages were waiting for the same host, and a transport
7227 filter was specified for the transport, Exim sent two messages over the
7228 same TCP/IP connection, and then failed with "socket operation on non-
7229 socket" when it tried to send the third.
7231 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7233 27. The extern definition of crypt16() in expand.c was not being excluded when
7234 the OS had its own crypt16() function.
7236 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7237 as a preferred synonym for return_size_limit, both as an option and as an
7240 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7242 30. Changed the default configuration ACL to relax the local part checking rule
7243 for addresses that are not in any local domains. For these addresses,
7244 slashes and pipe symbols are allowed within local parts, but the sequence
7245 /../ is explicitly forbidden.
7247 31. SPA server authentication was not clearing the challenge buffer before
7250 32. log_message in a "warn" ACL statement was writing to the reject log as
7251 well as to the main log, which contradicts the documentation and doesn't
7252 seem right (because no rejection is happening). So I have stopped it.
7254 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7255 However, I am unable to do any testing of this.
7257 34. Fixed an infelicity in the appendfile transport. When checking directories
7258 for a mailbox, to see if any needed to be created, it was accidentally
7259 using path names with one or more superfluous leading slashes; tracing
7260 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7262 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7263 contents are added to the log line that is written for every discarded
7264 recipient. (Previously a log_message setting was ignored.)
7266 36. The ${quote: operator now quotes the string if it is empty.
7268 37. The install script runs exim in order to find its version number. If for
7269 some reason other than non-existence or emptiness, which it checks, it
7270 could not run './exim', it was installing it with an empty version number,
7271 i.e. as "exim-". This error state is now caught, and the installation is
7274 38. An argument was missing from the function that creates an error message
7275 when Exim fails to connect to the socket for saslauthd authentication.
7276 This could cause Exim to crash, or give a corrupted message.
7278 39. Added isip, isip4, and isip6 to ${if conditions.
7280 40. The ACL variables $acl_xx are now saved with the message, and can be
7281 accessed later in routers, transports, and filters.
7283 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7284 strings in the file are not string-expanded.
7286 42. If a MAIL command specified a SIZE value that was too large to fit into an
7287 int variable, the check against message_size_limit failed. Such values are
7288 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7289 day this will have to be increased, but I don't think I want to be around
7290 when emails are that large.
7297 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7298 iconv() is not standard in FreeBSD.
7300 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7301 IPv6 enabled. The observed symptom was a segmentation fault on return from
7302 the function os_common_find_running_interfaces() in src/os.c.
7304 3. In the check_special_case() function in daemon.c I had used "errno" as an
7305 argument name, which causes warnings on some systems. This was basically a
7306 typo, since it was named "eno" in the comments!
7308 4. The code that waits for the clock to tick (at a resolution of some fraction
7309 of a second) so as to ensure message-id uniqueness was always waiting for
7310 at least one whole tick, when it could have waited for less. [This is
7311 almost certainly not relevant at current processor speeds, where it is
7312 unlikely to ever wait at all. But we try to future-proof.]
7314 5. The function that sleeps for a time interval that includes fractions of a
7315 second contained a race. It did not block SIGALRM between setting the
7316 timer, and suspending (a couple of lines later). If the interval was short
7317 and the sigsuspend() was delayed until after it had expired, the suspension
7318 never ended. On busy systems this could lead to processes getting stuck for
7321 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7322 process, before it forks any delivery processes. The open lookup caching
7323 mechanism meant that the open file or database connection was passed into
7324 the delivery process. The problem was that delivery processes always tidy
7325 up cached lookup data. This could cause a problem for the next delivery
7326 process started by the queue runner, because the external queue runner
7327 process does not know about the closure. So the next delivery process
7328 still has data in the lookup cache. In the case of a file lookup, there was
7329 no problem because closing a file descriptor in a subprocess doesn't affect
7330 the parent. However, if the lookup was caching a connection to a database,
7331 the connection was closed, and the second delivery process was likely to
7332 see errors such as "PGSQL: query failed: server closed the connection
7333 unexpectedly". The problem has been fixed by closing all cached lookups
7334 in a queue runner before running a delivery process.
7336 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7337 seem to have the "const" qualifier which it has on other OS. I've
7340 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7341 *of the same type* (client or server) with the same public name that an
7342 error should be diagnosed.
7344 9. When Exim looked up a host name for an IP address, but failed to find the
7345 original IP address when looking up the host name (a safety check), it
7346 output the message "<ip address> does not match any IP for NULL", which was
7347 confusing, to say the least. The bug was that the host name should have
7348 appeared instead of "NULL".
7350 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7351 user that is built into the binary, and the -C or -D options is used, root
7352 privilege is dropped before the configuration file is read. In addition,
7353 logging is switched to stderr instead of the normal log files. If the
7354 configuration then re-defines the Exim user, the unprivileged environment
7355 is probably not what is expected, so Exim logs a panic warning message (but
7358 However, if deliver_drop_privilege is set, the unprivileged state may well
7359 be exactly what is intended, so the warning has been cut out in that case,
7360 and Exim is allowed to try to write to its normal log files.
7366 1. smtp_return_error_details was not giving details for temporary sender
7367 or receiver verification errors.
7369 2. Diagnose a configuration error if two authenticators have the same public
7372 3. Exim used not to create the message log file for a message until the first
7373 delivery attempt. This could be confusing when incoming messages were held
7374 for policy or load reasons. The message log file is now created at the time
7375 the message is received, and an initial "Received" line is written to it.
7377 4. The automatically generated man page for command line options had a minor
7378 bug that caused no ill effects; however, a more serious problem was that
7379 the procedure for building the man page automatically didn't always
7380 operate. Consequently, release 4.20 contains an out-of-date version. This
7381 shouldn't happen again.
7383 5. When building Exim with embedded Perl support, the script that builds the
7384 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7385 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7387 6. The freeze_tell option was not being used for messages that were frozen on
7388 arrival, either by an ACL or by local_scan().
7390 7. Added the smtp_incomplete_transaction log selector.
7392 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7393 was accepting AUTH without a new EHLO.
7395 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7396 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7399 10. Logging of TCP/IP connections (when configured) now happens in the main
7400 daemon process instead of the child process, so that the TCP/IP connection
7401 count is more accurate (but it can never be perfect).
7403 11. The use of "drop" in a nested ACL was not being handled correctly in the
7404 outer ACL. Now, if condition failure induced by the nested "drop" causes
7405 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7406 or "require"), the connection is dropped.
7408 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7409 that yield "discard" can now be used with an "accept" or a "discard" verb,
7410 but an error is generated for any others (because I can't see a useful way
7411 to define what should happen).
7413 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7414 are now processed in the same way as lines in the Exim configuration file.
7415 In particular, continuation lines are supported.
7417 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7419 15. Added -ti meaning -t -i.
7421 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7422 domains, and warn by logging if others are found.
7424 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7425 returned from GIFCONF when Exim is trying to find the list of interfaces on
7426 a host. The code in os.c has been modified to copy each ifreq to an aligned
7427 structure in all cases.
7429 Also, in some cases, the returned ifreq's were being copied to a 'struct
7430 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7431 means the last couple of bytes of an IPv6 address could be chopped if the
7432 ifreq contained only a normal sockaddr (14 bytes storage).
7434 18. Named domain lists were not supported in the hosts_treat_as_local option.
7435 An entry such as +xxxx was not recognized, and was treated as a literal
7438 19. Ensure that header lines added by a DATA ACL are included in the reject log
7439 if the ACL subsequently rejects the message.
7441 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7442 MD5 (which is deprecated).
7444 21. When testing a filter file using -bf, Exim was writing a message when it
7445 took the sender from a "From " line in the message, but it was not doing so
7446 when it took $return_path from a Return-Path: header line. It now does.
7448 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7449 with a valid header line field name (a series of printing characters
7450 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7452 23. Changed "disc" in the source to "disk" to conform to the documentation and
7453 the book and for uniformity.
7455 24. Ignore Sendmail's -Ooption=value command line item.
7457 25. When execve() failed while trying to run a command in a pipe transport,
7458 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7459 could be confused with a return value of 69 from the command itself. This
7460 has been changed to 127, the value the shell returns if it is asked to run
7461 a non-existent command. The wording for the related log line suggests a
7462 non-existent command as the problem.
7464 26. If received_header_text expands to an empty string, do not add a Received:
7465 header line to the message. (Well, it adds a token one on the spool, but
7466 marks it "old" so that it doesn't get used or transmitted.)
7468 27. Installed eximstats 1.28 (addition of -nt option).
7470 28. There was no check for failure on the call to getsockname() in the daemon
7471 code. This can fail if there is a shortage of resources on the system, with
7472 ENOMEM, for example. A temporary error is now given on failure.
7474 29. Contrary to the C standard, it seems that in some environments, the
7475 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7476 program. Exim now does this explicitly; it affects the formatting of
7477 timestamps using strftime().
7479 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7480 complaints. I've now initialized all the variables, to avoid this.
7482 32. Header lines added by a system filter were not being "seen" during
7483 transport-time rewrites.
7485 33. The info_callback() function passed to OpenSSL is set up with type void
7486 (*)(SSL *, int, int), as described somewhere. However, when calling the
7487 function (actually a macro) that sets it up, the type void(*)() is
7488 expected. I've put in a cast to prevent warnings from picky compilers.
7490 34. If a DNS black list lookup found a CNAME record, but there were no A
7491 records associated with the domain it pointed at, Exim crashed.
7493 35. If a DNS black list lookup returned more than one A record, Exim ignored
7494 all but the first. It now scans all returned addresses if a particular IP
7495 value is being sought. In this situation, the contents of the
7496 $dnslist_value variable are a list of all the addresses, separated by a
7499 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7500 to accept a host name and all its aliases if the forward lookup for any of
7501 them yielded the IP address of the incoming connection. Now it accepts only
7502 those names whose forward lookup yields the correct IP address. Any other
7503 names are discarded. This closes a loophole whereby a rogue DNS
7504 administrator could create reverse DNS records to break through a
7505 wildcarded host restriction in an ACL.
7507 37. If a user filter or a system filter that ran in a subprocess used any of
7508 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7509 the wrong values were passed to the pipe command ($thisaddress had the
7510 value of $0, $0 had the value of $1, etc). This bug was introduced by
7511 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7513 38. Improved the line breaking for long SMTP error messages from ACLs.
7514 Previously, if there was no break point between 40 and 75 characters, Exim
7515 left the rest of the message alone. Two changes have been made: (a) I've
7516 reduced the minimum length to 35 characters; (b) if it can't find a break
7517 point between 35 and 75 characters, it looks ahead and uses the first one
7518 that it finds. This may give the occasional overlong line, but at least the
7519 remaining text gets split now.
7521 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7522 file descriptors might be low, and that setting 1000 would always raise it.
7523 It turns out that in some environments, the limit is already over 1000 and
7524 that lowering it causes trouble. So now Exim takes care not to decrease it.
7526 40. When delivering a message, the value of $return_path is set to $sender_
7527 address at the start of routing (routers may change the value). By an
7528 oversight, this default was not being set up when an address was tested by
7529 -bt or -bv, which affected the outcome if any router or filter referred to
7532 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7533 log only when "message" or "log_message" are set. However, if one of the
7534 conditions was an address verification, or a call to a nested ACL, the
7535 messages generated by the underlying test were being passed through. This
7536 no longer happens. The underlying message is available in $acl_verify_
7537 message for both "message" and "log_message" expansions, so it can be
7538 passed through if needed.
7540 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7541 new expansion $bh_ to give the encoded byte string without charset
7542 translation. Translation happens only if iconv() is available; HAVE_ICONV
7543 indicates this at build time. HEADERS_CHARSET gives the charset to
7544 translate to; headers_charset can change it in the configuration, and
7545 "headers charset" can change it in an individual filter file.
7547 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7548 that creates RFC 2047 encoded "words" labels them as that charset instead
7549 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7550 expansion operator; (ii) when Exim creates a From: line for a local
7551 message; (iii) when a header line is rewritten to include a "phrase" part.
7553 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7554 buggy, causing it to skip the first lines of messages whose message ID
7555 ended in 'D'. This would not have bitten before Exim release 4.14, because
7556 message IDs were unlikely to end in 'D' before then. The effect was to have
7557 incorrect size information for certain domains.
7559 45. #include "config.h" was missing at the start of the crypt16.c module. This
7560 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7563 46. If there was a timeout during a "random" callout check, Exim treated it as
7564 a failure of the random address, and carried on sending RSET and the real
7565 address. If the delay was just some slowness somewhere, the response to the
7566 original RCPT would be taken as a response to RSET and so on, causing
7567 mayhem of various kinds.
7569 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7570 when I implemented it. It didn't allow for the fact that some option values
7571 may legitimately be negative (e.g. size_addition), and it didn't even do
7572 the right test for positive values.
7574 48. Domain names in DNS records are case-independent. Exim always looks them up
7575 in lower case. Some resolvers return domain names in exactly the case they
7576 appear in the zone file, that is, they may contain uppercase letters. Not
7577 all resolvers do this - some return always lower case. Exim was treating a
7578 change of case by a resolver as a change of domain, similar to a widening
7579 of a domain abbreviation. This triggered its re-routing code and so it was
7580 trying to route what was effectively the same domain again. This normally
7581 caused routing to fail (because the router wouldn't handle the domain
7582 twice). Now Exim checks for this case specially, and just changes the
7583 casing of the domain that it ultimately uses when it transmits the message
7586 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7589 50. If a filter generated a file delivery with a non-absolute name (possible if
7590 no home directory exists for the router), the forbid_file option was not
7593 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7594 the existing equality matching.
7596 52. Exim was using ints instead of ino_t variables in some places where it was
7597 dealing with inode numbers.
7599 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7600 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7601 called TMPDIR, and if it finds it is different, it changes its value.
7603 54. The smtp_printf() function is now made available to local_scan() so
7604 additional output lines can be written before returning. There is also an
7605 smtp_fflush() function to enable the detection of a dropped connection.
7606 The variables smtp_input and smtp_batched_input are exported to
7609 55. Changed the default runtime configuration: the message "Unknown user"
7610 has been removed from the ACL, and instead placed on the localuser router,
7611 using the cannot_route_message feature. This means that any verification
7612 failures that generate their own messages won't get overridden. Similarly,
7613 the "Unrouteable address" message that was in the ACL for unverifiable
7614 relay addresses has also been removed.
7616 56. Added hosts_avoid_esmtp to the smtp transport.
7618 57. The exicyclog script was not checking for the esoteric option
7619 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7620 will work only if exicyclog is run under the appropriate euid.
7622 58. Following a discussion on the list, the rules by which Exim recognises line
7623 endings on incoming messages have been changed. The -dropcr and drop_cr
7624 options are now no-ops, retained only for backwards compatibility. The
7625 following line terminators are recognized: LF CRLF CR. However, special
7626 processing applies to CR:
7628 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7629 nor a local message in the state where . is a terminator.
7631 (ii) If a bare CR is encountered in a header line, an extra space is added
7632 after the line terminator so as not to end the header. The reasoning
7633 behind this is that bare CRs in header lines are most likely either
7634 to be mistakes, or people trying to play silly games.
7636 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7637 was being incorrectly given as 18 bytes larger than it should have been.
7638 This is a VOB (very old bug).
7640 60. This may never have affected anything current, but just in case it has:
7641 When the local host is found other than at the start of a list of hosts,
7642 the local host, those with the same MX, and any that follow, are discarded.
7643 When the list in question was part of a longer list of hosts, the following
7644 hosts (not currently being processed) were also being discarded. This no
7645 longer happens. I'm not sure if this situation could ever has previously
7648 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7651 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7652 header. This is something that is recommended in a new Internet Draft, and
7653 is something that is documented as being done by Sendmail. There are two
7654 possible values. For messages generated by the autoreply transport, Exim
7657 Auto-Submitted: auto-replied
7659 whereas for all other generated messages (e.g. bounces) it adds
7661 Auto-Submitted: auto-generated
7663 63. The "personal" condition in filters now includes a test for the
7664 Auto-Submitted: header. If it contains the string "auto-" the message it
7665 not considered personal.
7667 64. Added rcpt_include_affixes as a generic transport option.
7669 65. Added queue_only_override (default true).
7671 66. Added the syslog_duplication option.
7673 67. If what should have been the first header line of a message consisted of
7674 a space followed by a colon, Exim was mis-interpreting it as a header line.
7675 It isn't of course - it is syntactically invalid and should therefore be
7676 treated as the start of the message body. The misbehaviour could have
7677 caused a number of strange effects, including loss of data in subsequent
7678 header lines, and spool format errors.
7680 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7681 client host had authenticated. This control can now be exercised by an ACL
7682 for more flexibility.
7684 69. By default, callouts do not happen when testing with -bh. There is now a
7685 variant, -bhc, which does actually run the callout code, including
7686 consulting and updating the callout cache.
7688 70. Added support for saslauthd authentication, courtesy of Alexander
7691 71. If statvfs() failed on the spool or log directories while checking their
7692 size for availability, Exim confusingly gave the error "space shortage".
7693 Furthermore, in debugging mode it crashed with a floating point exception.
7694 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7695 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7696 serious problem, Exim now writes to the main and panic logs when this
7697 happens, with details of the failure. It then refuses to accept the
7698 incoming message, giving the message "spool directory problem" or "log
7699 directory problem" with a 421 code for SMTP messages.
7701 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7702 0, 1, and 2 exist, because some OS complain for execs without them (see
7703 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7704 descriptors. However, the code omitted to check that the open succeeded,
7705 causing mysterious errors if for some reason the permissions on /dev/null
7706 got screwed. Now Exim writes a message to the main and panic logs, and
7707 bombs out if it can't open /dev/null.
7709 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7710 interact so that it is all more flexible. It is supposed to remain
7711 backwards compatible. Also added extra_local_interfaces.
7713 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7714 to bomb out with an assertion failure - to the client this appears as a
7715 connection drop. This problem occurs in the part of the code that was taken
7716 from the Samba project. Fortunately, the assertion is in a very simple
7717 function, so I have fixed this by reproducing the function inline in the
7718 one place where it is called, and arranging for authentication to fail
7719 instead of killing the process with assert().
7721 75. The SPA client code was not working when the server requested OEM rather
7722 than Unicode encoding.
7724 76. Added code to make require_files with a specific uid setting more usable in
7725 the case where statting the file as root fails - usually a non-root-mounted
7726 NFS file system. When this happens and the failure is EACCES, Exim now
7727 forks a subprocess and does the per-uid checking as the relevant uid.
7729 77. Added process_log_path.
7731 78. If log_file_path was not explicitly set, a setting of check_log_space or
7732 check_log_inodes was ignored.
7734 79. If a space check for the spool or log partitions fails, the incident is now
7735 logged. Of course, in the latter case the data may get lost...
7737 80. Added the %p formatting code to string_format() so that it can be used to
7738 print addresses in debug_print(). Adjusted all the address printing in the
7739 debugging in store.c to use %p rather than %d.
7741 81. There was a concern that a line of code in smtp_in.c could overflow a
7742 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7743 initially expressed, the concern was not well-founded, because trailing
7744 spaces are removed early. However, if the trailing spaces were followed by
7745 a NULL, they did not get removed, so the overflow was possible. Two fixes
7748 (a) I re-wrote the offending code in a cleaner fashion.
7749 (b) If an incoming SMTP command contains a NULL character, it is rejected
7752 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7753 runs initgroups(), so that if the Exim user is in any additional groups,
7754 they will be used during message reception.
7760 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob