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.
182 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
183 and an ACL use a lookup for checking either the local_part or domain,
184 stale data could be accessed. Ensure that variable references are
185 dropped between messages.
187 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
188 by the client was not checked as pointing within response data before
189 being used. A malicious client could thus cause an out-of-bounds read and
190 possibly gain authentication. Fix by adding the check.
196 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
197 side, to decode a TLS 1.3 packet capture.
199 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
200 Previously the default library behaviour applied, sending two, each in
203 JH/03 Debug output for ACL now gives the config file name and line number for
206 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
208 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
210 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
211 buffer overrun for (non-chunking) other transports.
213 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
214 TLS1.3, means that a server rejecting a client certificate is not visible
215 to the client until the first read of encrypted data (typically the
216 response to EHLO). Add detection for that case and treat it as a failed
217 TLS connection attempt, so that the normal retry-in-clear can work (if
218 suitably configured).
220 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
221 and/or domain. Found and fixed by Jason Betts.
223 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
224 configuration). If a CNAME target was not a wellformed name pattern, a
227 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
228 the OS reports them interleaved with other addresses.
230 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
231 used both for input and for a verify callout, both encrypted, SMTP
232 responses being sent by the server could be lost. This resulted in
233 dropped connections and sometimes bounces generated by a peer sending
236 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
237 string. Previously it was possible to cause undefined behaviour in a
238 library routine (usually a crash). Found by "zerons".
240 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
243 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
244 API was removed, so update to use the newer ones.
246 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
247 any timeout set, is taking a long time. Previously we would hang on to a
248 rotated logfile "forever" if the input was arriving with long gaps
249 (a previous attempt to fix addressed lack, for a long time, of initial
252 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
253 shared (NFS) environment. The length of the tempfile name is now
254 4 + 16 ("hdr.$message_exim_id") which might break on file
255 systems which restrict the file name length to lower values.
258 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
259 shared (NFS) environment.
261 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
262 did for all versions <4.90). Notably -M, -m, --invert, -I may be
265 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
266 on some platforms for bit 31.
268 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
269 to changes apparently associated with TLS1.3 handling some of the APIs
270 previously used were either nonfunctional or inappropriate. Strings
271 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
272 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
273 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
274 This affects log line X= elements, the $tls_{in,out}_cipher variables,
275 and the use of specific cipher names in the encrypted= ACL condition.
277 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
279 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
280 verification result was not updated unless hosts_require_ocsp applied.
282 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
283 queue_list_requires_admin set to false, non-admin users were denied the
286 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
287 directory-of-certs mode. Previously they were advertised despite the
290 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
291 A single TCP connection by a client will now hold a TLS connection open
292 for multiple message deliveries, by default. Previoud the default was to
295 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
296 default. If built with the facility, DANE will be used. The facility
297 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
299 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
300 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
301 must be defined and you must still, unless you define DISABLE_TLS, manage
302 the the include-dir and library-file requirements that go with that
303 choice. Non-TLS builds are still supported.
305 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
308 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
309 default. If the platform supports and has the facility enabled, it will
310 be requested on all coneections.
312 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
313 controlled by the build-time option SUPPORT_PIPE_CONNECT.
315 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
317 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
318 success-DSN messages. Previously the From: header was always the default
319 one for these; the option was ignored.
321 JH/28 Fix the timeout on smtp response to apply to the whole response.
322 Previously it was reset for every read, so a teergrubing peer sending
323 single bytes within the time limit could extend the connection for a
324 long time. Credit to Qualsys Security Advisory Team for the discovery.
326 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
327 delivery address, which leaked information of the results of local
328 forwarding. Change to the original envelope recipient address, per
331 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
332 requested. Previously not bounce was generated and a log entry of
333 error ignored was made.
335 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
337 JH/32 Introduce a general tainting mechanism for values read from the input
338 channel, and values derived from them. Refuse to expand any tainted
339 values, to catch one form of exploit.
341 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
342 was unused and the unexpanded text used for the test. Found and
343 fixed by Ruben Jenster.
345 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
346 an attempt to use a TLS library read routine dereffed a nul pointer,
349 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
350 them in our smtp response.
352 JH/36 Have the general router option retry_use_local_part default to true when
353 any of the restrictive preconditions are set (to anything). Previously it
354 was only for check_local user. The change removes one item of manual
355 configuration which is required for proper retries when a remote router
356 handles a subset of addresses for a domain.
358 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
359 link count into consideration.
361 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
362 caused the extension of big_buffer, the following lines were ignored.
364 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
365 accordance with RFC 2308. Previously there was no expiry, so a longlived
366 receive process (eg. due to ACL delays) versus a short SOA value could
369 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
371 JH/39 Promote DMARC support to mainline.
373 JH/40 Bug 2452: Add a References: header to DSNs.
375 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
376 parameters. The relevant library call is documented as "Deprecated: This
377 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
378 3.6.0, DH parameters are negotiated following RFC7919."
380 HS/06 Change the default of dnssec_request_domains to "*"
382 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
383 carried on and emitted a BDAT command, even when PIPELINING was not
386 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
387 buffer was used for the filename, resulting in a trap when tainted
388 arguments (eg. $domain) were used.
390 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
391 recommended to avoid a possible server-load attack. The feature can be
392 re-enabled via the openssl_options main cofiguration option.
394 JH/45 local_scan API: documented the current smtp_printf() call. This changed
395 for version 4.90 - adding a "more data" boolean to the arguments.
396 Bumped the ABI version number also, this having been missed previously;
397 release versions 4.90 to 4.92.3 inclusive were effectively broken in
398 respect of usage of smtp_printf() by either local_scan code or libraries
399 accessed via the ${dlfunc } expansion item. Both will need coding
400 adjustment for any calls to smtp_printf() to match the new function
401 signature; a FALSE value for the new argument is always safe.
403 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
404 the file-offset (which the Linux syscall does, and exim expects); this
405 resulted in an indefinite loop.
407 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
408 to do ARC verification. The Authentication-Results: header line added
409 by the configuration then had no ARC item.
415 JH/01 Remove code calling the customisable local_scan function, unless a new
416 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
418 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
419 non-signal-safe functions being used.
421 JH/03 Bug 2269: When presented with a received message having a stupidly large
422 number of DKIM-Signature headers, disable DKIM verification to avoid
423 a resource-consumption attack. The limit is set at twenty.
425 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
426 report of oldest_pass in ${authres } in consequence, and separate out
427 some descriptions of reasons for verification fail.
429 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
430 files in the spool were present and unlocked. A queue-runner could spot
431 them, resulting in a duplicate delivery. Fix that by doing the unlock
432 after the unlink. Investigation by Tim Stewart. Take the opportunity to
433 add more error-checking on spoolfile handling while that code is being
436 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
437 No known attacks, no CVE, this is defensive hardening.
439 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
440 a queue-runner could start a delivery while other operations were ongoing.
441 Cutthrough delivery was a common victim, resulting in duplicate delivery.
442 Found and investigated by Tim Stewart. Fix by using the open message data
443 file handle rather than opening another, and not locally closing it (which
444 releases a lock) for that case, while creating the temporary .eml format
445 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
447 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
448 $sender_verify_failure/$recipient_verify_failure to "random".
450 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
453 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
454 Previously this would segfault.
456 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
459 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
460 like zero, since the resolver should be doing this for us, But we need one
461 as a CNAME but no MX presence gets the CNAME returned; we need to check
462 that doesn't point to an MX to declare it "no MX returned" rather than
463 "error, loop". A new main option is added so the older capability of
464 following some limited number of chain links is maintained.
466 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
468 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
469 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
470 now-deprecated earlier definitions used only specified the range up to TLS
471 1.2 (in the older-version library docs).
473 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
475 JH/15 Rework TLS client-side context management. Stop using a global, and
476 explicitly pass a context around. This enables future use of TLS for
477 connections to service-daemons (eg. malware scanning) while a client smtp
478 connection is using TLS; with cutthrough connections this is quite likely.
480 JH/16 Fix ARC verification to do AS checks in reverse order.
482 JH/17 Support a "tls" option on the ${readsocket } expansion item.
484 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
485 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
486 Previously the "utf8" would be re-prepended for every additional message.
488 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
489 Previously thery were accepted, resulting in issues when attempting to
490 forward messages to a non-supporting MTA.
492 PP/02 Let -n work with printing macros too, not just options.
494 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
495 one parent address was copied, and bogus data was used at delivery-logging
496 time. Either a crash (after delivery) or bogus log data could result.
497 Discovery and analysis by Tim Stewart.
499 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
500 Previously if the string ended mid-character, we did not insert the
501 promised '?' replacement.
503 PP/04 Documentation: current string operators work on bytes, not codepoints.
505 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
506 should pack well giving a smaller memory footprint so better caching and
507 therefore performance. Group the declarations where this can't be done so
508 that the byte-sized flag variables are not interspersed among pointer
509 variables, giving a better chance of good packing by the compiler.
511 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
512 non-null, to avoid issues with sites running BATV. Previously reports were
513 sent with an empty envelope sender so looked like bounces.
515 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
516 The ignore_error flag wasn't being returned from the filter subprocess so
517 was not set for later routers. Investigation and fix by Matthias Kurz.
519 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
520 and a msg:complete for the whole, when a message is manually removed using
521 -Mrm. Developement by Matthias Kurz, hacked on by JH.
523 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
524 a "Gnu special" function, asprintf() in the DB utility binary builds; I
525 hope that is portable enough.
527 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
528 requiring a known-CA anchor certificate; make it now rely entirely on the
529 TLSA as an anchor. Checking the name on the leaf cert against the name
530 on the A-record for the host is still done for TA (but not for EE mode).
532 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
533 included in delivery lines for non-proxied connections, when compiled with
534 SUPPORT_SOCKS and running with proxy logging enabled.
536 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
537 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
538 move the existing event to fire before the normal logging of message
539 failure so that custom logging is bracketed by normal logging.
541 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
542 msg:fail:internal event. Developement by Matthias Kurz.
544 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
545 far too small for todays use of crypto signatures stored there. Go all
546 the way to the max DNS message size of 64kB, even though this might be
547 overmuch for IOT constrained device use.
549 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
550 copy a string over itself. The library routine is documented as not
551 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
553 JH/32 For main options check_spool_space and check_inode_space, where the
554 platform supports 64b integers, support more than the previous 2^31 kB
555 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
556 the previous G, M, k.
558 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
559 $authenticated_fail_id variable on authentication failure. Previously
562 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
563 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
564 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
565 GNUTLS_SEC_PARAM_MEDIUM.
567 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
568 side. Previously we would continue as if no SNI had been received.
570 JH/36 Harden the handling of string-lists. When a list consisted of a sole
571 "<" character, which should be a list-separator specification, we walked
572 off past the nul-terimation.
574 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
575 causes) even when the retry time is not yet met. Previously they were
576 not, meaning that when (say) an account was over-quota and temp-rejecting,
577 and multiple senders' messages were queued, only one sender would get
578 notified on each configured delay_warning cycle.
580 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
582 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
583 cipher-suites, an error can be left on the stack even for a succeeding
584 accept; this results in impossible error messages when a later operation
587 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
588 return error codes indicating retry. Under TLS1.3 this becomes required.
590 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
591 it only wrote the new authenticators, resulting in a lack of tracking of
592 peer changes of ESMTP extensions until the next cache flush.
594 JH/41 Fix the loop reading a message header line to check for integer overflow,
595 and more-often against header_maxsize. Previously a crafted message could
596 induce a crash of the recive process; now the message is cleanly rejected.
598 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
599 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
605 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
606 When redis_servers is set to a list of > 1 element, and the Redis servers
607 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
608 case of MOVED into a DEFER case instead, thus moving the query onto the
609 next server in the list. For a cluster of N elements, all N servers must
610 be defined in redis_servers.
612 GF/02 Catch and remove uninitialized value warning in exiqsumm
613 Check for existence of @ARGV before looking at $ARGV[0]
615 JH/01 Replace the store_release() internal interface with store_newblock(),
616 which internalises the check required to safely use the old one, plus
617 the allocate and data copy operations duplicated in both (!) of the
618 extant use locations.
620 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
621 modifier. This matches the restriction on the commandline.
623 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
624 Previously only the last row was returned.
626 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
627 we assumed that tags in the header were well-formed, and parsed the
628 element content after inspecting only the first char of the tag.
629 Assumptions at that stage could crash the receive process on malformed
632 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
633 While running the DKIM ACL we operate on the Permanent memory pool so that
634 variables created with "set" persist to the DATA ACL. Also (at any time)
635 DNS lookups that fail create cache records using the Permanent pool. But
636 expansions release any allocations made on the current pool - so a dnsdb
637 lookup expansion done in the DKIM ACL releases the memory used for the
638 DNS negative-cache, and bad things result. Solution is to switch to the
639 Main pool for expansions.
640 While we're in that code, add checks on the DNS cache during store_reset,
641 active in the testsuite.
642 Problem spotted, and debugging aided, by Wolfgang Breyha.
644 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
645 When none of the hosts presented to a transport match an already-open
646 connection, close it and proceed with the list. Previously we would
647 queue the message. Spotted by Lena with Yahoo, probably involving
650 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
651 Previously a spurious "250 OK id=" response was appended to the proper
654 JH/08 The "support for" informational output now, which built with Content
655 Scanning support, has a line for the malware scanner interfaces compiled
656 in. Interface can be individually included or not at build time.
658 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
659 by the template makefile "src/EDITME". The "STREAM" support for an older
660 ClamAV interface method is removed.
662 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
663 rows affected is given instead).
665 JH/11 The runtime Berkeley DB library version is now additionally output by
666 "exim -d -bV". Previously only the compile-time version was shown.
668 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
669 SMTP connection. Previously, when one had more receipients than the
670 first, an abortive onward connection was made. Move to full support for
671 multiple onward connections in sequence, handling cutthrough connection
672 for all multi-message initiating connections.
674 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
675 routers. Previously, a multi-recipient message would fail to match the
676 onward-connection opened for the first recipient, and cause its closure.
678 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
679 a timeout on read on a GnuTLS initiating connection, resulting in the
680 initiating connection being dropped. This mattered most when the callout
681 was marked defer_ok. Fix to keep the two timeout-detection methods
684 JH/15 Relax results from ACL control request to enable cutthrough, in
685 unsupported situations, from error to silently (except under debug)
686 ignoring. This covers use with PRDR, frozen messages, queue-only and
689 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
691 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
692 metadata, resulting in a crash in free().
694 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
695 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
696 Broken also in d185889f4, with init system revamp.
698 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
699 Previously we abruptly closed the connection after reading a malware-
700 found indication; now we go on to read the "scan ok" response line,
703 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
704 ACL. Previously, a crash would result.
706 JH/19 Speed up macro lookups during configuration file read, by skipping non-
707 macro text after a replacement (previously it was only once per line) and
708 by skipping builtin macros when searching for an uppercase lead character.
710 JH/20 DANE support moved from Experimental to mainline. The Makefile control
711 for the build is renamed.
713 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
714 was allocated for every new TLS startup, meaning one per message. Fix
715 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
717 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
718 reported the original. Fix to report (as far as possible) the ACL
719 result replacing the original.
721 JH/23 Fix memory leak during multi-message connections using STARTTLS under
722 OpenSSL. Certificate information is loaded for every new TLS startup,
723 and the resources needed to be freed.
725 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
727 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
730 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
731 DATA response info to the (existing) per-recipient response info for
732 the "C=" log element. It can have useful tracking info from the
733 destination system. Patch from Simon Arlott.
735 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
736 length value. Previously this would segfault.
738 HS/02 Support Avast multiline protoocol, this allows passing flags to
739 newer versions of the scanner.
741 JH/28 Ensure that variables possibly set during message acceptance are marked
742 dead before release of memory in the daemon loop. This stops complaints
743 about them when the debug_store option is enabled. Discovered specifically
744 for sender_rate_period, but applies to a whole set of variables.
745 Do the same for the queue-runner and queue-list loops, for variables set
746 from spool message files. Do the same for the SMTP per-message loop, for
747 certain variables indirectly set in ACL operations.
749 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
750 as a multi-recipient message from a mailinglist manager). The coding had
751 an arbitrary cutoff number of characters while checking for more input;
752 enforced by writing a NUL into the buffer. This corrupted long / fast
753 input. The problem was exposed more widely when more pipelineing of SMTP
754 responses was introduced, and one Exim system was feeding another.
755 The symptom is log complaints of SMTP syntax error (NUL chars) on the
756 receiving system, and refused recipients seen by the sending system
757 (propating to people being dropped from mailing lists).
758 Discovered and pinpointed by David Carter.
760 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
761 replaced by the ${authresults } expansion.
763 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
765 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
766 allows proper process termination in container environments.
768 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
769 Previously the "final dot" had a newline after it; ensure it is CR,LF.
771 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
772 and "err_perm", deprecated since 4.83 when the RFC-defined words
773 "temperror" and "permerror" were introduced.
775 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
776 transport-filters or DKIM-signing. The restriction was lost in the
777 consolidation of verify-callout and delivery SMTP handling.
778 Extend the restriction to also cover ARC-signing.
780 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
781 in defer=pass mode supply a 450 to the initiator. Previously the message
784 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
785 tls_require_ciphers is used as before.
787 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
788 "pass_unscanned". Only tmpfails from the scanner are written to
789 the paniclog, as they may require admin intervention (permission
790 denied, license issues). Other scanner errors (like decompression
791 bombs) do not cause a paniclog entry.
793 JH/36 Fix reinitialisation of DKIM logging variable between messages.
794 Previously it was possible to log spurious information in receive log
797 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
798 triggered odd behaviour from Outlook Express clients.
800 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
803 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
804 since the IETF WG has not yet settled on that versus the original
805 "bare" representation.
807 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
808 Previously the millisecond value corrupted the output.
809 Fix also for syslog_pid=no and log_selector +pid, for which the pid
810 corrupted the output.
816 JH/01 Rework error string handling in TLS interface so that the caller in
817 more cases is responsible for logging. This permits library-sourced
818 string to be attached to addresses during delivery, and collapses
819 pairs of long lines into single ones.
821 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
822 during configuration. Wildcards are allowed and expanded.
824 JH/02 Rework error string handling in DKIM to pass more info back to callers.
825 This permits better logging.
827 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
828 do not close it down and have the child transport start it up again on
829 the passed-on TCP connection. Instead, proxy the child (and any
830 subsequent ones) for TLS via a unix-domain socket channel. Logging is
831 affected: the continued delivery log lines do not have any DNSSEC, TLS
832 Certificate or OCSP information. TLS cipher information is still logged.
834 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
835 identical IP addresses on different listening ports. Will also affect
838 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
839 add noisy ifdef guards to special-case this sillyness.
840 Patch from Bernd Kuhls.
842 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
843 than 255 are no longer allowed.
845 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
846 Disable the session-cache too, which might reduce our load. Since we
847 currrectly use a new context for every connection, both as server and
848 client, there is no benefit for these.
849 GnuTLS appears to not support tickets server-side by default (we don't
850 call gnutls_session_ticket_enable_server()) but client side is enabled
851 by default on recent versions (3.1.3 +) unless the PFS priority string
854 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
855 <https://reproducible-builds.org/specs/source-date-epoch/>.
857 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
858 the check for any unsuccessful recipients did not notice the limit, and
859 erroneously found still-pending ones.
861 JH/08 Pipeline CHUNKING command and data together, on kernels that support
862 MSG_MORE. Only in-clear (not on TLS connections).
864 JH/09 Avoid using a temporary file during transport using dkim. Unless a
865 transport-filter is involved we can buffer the headers in memory for
866 creating the signature, and read the spool data file once for the
867 signature and again for transmission.
869 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
870 4.77 as the kernel support then wasn't solid, having issues in 64bit
871 mode. Now, it's been long enough. Add support for FreeBSD also.
873 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
874 case where the routing stage had gathered several addresses to send to
875 a host before calling the transport for the first, we previously failed
876 to close down TLS in the old transport process before passing the TCP
877 connection to the new process. The new one sent a STARTTLS command
878 which naturally failed, giving a failed delivery and bloating the retry
879 database. Investigation and fix prototype from Wolfgang Breyha.
881 JH/12 Fix check on SMTP command input synchronisation. Previously there were
882 false-negatives in the check that the sender had not preempted a response
883 or prompt from Exim (running as a server), due to that code's lack of
884 awareness of the SMTP input buffering.
886 PP/04 Add commandline_checks_require_admin option.
887 Exim drops privileges sanely, various checks such as -be aren't a
888 security problem, as long as you trust local users with access to their
889 own account. When invoked by services which pass untrusted data to
890 Exim, this might be an issue. Set this option in main configuration
891 AND make fixes to the calling application, such as using `--` to stop
894 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
895 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
896 responses to those, into a single TLS record each way (this usually means
897 a single packet). As a side issue, smtp_enforce_sync now works on TLS
900 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
901 affects you only if you're dancing at the edge of the param size limits.
902 If you are, and this message makes sense to you, then: raise the
903 configured limit or use OpenSSL 1.1. Nothing we can do for older
906 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
907 cmdline element to get the documented default one. Previously it was
910 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
911 are unsuable for non-message-specific reasons - eg. connection timeout,
912 banner-time rejection.
914 JH/16 Fix logging of delivery remote port, when specified by router, under
917 PP/06 Repair manualroute's ability to take options in any order, even if one
918 is the name of a transport.
921 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
923 JH/17 Change the list-building routines interface to use the expanding-string
924 triplet model, for better allocation and copying behaviour.
926 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
927 Previously it was constructed the first time a possibly-matching string
928 was met in the configuration file input during startup; now it is done
931 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
932 compatible one, to avoid the (poorly documented) possibility of a config
933 file in the working directory redirecting the DB files, possibly correpting
934 some existing file. CVE-2017-10140 assigned for BDB.
936 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
937 cache-hot. Previously, although the result was properly cached, the
938 initial verify call returned a defer.
940 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
941 the main verify for receipient in uncached-mode.
943 JH/22 Retire historical build files to an "unsupported" subdir. These are
944 defined as "ones for which we have no current evidence of testing".
946 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
947 if present. Previously it was ignored.
949 JH/24 Start using specified-initialisers in C structure init coding. This is
950 a C99 feature (it's 2017, so now considered safe).
952 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
953 if was a fixed-sized field and bitmask ops via macros; it is now more
956 PP/07 GitHub PR 56: Apply MariaDB build fix.
957 Patch provided by Jaroslav Škarvada.
959 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
960 during Coverity cleanups [4.87 JH/47]
961 Diagnosis and fix provided by Michael Fischer v. Mollard.
963 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
964 the right size to place the terminating semicolon on its own folded
965 line, the header hash was calculated to an incorrect value thanks to
966 the (relaxed) space the fold became.
968 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
969 and confused the parent.
971 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
972 which could crash as a result. This could lead to undeliverable messages.
974 JH/28 Logging: "next input sent too soon" now shows where input was truncated
977 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
978 matters on fast-turnover and PID-randomising systems, which were getting
979 out-of-order delivery.
981 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
982 a possibly-overlapping copy. The symptom was that "Remote host closed
983 connection in response to HELO" was logged instead of the actual 4xx
986 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
987 Previously only that bufferd was discarded, resulting in SYMTP command
990 JH/32 DKIM: when a message has multiple signatures matching an identity given
991 in dkim_verify_signers, run the dkim acl once for each. Previously only
992 one run was done. Bug 2189.
994 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
995 "panic the current process" to "deliberately defer". The panic log is
996 still written with the problem list name; the mail and reject logs now
997 get a temp-reject line for the message that was being handled, saying
998 something like "domains check lookup or other defer". The SMTP 451
999 message is still "Temporary local problem".
1001 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1002 A crafted sequence of BDAT commands could result in in-use memory beeing
1003 freed. CVE-2017-16943.
1005 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1006 from SMTP input. Previously it was always done; now only done for DATA
1007 and not BDAT commands. CVE-2017-16944.
1009 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1010 to the message (such as an overlong header line). Previously this was
1011 not done and we did not exit BDAT mode. Followon from the previous item
1012 though a different problem.
1018 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1019 than -2003 did; needs libidn2 in addition to libidn.
1021 JH/02 The path option on a pipe transport is now expanded before use.
1023 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1024 Patch provided by "Björn", documentation fix added too.
1026 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1027 missing a wire-to-host endian conversion.
1029 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1030 close after a BDAT command line could be taken as a following command,
1031 giving a synch failure. Fix by only checking for synch immediately
1032 before acknowledging the chunk.
1034 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1035 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1036 macro. Patches provided by Josh Soref.
1038 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1039 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1040 by IANA as a service extension. However, John Klensin suggests that we
1043 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1044 the dkim code may be unix-mode line endings rather than smtp wire-format
1045 CRLF, so prepend a CR to any bare LF.
1047 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1048 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1050 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1051 the first were themselves being wrongly included in the feed into dkim
1052 processing; with most chunk sizes in use this resulted in an incorrect
1053 body hash calculated value.
1055 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1056 DKIM signature block, for verification. Although advised against by
1057 standards it is specifically not ruled illegal.
1059 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1061 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1062 missing a body hash (the bh= tag).
1064 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1065 It seems that HAProxy sends the Proxy Protocol information in clear and
1066 only then does a TLS startup, so do the same.
1068 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1069 TCP connections (such as for Spamd) unless the daemon successfully set
1070 Fast Open mode on its listening sockets. This fixes breakage seen on
1071 too-old kernels or those not configured for Fast Open, at the cost of
1072 requiring both directions being enabled for TFO, and TFO never being used
1073 by non-daemon-related Exim processes.
1075 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1076 endings, at least on the first header line. Try to canonify any that get
1077 past that check, despite the cost.
1079 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1080 now limited to an arbitrary five deep, while parsing addresses with the
1081 strip_excess_angle_brackets option enabled.
1083 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1084 instead leave the unprompted TLS handshake in socket buffer for the
1085 TLS library to consume.
1087 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1089 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1091 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1092 region in one operation, for speed, and this leaves assigned pointers
1093 dangling. Add checks run only under the testsuite which checks all
1094 variables at a store-reset and panics on a dangling pointer; add code
1095 explicitly nulling out all the variables discovered. Fixes one known
1096 bug: a transport crash, where a dangling pointer for $sending_ip_address
1097 originally assigned in a verify callout, is re-used.
1099 PP/06 Drop '.' from @INC in various Perl scripts.
1101 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1103 PP/08 Reduce a number of compilation warnings under clang; building with
1104 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1105 should be warning-free.
1107 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1109 HS/01 Fix portability problems introduced by PP/08 for platforms where
1110 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1112 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1113 chunk. This allows us to accept broken chunked messages. We need a more
1114 general solution here.
1116 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1117 already-broken messages in the queue.
1119 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1121 JH/19 Fix reference counting bug in routing-generated-address tracking.
1127 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1128 supports it and a size is available (ie. the sending peer gave us one).
1130 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1131 years of being deprecated). The replacements are the ACLs
1132 acl_smtp_mime and acl_not_smtp_mime.
1134 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1135 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1136 or even an in-clear connection were permitted. Now, if the host lookup
1137 was dnssec and dane was requested then the host is only used if the
1138 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1139 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1140 if one fails this test.
1141 This means that a poorly-configured remote DNS will make it incommunicado;
1142 but it protects against a DNS-interception attack on it.
1144 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1145 non-noisy when a race steals the message being considered.
1147 JH/05 If main configuration option tls_certificate is unset, generate a
1148 self-signed certificate for inbound TLS connections.
1150 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1151 in rewrites and routers.
1153 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1154 and logged a warning sing 4.83; now they are a configuration file error.
1156 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1157 (lacking @domain). Apply the same qualification processing as RCPT.
1159 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1161 JH/10 Support ${sha256:} applied to a string (as well as the previous
1164 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1165 a cutthrough deliver is pending, as we always want to make a connection.
1166 This also avoids re-routing the message when later placing the cutthrough
1167 connection after a verify cache hit.
1168 Do not update it with the verify result either.
1170 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1171 when routing results in more than one destination address.
1173 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1174 signing (which inhibits the cutthrough capability). Previously only
1175 the presence of an option was tested; now an expansion evaluating as
1176 empty is permissible (obviously it should depend only on data available
1177 when the cutthrough connection is made).
1179 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1180 the relevant preceding SMTP command did not note the pipelining mode.
1182 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1183 Previously they were not counted.
1185 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1186 as one having no matching records. Previously we deferred the message
1187 that needed the lookup.
1189 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1190 distinguished as "(=".
1192 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1193 for missing MX records. Previously it only worked for missing A records.
1195 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1197 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1198 after the data-go-ahead and data-ack. Patch from Jason Betts.
1200 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1201 even for a "none" policy. Patch from Tony Meyer.
1203 JH/22 Fix continued use of a connection for further deliveries. If a port was
1204 specified by a router, it must also match for the delivery to be
1207 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1208 When one of the recipients of a message was unsuitable for the connection
1209 (has no matching addresses), we lost track of needing to mark it
1210 deferred. As a result mail would be lost.
1212 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1214 JH/25 Decoding ACL controls is now done using a binary search; the source code
1215 takes up less space and should be simpler to maintain. Merge the ACL
1216 condition decode tables also, with similar effect.
1218 JH/26 Fix problem with one_time used on a redirect router which returned the
1219 parent address unchanged. A retry would see the parent address marked as
1220 delivered, so not attempt the (identical) child. As a result mail would
1223 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1224 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1225 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1228 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1229 Main config options check_{log,spool}_{inodes,space} are now
1230 100 inodes, 10MB unless set otherwise in the configuration.
1232 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1233 Previously it only applied to the main-section connection policy
1236 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1238 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1239 by me. Added RFC7919 DH primes as an alternative.
1241 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1242 are not in the system include path.
1244 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1245 GnuTLS, when a session startup failed (eg because the client disconnected)
1246 Exim did stdio operations after fclose. This was exposed by a recent
1247 change which nulled out the file handle after the fclose.
1249 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1250 signed directly by the cert-signing cert, rather than an intermediate
1251 OCSP-signing cert. This is the model used by LetsEncrypt.
1253 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1255 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1256 an incoming connection.
1258 HS/02 Bug 1802: Do not half-close the connection after sending a request
1261 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1262 fallback to "prime256v1".
1264 JH/34 SECURITY: Use proper copy of DATA command in error message.
1265 Could leak key material. Remotely exploitable. CVE-2016-9963.
1271 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1272 and 3.4.4 - once the server is enabled to respond to an OCSP request
1273 it does even when not requested, resulting in a stapling non-aware
1274 client dropping the TLS connection.
1276 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1277 support variable-length bit vectors. No functional change.
1279 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1280 The I= interface field on outgoing lines is now after the H= remote
1281 host field, same as incoming lines. There is a separate
1282 outgoing_interface log selector which allows you to disable the
1285 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1286 If not running log_selector +smtp_connection the mainlog would be held
1287 open indefinitely after a "too many connections" event, including to a
1288 deleted file after a log rotate. Leave the per net connection logging
1289 leaving it open for efficiency as that will be quickly detected by the
1290 check on the next write.
1292 HS/01 Bug 1671: Fix post transport crash.
1293 Processing the wait-<transport> messages could crash the delivery
1294 process if the message IDs didn't exist for some reason. When
1295 using 'split_spool_directory=yes' the construction of the spool
1296 file name failed already, exposing the same netto behaviour.
1298 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1299 mime_regex ACL conditions.
1301 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1302 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1303 to HELO, local diagnostic string.
1305 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1306 log line to debug. Even when configured with a tls authenticator many
1307 client connections are expected to not authenticate in this way, so
1308 an authenticate fail is not an error.
1310 HS/02 Add the Exim version string to the process info. This way exiwhat
1311 gives some more detail about the running daemon.
1313 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1314 matter for fast-change records such as DNSBLs.
1316 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1317 retry record, even if constant. There may be multiple transports with
1318 different interface settings and the retry behaviour needs to be kept
1321 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1323 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1325 JH/10 Bug 840: fix log_defer_output option of pipe transport
1327 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1328 hosts_randomize. This matters a lot when combined with mua_wrapper.
1330 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1331 ${quote_pgsql:<string>} operator.
1333 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1335 JH/14 Change header file name for freeradius-client. Relevant if compiling
1336 with Radius support; from the Gentoo tree and checked under Fedora.
1338 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1340 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1341 option transport_filter, meaning no filtering.
1343 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1345 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1346 defaults to "*" (all hosts). The variable is now available when not built
1347 with TLS, default unset, mainly to enable keeping the testsuite sane.
1348 If a server certificate is not supplied (via tls_certificate) an error is
1349 logged, and clients will find TLS connections fail on startup. Presumably
1350 they will retry in-clear.
1351 Packagers of Exim are strongly encouraged to create a server certificate
1352 at installation time.
1354 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1355 with the $config_file variable.
1357 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1358 in transport context, after the attempt, and per-recipient. The latter type
1359 is per host attempted. The event data is the error message, and the errno
1360 information encodes the lookup type (A vs. MX) used for the (first) host,
1361 and the trailing two digits of the smtp 4xx response.
1363 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1364 to write to mainlog (or rejectlog, paniclog) in the window between file
1365 creation and permissions/ownership being changed. Particularly affects
1366 installations where exicyclog is run as root, rather than exim user;
1367 result is that the running daemon panics and dies.
1369 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1371 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1372 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1373 "pri" and "weight". Note that the previous implicit priority given by the
1374 list order is no longer honoured.
1376 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1377 for DKIM processing.
1379 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1380 by defining SUPPORT_SOCKS.
1382 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1383 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1384 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1385 variables are renamed to proxy_{local,external}_{address,port}.
1387 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1388 for a build by defining SUPPORT_I18N
1390 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1391 of the query string, and make ${quote_redis:} do that quoting.
1393 JH/29 Move Events support from Experimental to mainline, enabled by default
1394 and removable for a build by defining DISABLE_EVENT.
1396 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1398 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1399 cached by the daemon.
1401 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1402 by defining LOOKUP_REDIS. The libhiredis library is required.
1404 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1405 keys are given for lookup.
1407 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1408 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1409 only supported when built with TLS support. The PolarSSL SHA routines
1410 are still used when the TLS library is too old for convenient support.
1412 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1413 openssl_options), for security. OpenSSL forces this from version 1.1.0
1414 server-side so match that on older versions.
1416 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1417 allocation for $value could be released as the expansion processing
1418 concluded, but leaving the global pointer active for it.
1420 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1421 and to use the domains and local_parts ACL conditions.
1423 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1424 incorrectly not doubled on cutthrough transmission, hence seen as a
1425 body-termination at the receiving system - resulting in truncated mails.
1426 Commonly the sender saw a TCP-level error, and retransmitted the message
1427 via the normal store-and-forward channel. This could result in duplicates
1428 received - but deduplicating mailstores were liable to retain only the
1429 initial truncated version.
1431 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1433 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1435 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1436 we're in there, support oversigning also; bug 1309.
1438 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1440 HS/04 Add support for keep_environment and add_environment options.
1442 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1443 either intentional arithmetic overflow during PRNG, or testing config-
1446 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1447 delivery resulted in actual delivery. Cancel cutthrough before DATA
1450 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1451 rejects a recipient: pass the reject to the originator.
1453 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1454 Many were false-positives and ignorable, but it's worth fixing the
1457 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1458 for the new environment-manipulation done at startup. Move the routines
1459 from being local to tls.c to being global via the os.c file.
1461 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1462 an extract embedded as result-arg for a map, the first arg for extract
1463 is unavailable so we cannot tell if this is a numbered or keyed
1464 extraction. Accept either.
1470 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1473 JH/02 The smtp transport option "multi_domain" is now expanded.
1475 JH/03 The smtp transport now requests PRDR by default, if the server offers
1478 JH/04 Certificate name checking on server certificates, when exim is a client,
1479 is now done by default. The transport option tls_verify_cert_hostnames
1480 can be used to disable this per-host. The build option
1481 EXPERIMENTAL_CERTNAMES is withdrawn.
1483 JH/05 The value of the tls_verify_certificates smtp transport and main options
1484 default to the word "system" to access the system default CA bundle.
1485 For GnuTLS, only version 3.0.20 or later.
1487 JH/06 Verification of the server certificate for a TLS connection is now tried
1488 (but not required) by default. The verification status is now logged by
1489 default, for both outbound TLS and client-certificate supplying inbound
1492 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1495 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1496 Status Notification (bounce) messages are now MIME format per RFC 3464.
1497 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1498 under the control of the dsn_advertise_hosts option, and routers may
1499 have a dsn_lasthop option.
1501 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1502 default, modifiable by a malware= option. The list separator for
1503 the options can now be changed in the usual way. Bug 68.
1505 JH/10 The smtp_receive_timeout main option is now expanded before use.
1507 JH/11 The incoming_interface log option now also enables logging of the
1508 local interface on delivery outgoing connections.
1510 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1511 if the interface and destination host and port all match.
1513 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1516 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1517 Patch from Andrew Lewis.
1519 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1520 now supports optional time-restrictions, weighting, and priority
1521 modifiers per server. Patch originally by <rommer@active.by>.
1523 JH/16 The spamd_address main option now supports a mixed list of local
1524 and remote servers. Remote servers can be IPv6 addresses, and
1525 specify a port-range.
1527 JH/17 Bug 68: The spamd_address main option now supports an optional
1528 timeout value per server.
1530 JH/18 Bug 1581: Router and transport options headers_add/remove can
1531 now have the list separator specified.
1533 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1536 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1539 JH/21 Support for the A6 type of dns record is withdrawn.
1541 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1542 rather than the verbs used.
1544 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1545 from 255 to 1024 chars.
1547 JH/24 Verification callouts now attempt to use TLS by default.
1549 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1550 are generic router options now. The defaults didn't change.
1552 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1553 Original patch from Alexander Shikoff, worked over by JH.
1555 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1556 files if ZCAT_COMMAND is not executable.
1558 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1560 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1562 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1563 Normally benign, it bites when the pair was led to by a CNAME;
1564 modern usage is to not canonicalize the domain to a CNAME target
1565 (and we were inconsistent anyway for A-only vs AAAA+A).
1567 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1569 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1570 when evaluating $sender_host_dnssec.
1572 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1573 $sender_helo_dnssec variable.
1575 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1577 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1579 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1581 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1582 documented as working, but never had. Support all but $spam_report.
1584 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1585 added for tls authenticator.
1587 HS/03 Add perl_taintmode main config option
1593 TL/01 When running the test suite, the README says that variables such as
1594 no_msglog_check are global and can be placed anywhere in a specific
1595 test's script, however it was observed that placement needed to be near
1596 the beginning for it to behave that way. Changed the runtest perl
1597 script to read through the entire script once to detect and set these
1598 variables, reset to the beginning of the script, and then run through
1599 the script parsing/test process like normal.
1601 TL/02 The BSD's have an arc4random API. One of the functions to induce
1602 adding randomness was arc4random_stir(), but it has been removed in
1603 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1604 function when detected.
1606 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1607 cause callback expansion.
1609 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1610 syntax errors in an expansion can be treated as a string instead of
1611 logging or causing an error, due to the internal use of bool_lax
1612 instead of bool when processing it.
1614 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1615 server certificates when making smtp deliveries.
1617 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1619 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1621 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1623 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1624 Merged patch from Sebastian Wiedenroth.
1626 JH/05 Fix results-pipe from transport process. Several recipients, combined
1627 with certificate use, exposed issues where response data items split
1628 over buffer boundaries were not parsed properly. This eventually
1629 resulted in duplicates being sent. This issue only became common enough
1630 to notice due to the introduction of connection certificate information,
1631 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1633 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1634 size buffer was used, resulting in syntax errors when an expansion
1637 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1638 version 3.3.6 or later.
1640 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1641 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1642 both become "event_action", the variables become $event_name, $event_data
1643 and $event_defer_errno. There is a new variable $verify_mode, usable in
1644 routers, transports and related events. The tls:cert event is now also
1645 raised for inbound connections, if the main configuration event_action
1648 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1649 early OCSP support, but no stapling (appears to be less than 1.0.0).
1651 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1652 server certificate names available under the smtp transport option
1653 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1656 JH/10 Time-related extraction expansions from certificates now use the main
1657 option "timezone" setting for output formatting, and are consistent
1658 between OpenSSL and GnuTLS compilations. Bug 1541.
1660 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1661 encoded parameter in the incoming message. Bug 1558.
1663 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1664 include certificate info, eximon was claiming there were spoolfile
1667 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1669 JH/14 Log delivery-related information more consistently, using the sequence
1670 "H=<name> [<ip>]" wherever possible.
1672 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1673 are problematic for Debian distribution, omit them from the release
1676 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1678 JH/16 Fix string representation of time values on 64bit time_t architectures.
1681 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1682 output list separator was used.
1687 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1688 checkers that were complaining about end of non-void function with no
1691 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1692 This was a regression introduced in 4.83 by another bugfix.
1694 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1696 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1697 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1703 TF/01 Correctly close the server side of TLS when forking for delivery.
1705 When a message was received over SMTP with TLS, Exim failed to clear up
1706 the incoming connection properly after forking off the child process to
1707 deliver the message. In some situations the subsequent outgoing
1708 delivery connection happened to have the same fd number as the incoming
1709 connection previously had. Exim would try to use TLS and fail, logging
1710 a "Bad file descriptor" error.
1712 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1713 utilities have not been installed.
1715 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1716 temporary space as the ACL may create new global variables.
1718 TL/01 LDAP support uses per connection or global context settings, depending
1719 upon the detected version of the libraries at build time.
1721 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1722 to extract and use the src ip:port in logging and expansions as if it
1723 were a direct connection from the outside internet. PPv2 support was
1724 updated based on HAProxy spec change in May 2014.
1726 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1728 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1729 Properly escape header and check for NULL return.
1731 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1734 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1736 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1737 characters in header names, implemented as a verify condition.
1738 Contributed by Michael Fischer v. Mollard.
1740 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1741 results permerror and temperror. Previous values are deprecated but
1742 still accepted. In a future release, err_perm and err_temp will be
1743 completely removed, which will be a backward incompatibility if the
1744 ACL tests for either of these two old results. Patch contributed by
1745 user bes-internal on the mailing list.
1747 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1749 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1750 selectors, in both main and reject logs.
1752 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1755 JH/07 Add malware type "sock" for talking to simple daemon.
1757 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1759 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1760 routers/transports under cutthrough routing.
1762 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1763 numbers. Touch up "bool" conditional to keep the same definition.
1765 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1767 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1769 JH/12 Expand items in router/transport headers_add or headers_remove lists
1770 individually rather than the list as a whole. Bug 1452.
1772 Required for reasonable handling of multiple headers_ options when
1773 they may be empty; requires that headers_remove items with embedded
1774 colons must have them doubled (or the list-separator changed).
1776 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1777 view the policy declared in the DMARC record. Currently, $dmarc_status
1778 is a combined value of both the record presence and the result of the
1781 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1783 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1784 dnslookup router and the smtp transport (applying to the forward
1787 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1788 of ldap servers used for a specific lookup. Patch provided by Heiko
1791 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1792 New variable $lookup_dnssec_authenticated for observability.
1794 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1795 Patch submitted by Lars Timman.
1797 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1799 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1800 Requires trusted mode and valid format message id, aborts otherwise.
1801 Patch contributed by Heiko Schlichting.
1803 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1804 certextract with support for various fields. Bug 1358.
1806 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1807 is requested by default, modifiable by smtp transport option
1810 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1811 operate on certificate variables to give certificate fingerprints
1812 Also new ${sha256:cert_variable}.
1814 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1816 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1817 Christian Aistleitner.
1819 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1821 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1822 file. Patch from Wolfgang Breyha.
1824 JH/25 Expand the coverage of the delivery $host and $host_address to
1825 client authenticators run in verify callout. Bug 1476.
1827 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1828 align with daemon_smtp_ports. Bug 72.
1830 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1831 support and error reporting did not work properly.
1833 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1834 and is readable. Patch from Andrew Colin Kissa.
1836 TL/14 Enhance documentation of ${run expansion and how it parses the
1837 commandline after expansion, particularly in the case when an
1838 unquoted variable expansion results in an empty value.
1840 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1842 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1843 by renaming away from it. Addresses GH issue 15, reported by
1846 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1848 TL/15 SECURITY: prevent double expansion in math comparison functions
1849 (can expand unsanitized data). Not remotely exploitable.
1856 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1858 PP/02 Make -n do something, by making it not do something.
1859 When combined with -bP, the name of an option is not output.
1861 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1864 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1865 $sender_host_name and config options to manage this, and basic check
1868 PP/05 DSCP support for outbound connections and control modifier for inbound.
1870 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1871 (Only plugin which currently uses this is kerberos4, which nobody should
1872 be using, but we should make it available and other future plugins might
1873 conceivably use it, even though it would break NAT; stuff *should* be
1874 using channel bindings instead).
1876 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1877 name; added for Sendmail compatibility; requires admin caller.
1878 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1879 just ignore it); requires trusted caller.
1880 Also parse but ignore: -Ac -Am -X<logfile>
1883 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1885 TL/02 Add +smtp_confirmation as a default logging option.
1887 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1888 Patch by Magnus Holmgren from 2007-02-20.
1890 TL/04 Bugzilla 1281 - Spec typo.
1891 Bugzilla 1283 - Spec typo.
1892 Bugzilla 1290 - Spec grammar fixes.
1894 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1896 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1898 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1899 mailing list by by Dmitry Isaikin.
1901 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1903 JH/02 Support "G" suffix to numbers in ${if comparisons.
1905 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1907 NM/01 Bugzilla 1197 - Spec typo
1908 Bugzilla 1196 - Spec examples corrections
1910 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1912 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1913 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1916 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1917 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1919 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1920 "acl {{name}{arg}...}", and optional args on acl condition
1923 JH/05 Permit multiple router/transport headers_add/remove lines.
1925 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1927 JH/07 Avoid using a waiting database for a single-message-only transport.
1928 Performance patch from Paul Fisher. Bugzilla 1262.
1930 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1933 JH/09 Add $headers_added variable, with content from use of ACL modifier
1934 add_header (but not yet added to the message). Bugzilla 199.
1936 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1937 Pulled from Bugzilla 817 by Wolfgang Breyha.
1939 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1941 (nb: this is the same fix as in Exim 4.80.1)
1943 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1946 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1949 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1951 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1952 advertise SMTP AUTH mechanism to us, instead of a generic
1953 protocol violation error. Also, make Exim more robust to bad
1954 data from the Dovecot auth socket.
1956 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1958 When a queue runner is handling a message, Exim first routes the
1959 recipient addresses, during which it prunes them based on the retry
1960 hints database. After that it attempts to deliver the message to
1961 any remaining recipients. It then updates the hints database using
1964 So if a recipient address works intermittently, it can get repeatedly
1965 deferred at routing time. The retry hints record remains fresh so the
1966 address never reaches the final cutoff time.
1968 This is a fairly common occurrence when a user is bumping up against
1969 their storage quota. Exim had some logic in its local delivery code
1970 to deal with this. However it did not apply to per-recipient defers
1971 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1973 This change adds a proper retry rule check during routing so that the
1974 final cutoff time is checked against the message's age. We only do
1975 this check if there is an address retry record and there is not a
1976 domain retry record; this implies that previous attempts to handle
1977 the address had the retry_use_local_parts option turned on. We use
1978 this as an approximation for the destination being like a local
1979 delivery, as in LMTP.
1981 I suspect this new check makes the old local delivery cutoff check
1982 redundant, but I have not verified this so I left the code in place.
1984 TF/02 Correct gecos expansion when From: is a prefix of the username.
1986 Test 0254 submits a message to Exim with the header
1990 When I ran the test suite under the user fanf2, Exim expanded
1991 the header to contain my full name, whereas it should have added
1992 a Resent-Sender: header. It erroneously treats any prefix of the
1993 username as equal to the username.
1995 This change corrects that bug.
1997 GF/01 DCC debug and logging tidyup
1998 Error conditions log to paniclog rather than rejectlog.
1999 Debug lines prefixed by "DCC: " to remove any ambiguity.
2001 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2003 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2004 Bug spotted by Jeremy Harris; was flawed since initial commit.
2005 Would have resulted in OCSP responses post-SNI triggering an Exim
2006 NULL dereference and crash.
2008 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2010 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2011 Bug detection, analysis and fix by Samuel Thibault.
2012 Bugzilla 1331, Debian bug #698092.
2014 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2016 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2017 Server implementation by Todd Lyons, client by JH.
2018 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2019 config variable "prdr_enable" controls whether the server
2020 advertises the facility. If the client requests PRDR a new
2021 acl_data_smtp_prdr ACL is called once for each recipient, after
2022 the body content is received and before the acl_smtp_data ACL.
2023 The client is controlled by both of: a hosts_try_prdr option
2024 on the smtp transport, and the server advertisement.
2025 Default client logging of deliveries and rejections involving
2026 PRDR are flagged with the string "PRDR".
2028 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2029 fclose(). Diagnosis by Todd Lyons.
2031 PP/17 Update configure.default to handle IPv6 localhost better.
2032 Patch by Alain Williams (plus minor tweaks).
2035 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2036 This is now consistent with GnuTLS, and is now documented: the
2037 previous undocumented portable approach to treating the option as
2038 unset was to force an expansion failure. That still works, and
2039 an empty string is now equivalent.
2041 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2042 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2043 not performing validation itself.
2045 PP/20 Added force_command boolean option to pipe transport.
2046 Patch from Nick Koston, of cPanel Inc.
2048 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2051 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2053 PP/21 Fix eximon continuous updating with timestamped log-files.
2054 Broken in a format-string cleanup in 4.80, missed when I repaired the
2055 other false fix of the same issue.
2056 Report and fix from Heiko Schlichting.
2059 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2060 Report from Prashanth Katuri.
2062 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2063 It's SecureTransport, so affects any MacOS clients which use the
2064 system-integrated TLS libraries, including email clients.
2066 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2067 using a MIME ACL for non-SMTP local injection.
2068 Report and assistance in diagnosis by Warren Baker.
2070 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2072 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2074 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2075 last id that failed so it may be referenced in subsequent ACL's.
2077 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2080 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2081 ldap library initialization, allowing self-signed CA's to be
2082 used. Also properly sets require_cert option later in code by
2083 using NULL (global ldap config) instead of ldap handle (per
2084 session). Bug diagnosis and testing by alxgomz.
2086 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2087 the src/util/ subdirectory.
2089 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2090 renamed to Transport Post Delivery Action by Jeremy Harris, as
2093 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2094 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2095 redis_servers = needs to be configured which will be used by the redis
2096 lookup. Patch from Warren Baker, of The Packet Hub.
2098 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2100 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2101 hostname or reverse DNS when processing a host list. Used suggestions
2102 from multiple comments on this bug.
2104 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2106 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2107 Missed a few lines, added it to make the runtest require no keyboard
2110 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2111 contains upper case chars. Make router use caseful_local_part.
2113 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2114 support when GnuTLS has been built with p11-kit.
2120 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2122 This, or similar/improved, will also be change PP/11 of 4.82.
2128 PP/01 Handle short writes when writing local log-files.
2129 In practice, only affects FreeBSD (8 onwards).
2130 Bugzilla 1053, with thanks to Dmitry Isaikin.
2132 NM/01 Bugzilla 949 - Documentation tweak
2134 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2137 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2139 PP/02 Implemented gsasl authenticator.
2141 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2143 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2144 `pkg-config foo` for cflags/libs.
2146 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2147 with rest of GSASL and with heimdal_gssapi.
2149 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2150 `pkg-config foo` for cflags/libs for the TLS implementation.
2152 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2153 properties get this fed in as external SSF. A number of robustness
2154 and debugging improvements to the cyrus_sasl authenticator.
2156 PP/08 cyrus_sasl server now expands the server_realm option.
2158 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2159 Patch by Jeremy Harris.
2161 PP/10 Added dbmjz lookup type.
2163 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2165 PP/12 MAIL args handles TAB as well as SP, for better interop with
2166 non-compliant senders.
2167 Analysis and variant patch by Todd Lyons.
2169 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2170 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2171 Patch from Dirk Mueller <dmueller@suse.com>
2173 PP/13 tls_peerdn now print-escaped for spool files.
2174 Observed some $tls_peerdn in wild which contained \n, which resulted
2175 in spool file corruption.
2177 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2178 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2179 or write after TLS renegotiation, which otherwise led to messages
2182 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2183 as a tracking header (ie: a signed header comes before the signature).
2184 Patch from Wolfgang Breyha.
2186 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2187 comma-sep list; embedded commas doubled.
2189 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2191 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2193 Report and patch from Dmitry Banschikov.
2195 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2196 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2197 failures. We appear to now support TLS1.1+ with Exim.
2199 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2200 lets Exim select keys and certificates based upon TLS SNI from client.
2201 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2202 before an outbound SMTP session. New log_selector, +tls_sni.
2204 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2205 NULL dereference. Report and patch from Alun Jones.
2207 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2208 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2209 Not seeing resolver debug output on NetBSD, but suspect this is a
2210 resolver implementation change.
2212 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2213 Left warnings. Added "eximon gdb" invocation mode.
2215 PP/21 Defaulting "accept_8bitmime" to true, not false.
2217 PP/22 Added -bw for inetd wait mode support.
2219 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2220 locate the relevant includes and libraries. Made this the default.
2222 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2223 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2225 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2226 This may cause build issues on older platforms.
2228 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2229 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2230 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2231 Added SNI support via GnuTLS too.
2232 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2234 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2236 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2237 Applied second patch from Janne, implementing suggestion to default
2238 multiple-strings-in-record handling to match SPF spec.
2240 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2242 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2243 read-only, out of scope).
2244 Patch from Wolfgang Breyha, report from Stuart Northfield.
2246 PP/29 Fix three issues highlighted by clang analyser static analysis.
2247 Only crash-plausible issue would require the Cambridge-specific
2248 iplookup router and a misconfiguration.
2249 Report from Marcin Mirosław.
2251 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2253 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2254 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2255 As part of this, removing so much warning spew let me fix some minor
2256 real issues in debug logging.
2258 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2259 assignment on my part. Fixed.
2261 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2262 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2263 Janne Snabb (who went above and beyond: thank you).
2265 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2266 string otherwise requires a connection and a bunch more work and it's
2267 relatively easy to get wrong. Should also expose TLS library linkage
2270 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2271 64-bit ${eval} (JH/03).
2273 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2274 GNU libc to support some of the 64-bit stuff, should not lead to
2275 conflicts. Defined before os.h is pulled in, so if a given platform
2276 needs to override this, it can.
2278 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2279 protection layer was required, which is not implemented.
2280 Bugzilla 1254, patch from Wolfgang Breyha.
2282 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2283 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2284 tls_dhparam take prime identifiers. Also unbreak combination of
2285 OpenSSL+DH_params+TLSSNI.
2287 PP/39 Disable SSLv2 by default in OpenSSL support.
2293 PP/01 Solaris build fix for Oracle's LDAP libraries.
2294 Bugzilla 1109, patch from Stephen Usher.
2296 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2298 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2301 TF/02 Fix a couple more cases where we did not log the error message
2302 when unlink() failed. See also change 4.74-TF/03.
2304 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2305 lock up or crash if it happened to be inside a call to libc when it
2306 got a SIGUSR1 from exiwhat.
2308 The SIGUSR1 handler appends the current process status to the process
2309 log which is later printed by exiwhat. It used to use the general
2310 purpose logging code to do this, but several functions it calls are
2311 not safe for signals.
2313 The new output code in the SIGUSR1 handler is specific to the process
2314 log, and simple enough that it's easy to inspect for signal safety.
2315 Removing some special cases also simplifies the general logging code.
2316 Removing the spurious timestamps from the process log simplifies
2319 TF/04 Improved ratelimit ACL condition.
2321 The /noupdate option has been deprecated in favour of /readonly which
2322 has clearer semantics. The /leaky, /strict, and /readonly update modes
2323 are mutually exclusive. The update mode is no longer included in the
2324 database key; it just determines when the database is updated. (This
2325 means that when you upgrade Exim will forget old rate measurements.)
2327 Exim now checks that the per_* options are used with an update mode that
2328 makes sense for the current ACL. For example, when Exim is processing a
2329 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2330 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2331 must specify per_mail/readonly. If you omit the update mode it defaults to
2332 /leaky where that makes sense (as before) or /readonly where required.
2334 The /noupdate option is now undocumented but still supported for
2335 backwards compatibility. It is equivalent to /readonly except that in
2336 ACLs where /readonly is required you may specify /leaky/noupdate or
2337 /strict/noupdate which are treated the same as /readonly.
2339 A useful new feature is the /count= option. This is a generalization
2340 of the per_byte option, so that you can measure the throughput of other
2341 aggregate values. For example, the per_byte option is now equivalent
2342 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2344 The per_rcpt option has been generalized using the /count= mechanism
2345 (though it's more complicated than the per_byte equivalence). When it is
2346 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2347 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2348 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2349 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2350 Note that using per_rcpt with a non-readonly update mode in more than
2351 one ACL will cause the recipients to be double-counted. (The per_mail
2352 and per_byte options don't have this problem.)
2354 The handling of very low rates has changed slightly. If the computed rate
2355 is less than the event's count (usually one) then this event is the first
2356 after a long gap. In this case the rate is set to the same as this event's
2357 count, so that the first message of a spam run is counted properly.
2359 The major new feature is a mechanism for counting the rate of unique
2360 events. The new per_addr option counts the number of different
2361 recipients that someone has sent messages to in the last time period. It
2362 behaves like per_rcpt if all the recipient addresses are different, but
2363 duplicate recipient addresses do not increase the measured rate. Like
2364 the /count= option this is a general mechanism, so the per_addr option
2365 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2366 example, measure the rate that a client uses different sender addresses
2367 with the options per_mail/unique=$sender_address. There are further
2368 details in the main documentation.
2370 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2372 TF/06 Removed a few PCRE remnants.
2374 TF/07 Automatically extract Exim's version number from tags in the git
2375 repository when doing development or release builds.
2377 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2378 Bugzilla 879. Patch from Paul Fisher.
2380 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2381 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2384 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2386 PP/05 Variable $av_failed, true if the AV scanner deferred.
2387 Bugzilla 1078. Patch from John Horne.
2389 PP/06 Stop make process more reliably on build failure.
2390 Bugzilla 1087. Patch from Heiko Schlittermann.
2392 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2393 Bugzilla 1089. Patch from Heiko Schlittermann.
2395 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2396 Bugzilla 1131. Patch from Holger Weiß.
2398 PP/09 Handle IPv6 addresses with SPF.
2399 Bugzilla 860. Patch from Wolfgang Breyha.
2401 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2403 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2406 PP/11 match_* no longer expand right-hand-side by default.
2407 New compile-time build option, EXPAND_LISTMATCH_RHS.
2408 New expansion conditions, "inlist", "inlisti".
2410 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2412 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2414 PP/14 fix log_write() format string regression from TF/03.
2415 Bugzilla 1152. Patch from Dmitry Isaikin.
2421 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2423 PP/02 Harmonised TLS library version reporting; only show if debugging.
2424 Layout now matches that introduced for other libraries in 4.74 PP/03.
2426 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2428 PP/04 New "dns_use_edns0" global option.
2430 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2433 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2434 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2436 TK/01 Updated PolarSSL code to 0.14.2.
2437 Bugzilla 1097. Patch from Andreas Metzler.
2439 PP/07 Catch divide-by-zero in ${eval:...}.
2440 Fixes bugzilla 1102.
2442 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2445 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2446 format-string attack -- SECURITY: remote arbitrary code execution.
2448 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2449 time unintentionally subject to list matching rules, letting the header
2450 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2451 arbitrary string expansion). This allowed for information disclosure.
2453 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2454 INT_MIN/-1 -- value coerced to INT_MAX.
2460 NM/01 Workaround for PCRE version dependency in version reporting
2463 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2464 This fixes portability to compilers other than gcc, notably
2465 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2467 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2468 makefiles for portability to HP-UX and POSIX correctness.
2470 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2471 Also via indented variable definition in the Makefile.
2472 (Debugging by Oliver Heesakkers).
2474 PP/02 Restore caching of spamd results with expanded spamd_address.
2475 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2477 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2478 Improves build reliability. Fix from: Frank Elsner
2480 NM/02 Fix wide character breakage in the rfc2047 coding
2481 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2483 NM/03 Allow underscore in dnslist lookups
2484 Fixes bug 1026. Patch from Graeme Fowler
2486 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2487 Code patches from Adam Ciarcinski of NetBSD.
2489 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2492 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2493 is logged, to avoid truncation. Patch from John Horne.
2495 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2496 Patch from Jakob Hirsch.
2498 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2499 SQL string expansion failure details.
2500 Patch from Andrey Oktyabrski.
2502 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2503 Patch from Simon Arlott.
2505 PP/09 New lookups functionality failed to compile on old gcc which rejects
2506 extern declarations in function scope.
2507 Patch from Oliver Fleischmann
2509 PP/10 Use sig_atomic_t for flags set from signal handlers.
2510 Check getgroups() return and improve debugging.
2511 Fixed developed for diagnosis in bug 927 (which turned out to be
2514 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2515 Patch from Mark Zealey.
2517 PP/12 Bugzilla 1056: Improved spamd server selection.
2518 Patch from Mark Zealey.
2520 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2521 Based on patch from Heiko Schlittermann.
2523 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2524 Patch from Uwe Doering, sign-off by Michael Haardt.
2526 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2527 variable declaration deep within a block. Bug and patch from
2530 PP/15 lookups-Makefile IRIX compatibility coercion.
2532 PP/16 Make DISABLE_DKIM build knob functional.
2534 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2535 Patch by Simon Arlott
2537 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2538 variable argument macros. Our copy now differs from upstream.
2544 TF/01 Failure to get a lock on a hints database can have serious
2545 consequences so log it to the panic log.
2547 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2548 controlled using the smtp_confirmation log selector.
2550 TF/03 Include the error message when we fail to unlink a spool file.
2552 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2553 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2554 for maintaining out-of-tree patches for some time.
2556 PP/01 Bugzilla 139: Documentation and portability issues.
2557 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2558 Handle per-OS dynamic-module compilation flags.
2560 PP/02 Let /dev/null have normal permissions.
2561 The 4.73 fixes were a little too stringent and complained about the
2562 permissions on /dev/null. Exempt it from some checks.
2563 Reported by Andreas M. Kirchwitz.
2565 PP/03 Report version information for many libraries, including
2566 Exim version information for dynamically loaded libraries. Created
2567 version.h, now support a version extension string for distributors
2568 who patch heavily. Dynamic module ABI change.
2570 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2571 privilege escalation vulnerability whereby the Exim run-time user
2572 can cause root to append content of the attacker's choosing to
2575 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2578 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2579 If dropping privileges for untrusted macros, we disabled normal logging
2580 on the basis that it would fail; for the Exim run-time user, this is not
2581 the case, and it resulted in successful deliveries going unlogged.
2582 Fixed. Reported by Andreas Metzler.
2588 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2589 only prepend for the Resent-* case. Fixes regression introduced in
2590 Exim 4.70 by NM/22 for Bugzilla 607.
2592 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2593 increasing numbers of administrators be bitten by this.
2595 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2597 PP/03 Bugzilla 994: added openssl_options main configuration option.
2599 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2601 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2603 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2605 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2606 without a peer certificate, leading to a segfault because of an
2607 assumption that peers always have certificates. Be a little more
2608 paranoid. Problem reported by Martin Tscholak.
2610 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2611 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2612 NB: ClamAV planning to remove STREAM in "middle of 2010".
2613 CL also introduces -bmalware, various -d+acl logging additions and
2614 more caution in buffer sizes.
2616 PP/09 Implemented reverse_ip expansion operator.
2618 PP/10 Bugzilla 937: provide a "debug" ACL control.
2620 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2622 PP/12 Bugzilla 973: Implement --version.
2624 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2626 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2628 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2630 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2631 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2632 ignore trailing whitespace.
2634 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2636 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2639 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2640 Notification from Dr Andrew Aitchison.
2642 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2643 ExtendedDetectionInfo response format.
2644 Notification from John Horne.
2646 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2649 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2650 XSL and documented dependency on system catalogs, with examples of how
2653 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2656 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2657 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2658 configuration file which is writeable by the Exim user or group.
2660 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2661 of configuration files to cover files specified with the -C option if
2662 they are going to be used with root privileges, not just the default
2665 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2666 option (effectively making it always true).
2668 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2669 files to be used while preserving root privileges.
2671 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2672 that rogue child processes cannot use them.
2674 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2675 run-time user, instead of root.
2677 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2678 Exim run-time user without dropping privileges.
2680 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2681 result string, instead of calling string_vformat() twice with the same
2684 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2685 for other users. Others should always drop root privileges if they use
2686 -C on the command line, even for a whitelisted configure file.
2688 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2690 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2696 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2697 $data_path, and $header_path variables; fixed documentation bugs and
2700 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2701 exipick to access non-standard spools, including the "frozen" queue
2704 NM/01 Bugzilla 965: Support mysql stored procedures.
2705 Patch from Alain Williams
2707 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2709 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2710 Patch from Andreas Metzler
2712 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2713 Patch from Kirill Miazine
2715 NM/05 Bugzilla 671: Added umask to procmail example.
2717 JJ/03 installed exipick 20100323.0, fixing doc bug
2719 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2720 directory. Notification and patch from Dan Rosenberg.
2722 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2724 TK/02 Improve log output when DKIM signing operation fails.
2726 MH/01 Treat the transport option dkim_domain as a colon separated
2727 list, not as a single string, and sign the message with each element,
2728 omitting multiple occurences of the same signer.
2730 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2731 Bugzilla 985, 986. Patch by Simon Arlott
2733 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2734 Patch by Simon Arlott
2736 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2737 MBX locking. Notification from Dan Rosenberg.
2743 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2745 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2747 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2749 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2751 NM/04 Bugzilla 915: Flag broken perl installation during build.
2757 TK/01 Added patch by Johannes Berg that expands the main option
2758 "spamd_address" if it starts with a dollar sign.
2760 TK/02 Write list of recipients to X-Envelope-Sender header when building
2761 the mbox-format spool file for content scanning (suggested by Jakob
2764 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2765 (http://www.dcc-servers.net/) support via dccifd. Activated by
2766 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2768 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2769 by Mark Daniel Reidel <mr@df.eu>.
2771 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2772 When building exim an external PCRE library is now needed -
2773 PCRE is a system library on the majority of modern systems.
2774 See entry on PCRE_LIBS in EDITME file.
2776 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2777 conversation. Added nologin parameter to request.
2778 Patch contributed by Kirill Miazine.
2780 TF/01 Do not log submission mode rewrites if they do not change the address.
2782 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2784 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2785 log files in place. Contributed by Roberto Lima.
2787 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2789 TF/03 Bugzilla 615: When checking the local_parts router precondition
2790 after a local_part_suffix or local_part_prefix option, Exim now
2791 does not use the address's named list lookup cache, since this
2792 contains cached lookups for the whole local part.
2794 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2795 Robert Millan. Documentation is in experimental-spec.txt.
2797 TF/04 Bugzilla 668: Fix parallel build (make -j).
2799 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2801 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2802 Patch provided by Jan Srzednicki.
2804 TF/05 Leading white space used to be stripped from $spam_report which
2805 wrecked the formatting. Now it is preserved.
2807 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2808 that they are available at delivery time.
2810 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2812 TF/08 TLS error reporting now respects the incoming_interface and
2813 incoming_port log selectors.
2815 TF/09 Produce a more useful error message if an SMTP transport's hosts
2816 setting expands to an empty string.
2818 NM/06 Bugzilla 744: EXPN did not work under TLS.
2819 Patch provided by Phil Pennock.
2821 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2822 Patch provided by Richard Godbee.
2824 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2825 acl_smtp_notquit, added index entry.
2827 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2828 Patch provided by Eugene Bujak.
2830 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2831 accept(). Patch provided by Maxim Dounin.
2833 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2834 Patch provided by Phil Pennock.
2836 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2838 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2839 Patch provided by Brad "anomie" Jorsch.
2841 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2842 Patch provided by Dean Brooks.
2844 TK/05 Add native DKIM support (does not depend on external libraries).
2846 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2847 Patch provided by Graeme Fowler.
2849 NM/16 Bugzilla 851: Documentation example syntax fix.
2851 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2853 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2856 NM/19 Bugzilla 745: TLS version reporting.
2857 Patch provided by Phil Pennock.
2859 NM/20 Bugzilla 167: bool: condition support.
2860 Patch provided by Phil Pennock.
2862 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2863 clients. Patch provided by Phil Pennock.
2865 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2866 Patch provided by Brad "anomie" Jorsch.
2868 NM/23 Bugzilla 687: Fix misparses in eximstats.
2869 Patch provided by Heiko Schlittermann.
2871 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2872 Patch provided by Heiko Schlittermann.
2874 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2875 plus update to original patch.
2877 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2879 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2880 Patch provided by David Brownlee.
2882 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2884 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2886 NM/30 Bugzilla 888: TLS documentation bugfixes.
2888 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2890 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2891 Unlike the original bugzilla I have changed all shell scripts in src tree.
2893 NM/33 Bugzilla 898: Transport filter timeout fix.
2894 Patch by Todd Rinaldo.
2896 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2897 Patch by Serge Demonchaux.
2899 NM/35 Bugzilla 39: Base64 decode bug fixes.
2900 Patch by Jakob Hirsch.
2902 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2904 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2906 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2908 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2914 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2915 ALT-N's libdkim that I have put here:
2916 http://duncanthrax.net/exim-experimental/
2918 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2919 were called 'true' and it seems that C99 defines that as a reserved
2920 keyword to be used with 'bool' variable types. That means you could
2921 not include C99-style headers which use bools without triggering
2922 build errors in sieve.c.
2924 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2925 as mailq or other aliases. Changed the --help handling significantly
2926 to do whats expected. exim_usage() emits usage/help information.
2928 SC/01 Added the -bylocaldomain option to eximstats.
2930 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2932 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2934 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2940 PH/01 Another patch from the Sieve maintainer.
2942 PH/02 When an IPv6 address is converted to a string for single-key lookup
2943 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2944 dots are used instead of colons so that keys in lsearch files need not
2945 contain colons. This was done some time before quoting was made available
2946 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2947 (notated using the quote facility) so as to distinguish them from IPv4
2948 keys. This meant that lookups for IP addresses in host lists did not work
2949 for iplsearch lookups.
2951 This has been fixed by arranging for IPv6 addresses to be expressed with
2952 colons if the lookup type is iplsearch. This is not incompatible, because
2953 previously such lookups could never work.
2955 The situation is now rather anomalous, since one *can* have colons in
2956 ordinary lsearch keys. However, making the change in all cases is
2957 incompatible and would probably break a number of configurations.
2959 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2962 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2963 conversion specification without a maximum field width, thereby enabling
2964 a rogue spamd server to cause a buffer overflow. While nobody in their
2965 right mind would setup Exim to query an untrusted spamd server, an
2966 attacker that gains access to a server running spamd could potentially
2967 exploit this vulnerability to run arbitrary code as the Exim user.
2969 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2970 $primary_hostname instead of what libspf2 thinks the hosts name is.
2972 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2973 a directory entry by the name of the lookup key. Previously, if a
2974 symlink pointed to a non-existing file or a file in a directory that
2975 Exim lacked permissions to read, a lookup for a key matching that
2976 symlink would fail. Now it is enough that a matching directory entry
2977 exists, symlink or not. (Bugzilla 503.)
2979 PH/03 The body_linecount and body_zerocount variables are now exported in the
2982 PH/04 Added the $dnslist_matched variable.
2984 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2985 This means they are set thereafter only if the connection becomes
2988 PH/06 Added the client_condition to authenticators so that some can be skipped
2989 by clients under certain conditions.
2991 PH/07 The error message for a badly-placed control=no_multiline_responses left
2992 "_responses" off the end of the name.
2994 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2996 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2997 (without spaces) instead of just copying the configuration text.
2999 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3001 PH/11 Added $max_received_linelength.
3003 PH/12 Added +ignore_defer and +include_defer to host lists.
3005 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3006 way in which PCRE > 7.0 is built.
3008 PH/14 Implemented queue_only_load_latch.
3010 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3011 MAIL command. The effect was to mangle the value on 64-bit systems.
3013 PH/16 Another patch from the Sieve maintainer.
3015 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3017 PH/18 If a system quota error occurred while trying to create the file for
3018 a maildir delivery, the message "Mailbox is full" was not appended to the
3019 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3020 only to a quota excession during the actual writing of the file.
3022 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3023 characters?) which causes problems in log lines. The DN values are now
3024 passed through string_printing() before being added to log lines.
3026 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3027 and InterBase are left for another time.)
3029 PH/21 Added message_body_newlines option.
3031 PH/22 Guard against possible overflow in moan_check_errorcopy().
3033 PH/23 POSIX allows open() to be a macro; guard against that.
3035 PH/24 If the recipient of an error message contained an @ in the local part
3036 (suitably quoted, of course), incorrect values were put in $domain and
3037 $local_part during the evaluation of errors_copy.
3043 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3044 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3047 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3048 issue a MAIL command.
3050 PH/02 In an ACL statement such as
3052 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3054 if a client was not listed at all, or was listed with a value other than
3055 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3056 the condition was not true (as it should be), so access was not denied.
3057 The bug was that the ! inversion was incorrectly passed on to the second
3058 item. This has been fixed.
3060 PH/03 Added additional dnslists conditions == and =& which are different from
3061 = and & when the dns lookup returns more than one IP address.
3063 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3064 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3066 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3067 FSYNC, which compiles an option called disable_fsync that allows for
3068 bypassing fsync(). The documentation is heavily laced with warnings.
3070 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3072 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3073 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3074 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3075 including adding "make clean"; (3) Added -fPIC when compiling the test
3076 dynamically loaded module, to get rid of a warning.
3078 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3079 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3080 = 0s. The bug is otherwise harmless.
3082 PH/07 There was a bug in the dovecot authenticator such that the value of
3083 $auth1 could be overwritten, and so not correctly preserved, after a
3084 successful authentication. This usually meant that the value preserved by
3085 the server_setid option was incorrect.
3087 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3089 PH/09 Installed PCRE release 7.0.
3091 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3092 run for batched SMTP input. It is now run at the start of every message
3093 in the batch. While fixing this I discovered that the process information
3094 (output by running exiwhat) was not always getting set for -bs and -bS
3095 input. This is fixed, and it now also says "batched" for BSMTP.
3097 PH/11 Added control=no_pipelining.
3099 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3100 patch, slightly modified), and move the expansion of helo_data till after
3101 the connection is made in the smtp transport (so it can use these
3104 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3106 PH/14 Added log_selector = +pid.
3108 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3110 PH/16 Add ${if forany and ${if forall.
3112 PH/17 Added dsn_from option to vary the From: line in DSNs.
3114 PH/18 Flush SMTP output before performing a callout, unless control =
3115 no_callout_flush is set.
3117 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3118 was true (the default) a successful delivery failed to delete the retry
3119 item, thus causing premature timeout of the address. The bug is now
3122 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3124 PH/21 Long custom messages for fakedefer and fakereject are now split up
3125 into multiline responses in the same way that messages for "deny" and
3126 other ACL rejections are.
3128 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3129 with slight modification.
3131 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3132 draft, changing the syntax and factoring some duplicate code".
3134 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3135 for deliveries of the second and subsequent messages over the same SMTP
3138 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3139 ${reduce, with only minor "tidies".
3141 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3143 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3144 expansion side effects.
3146 PH/27 When a message times out after an over-quota error from an Exim-imposed
3147 quota, the bounce message says "mailbox is full". This message was not
3148 being given when it was a system quota that was exceeded. It now should
3151 MH/03 Made $recipients available in local_scan(). local_scan() already has
3152 better access to the recipient list through recipients_list[], but
3153 $recipients can be useful in postmaster-provided expansion strings.
3155 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3156 in the case of a MAIL command with additional options following the
3157 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3158 were accidentally chopped off.
3160 PH/29 SMTP synchronization checks are implemented when a command is read -
3161 there is a check that no more input is waiting when there shouldn't be
3162 any. However, for some commands, a delay in an ACL can mean that it is
3163 some time before the response is written. In this time, more input might
3164 arrive, invalidly. So now there are extra checks after an ACL has run for
3165 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3166 pipelining has not been advertised.
3168 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3170 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3171 returned in the SMTP response when smtp_return_error_details was false.
3172 This has been fixed.
3174 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3175 instead of the C I/O that was originally supplied, because problems were
3176 reported on Solaris.
3178 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3179 Exim which did not show up earlier: it was assuming that a call to
3180 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3181 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3182 was a macro that became an assignment, so it seemed to work. This has
3183 changed to a proper function call with a void return, hence the compile
3184 error. Exim's code has been fixed.
3186 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3189 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3191 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3193 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3194 because it turns out to be too big or there is a timeout) there is no
3195 "Completed" line in the log. When some messages of this type were
3196 selected by exigrep, they were listed as "not completed". Others were
3197 picked up by some special patterns. I have improved the selection
3198 criteria to be more general.
3200 PH/38 The host_find_failed option in the manualroute router can now be set
3201 to "ignore", to completely ignore a host whose IP address cannot be
3202 found. If all hosts are ignored, the behaviour is controlled by the new
3203 host_all_ignored option.
3205 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3206 homing or because of multiple MX records with /mx) generated more than
3207 one IP address, and the following item turned out to be the local host,
3208 all the secondary addresses of the first item were incorrectly removed
3209 from the list, along with the local host and any following hosts (which
3210 is what is supposed to happen).
3212 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3213 whoever called Exim into the -H file. In the case of the daemon it was
3214 behaving confusingly. When first started, it used values for whoever
3215 started the daemon, but after a SIGHUP it used the Exim user (because it
3216 calls itself on a restart). I have changed the code so that it now always
3219 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3220 message are rejected with the same error (e.g. no authentication or bad
3221 sender address), and a DATA command is nevertheless sent (as can happen
3222 with PIPELINING or a stupid MUA), the error message that was given to the
3223 RCPT commands is included in the rejection of the DATA command. This is
3224 intended to be helpful for MUAs that show only the final error to their
3227 PH/42 Another patch from the Sieve maintainer.
3229 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3230 Eximstats - Fixed some broken HTML links and added missing column headers
3232 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3233 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3235 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3237 PH/43 Yet another patch from the Sieve maintainer.
3239 PH/44 I found a way to check for a TCP/IP connection going away before sending
3240 the response to the final '.' that terminates a message, but only in the
3241 case where the client has not sent further data following the '.'
3242 (unfortunately, this is allowed). However, in many cases there won't be
3243 any further data because there won't be any more messages to send. A call
3244 to select() can be used: if it shows that the input is "ready", there is
3245 either input waiting, or the socket has been closed. An attempt to read
3246 the next input character can distinguish the two cases. Previously, Exim
3247 would have sent an OK response which the client would never have see.
3248 This could lead to message repetition. This fix should cure that, at
3249 least in a lot of common cases.
3251 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3252 advertised in response to EHLO.
3258 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3259 fixed by 4.65/MH/01 (is this a record?) are fixed:
3261 (i) An empty string was always treated as zero by the numeric comparison
3262 operators. This behaviour has been restored.
3264 (ii) It is documented that the numeric comparison operators always treat
3265 their arguments as decimal numbers. This was broken in that numbers
3266 starting with 0 were being interpreted as octal.
3268 While fixing these problems I realized that there was another issue that
3269 hadn't been noticed. Values of message_size_limit (both the global option
3270 and the transport option) were treated as octal if they started with 0.
3271 The documentation was vague. These values are now always treated as
3272 decimal, and I will make that clear in the documentation.
3278 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3279 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3282 MH/01 Don't check that the operands of numeric comparison operators are
3283 integers when their expansion is in "skipping" mode (fixes bug
3284 introduced by 4.64-PH/07).
3286 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3287 child addresses, Exim now panics and dies. Previously, because the count
3288 is held in a short int, deliveries were likely to be lost. As such a
3289 large number of recipients for a single message is ridiculous
3290 (performance will be very, very poor), I have chosen to impose a limit
3291 rather than extend the field.
3297 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3298 leftover -K file (the existence of which was triggered by #402).
3299 While we were at it, introduced process PID as part of the -K
3300 filename. This should rule out race conditions when creating
3303 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3304 processing considerably. Previous code took too long for large mails,
3305 triggering a timeout which in turn triggers #401.
3307 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3308 in the DK code in transports.c. sendfile() is not really portable,
3309 hence the _LINUX specificness.
3311 TF/01 In the add_headers option to the mail command in an Exim filter,
3312 there was a bug that Exim would claim a syntax error in any
3313 header after the first one which had an odd number of characters
3316 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3317 callout verification, Exim cached a "reject" for the entire domain. This
3318 is correct for most verifications, but it is not correct for a recipient
3319 verification with use_sender or use_postmaster set, because in that case
3320 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3321 case of MAIL FROM:<> rejection from other early rejections (e.g.
3322 rejection of HELO). When verifying a recipient using a non-null MAIL
3323 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3324 Whatever the result of the callout, the value of the domain cache is
3325 left unchanged (for any other kind of callout, getting as far as trying
3326 RCPT means that the domain itself is ok).
3328 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3331 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3332 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3333 session, the daemon ignores ECONNECT errors and logs others; it now
3334 ignores EPIPE as well.
3336 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3337 (quoted-printable decoding).
3339 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3340 later the small subsequent patch to fix an introduced bug.
3342 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3344 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3346 PH/08 An error is now given if message_size_limit is specified negative.
3348 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3349 to be given (somewhat) arbitrary names.
3351 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3354 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3355 miscellaneous code fixes
3357 PH/10 Added the log_reject_target ACL modifier to specify where to log
3360 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3361 hostname. This is wrong, because it relates to the incoming message (and
3362 probably the interface on which it is arriving) and not to the outgoing
3363 callout (which could be using a different interface). This has been
3364 changed to use the value of the helo_data option from the smtp transport
3365 instead - this is what is used when a message is actually being sent. If
3366 there is no remote transport (possible with a router that sets up host
3367 addresses), $smtp_active_hostname is used.
3369 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3370 tweaks were necessary in order to get it to work (see also 21 below):
3371 (a) The code assumed that strncpy() returns a negative number on buffer
3372 overflow, which isn't the case. Replaced with Exim's string_format()
3374 (b) There were several signed/unsigned issues. I just did the minimum
3375 hacking in of casts. There is scope for a larger refactoring.
3376 (c) The code used strcasecmp() which is not a standard C function.
3377 Replaced with Exim's strcmpic() function.
3378 (d) The code set only $1; it now sets $auth1 as well.
3379 (e) A simple test gave the error "authentication client didn't specify
3380 service in request". It would seem that Dovecot has changed its
3381 interface. Fortunately there's a specification; I followed it and
3382 changed what the client sends and it appears to be working now.
3384 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3387 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3388 address A is aliased to B and C, where B exists and C does not. Without
3389 -v the output is "A verified" because verification stops after a
3390 successful redirection if more than one address is generated. However,
3391 with -v the child addresses are also verified. Exim was outputting "A
3392 failed to verify" and then showing the successful verification for C,
3393 with its parentage. It now outputs "B failed to verify", showing B's
3394 parentage before showing the successful verification of C.
3396 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3397 look up a TXT record in a specific list after matching in a combined
3400 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3401 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3402 they consult the DNS. I had assumed they would set it the way they
3403 wanted; and indeed my experiments on Linux seem to show that in some
3404 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3405 To be on the safe side, however, I have now made the interface to
3406 host_find_byname() similar to host_find_bydns(), with an argument
3407 containing the DNS resolver options. The host_find_byname() function now
3408 sets these options at its start, just as host_find_bydns() does. The smtp
3409 transport options dns_qualify_single and dns_search_parents are passed to
3410 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3411 of host_find_byname() use the default settings of RES_DEFNAMES
3412 (qualify_single) but not RES_DNSRCH (search_parents).
3414 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3415 spool_read_header() do less string testing, by means of a preliminary
3416 switch on the second character of optional "-foo" lines. (This is
3417 overdue, caused by the large number of possibilities that now exist.
3418 Originally there were few.) While I was there, I also converted the
3419 str(n)cmp tests so they don't re-test the leading "-" and the first
3420 character, in the hope this might squeeze out yet more improvement.
3422 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3423 flag allowing group syntax was set by the header_syntax check but not
3424 turned off, possible causing trouble later; (2) The flag was not being
3425 set at all for the header_verify test, causing "group"-style headers to
3426 be rejected. I have now set it in this case, and also caused header_
3427 verify to ignore an empty address taken from a group. While doing this, I
3428 came across some other cases where the code for allowing group syntax
3429 while scanning a header line wasn't quite right (mostly, not resetting
3430 the flag correctly in the right place). These bugs could have caused
3431 trouble for malformed header lines. I hope it is now all correct.
3433 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3434 with the "reply" argument non-NULL. The code, however (which originally
3435 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3436 but it didn't always do it. This confused somebody who was copying the
3437 code for some other use. I have removed all the tests.
3439 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3440 feature that was used to support insecure browsers during the U.S. crypto
3441 embargo. It requires special client support, and Exim is probably the
3442 only MTA that supported it -- and would never use it because real RSA is
3443 always available. This code has been removed, because it had the bad
3444 effect of slowing Exim down by computing (never used) parameters for the
3445 RSA_EXPORT functionality.
3447 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3448 authenticator to fail if there's a tab character in the incoming data
3449 (there should never be unless someone is messing about, as it's supposed
3450 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3451 if the connection is using TLS or if the remote IP is the same as the
3452 local IP, and the "valid-client-cert option" if a client certificate has
3455 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3456 authenticators. This can be used for authorization after authentication
3457 succeeds. (In the case of plaintext, it servers for both authentication
3460 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3461 if any retry times were supplied.
3463 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3464 connection, where there is no client IP address to check. In this
3465 situation, the verify now always succeeds.
3467 PH/25 Applied John Jetmore's -Mset patch.
3469 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3471 PH/27 In a string expansion for a processed (not raw) header when multiple
3472 headers of the same name were present, leading whitespace was being
3473 removed from all of them, but trailing whitespace was being removed only
3474 from the last one. Now trailing whitespace is removed from each header
3475 before concatenation. Completely empty headers in a concatenation (as
3476 before) are ignored.
3478 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3479 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3481 PH/29 [Removed. This was a change that I later backed out, and forgot to
3482 correct the ChangeLog entry (that I had efficiently created) before
3483 committing the later change.]
3485 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3486 address errors (4xx response to RCPT) over the same connection as other
3487 messages routed to the same hosts. Such deliveries are always "forced",
3488 so retry times are not inspected. This resulted in far too many retries
3489 for the affected addresses. The effect occurred only when there were more
3490 hosts than the hosts_max_try setting in the smtp transport when it had
3491 the 4xx errors. Those hosts that it had tried were not added to the list
3492 of hosts for which the message was waiting, so if all were tried, there
3493 was no problem. Two fixes have been applied:
3495 (i) If there are any address or message errors in an SMTP delivery, none
3496 of the hosts (tried or untried) are now added to the list of hosts
3497 for which the message is waiting, so the message should not be a
3498 candidate for sending over the same connection that was used for a
3499 successful delivery of some other message. This seems entirely
3500 reasonable: after all the message is NOT "waiting for some host".
3501 This is so "obvious" that I'm not sure why it wasn't done
3502 previously. Hope I haven't missed anything, but it can't do any
3503 harm, as the worst effect is to miss an optimization.
3505 (ii) If, despite (i), such a delivery is accidentally attempted, the
3506 routing retry time is respected, so at least it doesn't keep
3507 hammering the server.
3509 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3510 in ${readsocket because some servers need this prod.
3512 PH/32 Added some extra debug output when updating a wait-xxx database.
3514 PH/33 The hint "could be header name not terminated by colon", which has been
3515 given for certain expansion errors for a long time, was not being given
3516 for the ${if def:h_colon_omitted{... case.
3518 PH/34 The spec says: "With one important exception, whenever a domain list is
3519 being scanned, $domain contains the subject domain." There was at least
3520 one case where this was not true.
3522 PH/35 The error "getsockname() failed: connection reset by peer" was being
3523 written to the panic log as well as the main log, but it isn't really
3524 panic-worthy as it just means the connection died rather early on. I have
3525 removed the panic log writing for the ECONNRESET error when getsockname()
3528 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3529 runs only) independently of the message's sender address. This meant
3530 that, if the 4xx error was in fact related to the sender, a different
3531 message to the same recipient with a different sender could confuse
3532 things. In particular, this can happen when sending to a greylisting
3533 server, but other circumstances could also provoke similar problems.
3534 I have changed the default so that the retry time for these errors is now
3535 based a combination of the sender and recipient addresses. This change
3536 can be overridden by setting address_retry_include_sender=false in the
3539 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3540 remote server are returned as part of bounce messages. This was not
3541 happening for LMTP over a pipe (the lmtp transport), but now it is the
3542 same for both kinds of LMTP.
3544 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3545 in header lines that were in fact CNAMEs. This is no longer the case.
3547 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3548 and queue runs started by the daemon processed all messages. This has
3549 been fixed so that -R and -S can now usefully be given with -q<time>.
3551 PH/40 Import PCRE release 6.7 (fixes some bugs).
3553 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3555 PH/42 Give an error if -q is specified more than once.
3557 PH/43 Renamed the variables $interface_address and $interface_port as
3558 $received_ip_address and $received_port, to make it clear that these
3559 values apply to message reception, and not to the outgoing interface when
3560 a message is delivered. (The old names remain recognized, of course.)
3562 PH/44 There was no timeout on the connect() call when using a Unix domain
3563 socket in the ${readsocket expansion. There now is.
3565 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3566 be meaningful with "accept".
3568 SC/01 Eximstats V1.43
3569 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3571 SC/02 Eximstats V1.44
3572 Use a glob alias rather than an array ref in the generated
3573 parser. This improves both readability and performance.
3575 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3576 Collect SpamAssassin and rejection statistics.
3577 Don't display local sender or destination tables unless
3578 there is data to show.
3579 Added average volumes into the top table text output.
3581 SC/04 Eximstats V1.46
3582 Collect data on the number of addresses (recipients)
3583 as well as the number of messages.
3585 SC/05 Eximstats V1.47
3586 Added 'Message too big' to the list of mail rejection
3587 reasons (thanks to Marco Gaiarin).
3589 SC/06 Eximstats V1.48
3590 Mainlog lines which have GMT offsets and are too short to
3591 have a flag are now skipped.
3593 SC/07 Eximstats V1.49 (Alain Williams)
3594 Added the -emptyok flag.
3596 SC/08 Eximstats V1.50
3597 Fixes for obtaining the IP address from reject messages.
3599 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3600 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3601 whitespace changes from 4.64-PH/27
3603 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3606 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3607 are found, allow negative numbers in numeric criteria)
3609 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3611 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3612 to match changes made in 4.64-PH/43
3614 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3616 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3617 "domain" is actually an IP address, and force a failure. This locks out
3618 those revolvers/nameservers that support "A-for-A" lookups, in
3619 contravention of the specifications.
3621 PH/48 When a host name was looked up from an IP address, and the subsequent
3622 forward lookup of the name timed out, the host name was left in
3623 $sender_host_name, contrary to the specification.
3625 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3626 restricted to single-key lookups, Exim was not diagnosing an error if
3627 * or *@ was used with a query-style lookup.
3629 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3631 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3632 long ago, but noone interested enough thought of it. Let's just say that
3633 the "1.1" means that there are some new functions that weren't there at
3634 some point in the past.
3636 PH/51 Error processing for expansion failure of helo_data from an smtp
3637 transport during callout processing was broken.
3639 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3640 tested/used via the -bh/-bhc/-bs options.
3642 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3643 bug, fixed in subsequent PCRE releases).
3645 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3646 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3648 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3654 SC/01 Use a glob alias rather than an array ref in eximstats generated
3655 parser. This improves both readability and performance.
3657 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3658 Don't display local sender or destination tables in eximstats unless
3659 there is data to show.
3660 Added average volumes into the eximstats top table text output.
3662 SC/03 Collect data on the number of addresses (recipients) as well
3663 as the number of messages in eximstats.
3665 TF/01 Correct an error in the documentation for the redirect router. Exim
3666 does (usually) call initgroups() when daemonizing.
3668 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3669 with consistent privilege compared to when running as a daemon.
3671 TF/03 Note in the spec that $authenticated_id is not set for local
3672 submissions from trusted users.
3674 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3675 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3677 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3678 by adding some example configuration directives to the default
3679 configuration file. A little bit of work is required to uncomment the
3680 directives and define how usernames and passwords are checked, but
3681 there is now a framework to start from.
3683 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3684 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3685 without this. I don't know how relevant this is to other LDAP libraries.
3687 PH/02 Add the verb name to the "unknown ACL verb" error.
3689 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3691 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3693 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3694 directory not expanded when it should be if an expanded home directory
3695 was set for the address (which is overridden by the transport).
3697 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3700 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3701 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3702 because it is too late at that time, and has no effect.
3704 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3705 security issue with \' (bugzilla #107). I could not use the
3706 PQescapeStringConn() function, because it needs a PGconn value as one of
3709 PH/08 When testing addresses using -bt, indicate those final addresses that
3710 are duplicates that would not cause an additional delivery. At least one
3711 person was confused, thinking that -bt output corresponded to deliveries.
3712 (Suppressing duplicates isn't a good idea as you lose the information
3713 about possibly different redirections that led to the duplicates.)
3715 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3716 systems where poll() doesn't work, in particular OS X.
3718 PH/10 Added more information to debugging output for retry time not reached.
3720 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3721 operations in malware.c.
3723 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3726 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3727 syslog_duplication set false, log lines that would normally be written
3728 both the the main log and to the reject log were not written to syslog at
3731 PH/14 In the default configuration, change the use of "message" in ACL warn
3732 statements to "add_header".
3734 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3735 not followed by a command (e.g. "seen endif").
3737 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3738 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3741 PH/17 Added extra conditions to the default value of delay_warning_condition
3745 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3746 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3747 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3750 The Auto-Submitted: and various List- headers are standardised, whereas I
3751 don't think Precedence: ever was.
3753 PH/18 Refactored debugging code in route_finduser() to show more information,
3754 in particular, the error code if getpwnam() issues one.
3756 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3757 This is apparently needed in addition to the PH/07 change above to avoid
3758 any possible encoding problems.
3760 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3761 but not after initializing Perl.
3763 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3764 output them only if debugging. By default they are written stderr,
3765 apparently, which is not desirable.
3767 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3770 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3773 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3775 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3776 authenticated or an ident call has been made. Suppress the default
3777 values for $authenticated_id and $authenticated_sender (but permit -oMai
3778 and -oMas) when testing with -bh.
3780 PH/24 Re-jigged the order of the tests in the default configuration so that the
3781 tests for valid domains and recipients precede the DNS black list and CSA
3782 tests, on the grounds that those ones are more expensive.
3784 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3785 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3786 This bug exists in every version of Exim that I still have, right back to
3789 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3790 However, an attempt to turn on case-sensitivity in a regex key by
3791 including (?-i) didn't work because the subject string was already
3792 lowercased, and the effects were non-intuitive. It turns out that a
3793 one-line patch can be used to allow (?-i) to work as expected.
3799 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3800 other effects) broke the use of negated acl sub-conditions.
3802 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3805 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3806 "Deny" causes Exim to reject the incoming connection with a 554 error.
3807 Unfortunately, if there is a major crisis, such as a disk failure,
3808 tcp-wrappers gives "deny", whereas what one would like would be some
3809 kind of temporary error. A kludge has been added to help with this.
3810 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3811 554 error is used if errno is still zero or contains ENOENT (which occurs
3812 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3815 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3817 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3818 errors. Otherwise a message that provokes a temporary error (when other
3819 messages do not) can cause a whole host to time out.
3821 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3822 addresses were routed directly to files or pipes from a redirect router.
3823 File deliveries just didn't batch; pipe deliveries might have suffered
3826 PH/06 A failure to get a lock for a hints database would erroneously always say
3827 "Failed to get write lock", even when it was really a read lock.
3829 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3830 of 0600. This has been changed to use the value of the lockfile_mode
3831 option (which defaults to 0600).
3833 PH/08 Applied small patch from the Sieve maintainer.
3835 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3836 folder from quota calculations, a direct delivery into this folder messed
3837 up the contents of the maildirsize file. This was because the regex was
3838 used only to exclude .Trash (or whatever) when the size of the mailbox
3839 was calculated. There was no check that a delivery was happening into an
3840 excluded directory. This bug has been fixed by ignoring all quota
3841 processing for deliveries into excluded directories.
3843 PH/10 Added the maildirfolder_create_regex option to appendfile.
3849 PH/01 The code for finding all the local interface addresses on a FreeBSD
3850 system running IPv6 was broken. This may well have applied to all BSD
3851 systems, as well as to others that have similar system calls. The broken
3852 code found IPv4 interfaces correctly, but gave incorrect values for the
3853 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3854 that it would not match correctly against @[] and not recognize the IPv6
3857 PH/02 The ipliteral router was not recognizing addresses of the form user@
3858 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3860 PH/03 Added disable_ipv6.
3862 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3863 decoded form, because it is most often used to construct To: headers
3864 lines in autoreplies, and the decoded form may well be syntactically
3865 invalid. However, $reply_address has leading white space removed, and all
3866 newlines turned into spaces so that the autoreply transport does not
3869 PH/05 If group was specified without a user on a router, and no group or user
3870 was specified on a transport, the group from the router was ignored.
3872 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3873 for visible compile-time settings that can be used to change these
3874 numbers, for those that want even more. Backwards compatibility with old
3875 spool files has been maintained. However, going back to a previous Exim
3876 release will lost any variables that are in spool files.
3878 PH/07 Two small changes when running in the test harness: increase delay when
3879 passing a TCP/IP connection to a new process, in case the original
3880 process has to generate a bounce, and remove special handling of
3881 127.0.0.2 (sic), which is no longer necessary.
3883 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3884 be the same on different OS.
3886 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3889 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3890 whether --show-vars was specified or not
3892 JJ/02 exipick: Added support for new ACL variable spool format introduced
3895 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3896 syntactically invalid From: or Reply-to: line, and a filter used this to
3897 generate an autoreply, and therefore failed to obtain an address for the
3898 autoreply, Exim could try to deliver to a non-existent relative file
3899 name, causing unrelated and misleading errors. What now happens is that
3900 it logs this as a hard delivery error, but does not attempt to create a
3903 PH/11 The exinext utility has a -C option for testing purposes, but although
3904 the given file was scanned by exinext itself; it wasn't being passed on
3905 when Exim was called.
3907 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3908 an end-of-file indication when reading a command response.
3910 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3911 compiled. In many other places in Exim, IPv6 addresses are always
3912 recognized, so I have changed this. It also means that IPv4 domain
3913 literals of the form [IPV4:n.n.n.n] are now always recognized.
3915 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3916 used if the router is not running as root, for example, when verifying at
3917 ACL time, or when using -bh. The debugging output from this situation was
3918 non-existent - all you got was a failure to exec. I have made two
3921 (a) Failures to set uid/gid, the current directory, or a process leader
3922 in a subprocess such as that created by queryprogram now generate
3923 suitable debugging output when -d is set.
3925 (b) The queryprogram router detects when it is not running as root,
3926 outputs suitable debugging information if -d is set, and then runs
3927 the subprocess without attempting to change uid/gid.
3929 PH/15 Minor change to Makefile for building test_host (undocumented testing
3932 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3933 additional section of a DNS packet that returns MX or SRV records.
3934 Instead, it always explicitly searches for A/AAAA records. This avoids
3935 major problems that occur when a DNS server includes only records of one
3936 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3937 fixed another bug: if SRV records were looked up and the corresponding
3938 address records were *not* found in the additional section, the port
3939 values from the SRV records were lost.
3941 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3942 using the correct key (the original address) when searching the retry
3943 rules in order to find which one to use for generating the retry hint.
3945 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3946 adding the default one. Similarly, if it contains a Reply-To: header, the
3947 errors_reply_to option, if set, is not used.
3949 PH/19 When calculating a retry time, Exim used to measure the "time since
3950 failure" by looking at the "first failed" field in the retry record. Now
3951 it does not use this if it is later than than the arrival time of the
3952 message. Instead it uses the arrival time. This makes for better
3953 behaviour in cases where some deliveries succeed, thus re-setting the
3954 "first failed" field. An example is a quota failure for a huge message
3955 when small messages continue to be delivered. Without this change, the
3956 "time since failure" will always be short, possible causing more frequent
3957 delivery attempts for the huge message than are intended.
3958 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3960 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3961 $1, $2, $3) because the numerical variables can be reset during some
3962 expansion items (e.g. "match"), thereby losing the authentication data.
3964 PH/21 Make -bV show the size of off_t variables so that the test suite can
3965 decide whether to run tests for quotas > 2G.
3967 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3968 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3969 filecount value is greater than 2G or if a quota value is greater than 2G
3970 on a system where the size of off_t is not greater than 4, a panic error
3973 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3974 never match. The debug and -bh output now contains an explicit error
3975 message indicating a malformed IPv4 address or mask.
3977 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3978 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3979 PH/23 above applies.
3981 PH/25 Do not write to syslog when running in the test harness. The only
3982 occasion when this arises is a failure to open the main or panic logs
3983 (for which there is an explicit test).
3985 PH/26 Added the /no_tell option to "control=freeze".
3987 PH/27 If a host name lookup failed very early in a connection, for example, if
3988 the IP address matched host_lookup and the reverse lookup yielded a name
3989 that did not have a forward lookup, an error message of the form "no IP
3990 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3991 could be logged. Now it outputs the IP address instead of "NULL".
3993 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3994 allows the sender and the authenticated sender to be set when
3995 submitting a message from within Exim. Since child_open_exim() is
3996 documented for local_scan(), the new function should be too.
3998 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3999 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4000 results in an empty string is now treated as unset.
4002 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4004 PH/31 Added sender_verify_fail logging option.
4006 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4007 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4008 tidied the source and removed it altogether.
4010 PH/33 When a queue run was abandoned because the load average was too high, a
4011 log line was always written; now it is written only if the queue_run log
4012 selector is set. In addition, the log line for abandonment now contains
4013 information about the queue run such as the pid. This is always present
4014 in "start" and "stop" lines but was omitted from the "abandon" line.
4016 PH/34 Omit spaces between a header name and the colon in the error message that
4017 is given when verify = headers_syntax fails (if there are lots of them,
4018 the message gets confusing).
4020 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4021 names, as there are now some PTR records that contain slashes. This check
4022 is only to protect against broken name servers that fall over on strange
4023 characters, so the fact that it applies to all lookups doesn't matter.
4025 PH/36 Now that the new test suite is complete, we can remove some of the
4026 special code in Exim that was needed for the old test suite. For example,
4027 sorting DNS records because real resolvers return them in an arbitrary
4028 order. The new test suite's fake resolver always returns records in the
4031 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4032 bounces) except when queue_only is set, to avoid logging races between
4033 the different processes.
4035 PH/38 Panic-die if .include specifies a non-absolute path.
4037 PH/39 A tweak to the "H" retry rule from its user.
4039 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4040 a label. They prevented compilation on older perls.
4042 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4043 a warning to be raised on newish perls.
4045 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4046 on queue. Changes to match documented behaviour of showing count of
4047 messages matching specified criteria.
4049 PH/40 Changed the default ident timeout from 30s to 5s.
4051 PH/41 Added support for the use of login_cap features, on those BSD systems
4052 that have them, for controlling the resources used by pipe deliveries.
4054 PH/42 The content-scanning code uses fopen() to create files in which to put
4055 message data. Previously it was not paying any attention to the mode of
4056 the files. Exim runs with umask(0) because the rest of the code creates
4057 files with open(), and sets the required mode explicitly. Thus, these
4058 files were ending up world-writeable. This was not a big issue, because,
4059 being within the spool directory, they were not world-accessible. I have
4060 created a function called modefopen, which takes an additional mode
4061 argument. It sets umask(777), creates the file, chmods it to the required
4062 mode, then resets the umask. All the relevant calls to fopen() in the
4063 content scanning code have been changed to use this function.
4065 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4066 to 24 hours. This avoids potential overflow problems when processing G
4067 and H retry rules. I suspect nobody ever tinkers with this value.
4069 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4071 PH/45 When the plaintext authenticator is running as a client, the server's
4072 challenges are checked to ensure they are valid base64 strings. By
4073 default, the authentication attempt is cancelled if an invalid string is
4074 received. Setting client_ignore_invalid_base64 true ignores these errors.
4075 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4076 they are received. Thus, the responses can be made to depend on the
4077 challenges. If an invalid string is ignored, an empty string is placed in
4080 PH/46 Messages that are created by the autoreply transport now contains a
4081 References: header, in accordance with RFCs 2822 and 3834.
4083 PH/47 Added authenticated_sender_force to the smtp transport.
4085 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4087 PH/49 Installed latest patch from the Sieve maintainer.
4089 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4090 was also set, the appendfile transport was unnecessarily scanning a
4091 directory of message files (e.g. for maildir delivery) to find the count
4092 of files (along with the size), even though it did not need this
4093 information. It now does the scan only if it needs to find either the
4094 size of the count of files.
4096 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4098 PH/52 Two bugs concerned with error handling when the smtp transport is
4101 (i) Exim was not creating retry information for temporary errors given
4102 for individual recipients after the DATA command when the smtp transport
4103 was used in LMTP mode. This meant that they could be retried too
4104 frequently, and not timed out correctly.
4106 (ii) Exim was setting the flag that allows error details to be returned
4107 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4108 recipients that were returned after the DATA command.
4110 PH/53 This is related to PH/52, but is more general: for any failing address,
4111 when detailed error information was permitted to be returned to the
4112 sender, but the error was temporary, then after the final timeout, only
4113 "retry timeout exceeded" was returned. Now it returns the full error as
4114 well as "retry timeout exceeded".
4116 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4117 do this, and (what is worse) MTAs that accept it.
4119 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4120 will now be deprecated.
4122 PH/56 New os.c-cygwin from the Cygwin maintainer.
4124 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4125 formats (previously only available in exim formats via -bpr, -bpru,
4126 and -bpra. Now also available in native and exiqgrep formats)
4128 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4129 with very large, slow to parse queues
4131 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4133 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4135 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4136 responses to authentication challenges, though it was showing the
4137 challenges; (ii) I've removed the CR characters from the debug output for
4140 PH/58 Allow for the insertion of a newline as well as a space when a string
4141 is turned into more than one encoded-word during RFC 2047 encoding. The
4142 Sieve code now uses this.
4144 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4145 data_4xx, lost_connection, tls_required.
4147 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4148 message was being sent as an SMTP response.
4150 PH/61 Add -l and -k options to exicyclog.
4152 PH/62 When verifying, if an address was redirected to one new address, so that
4153 verification continued, and the new address failed or deferred after
4154 having set something in $address_data, the value of $address_data was not
4155 passed back to the ACL. This was different to the case when no
4156 redirection occurred. The value is now passed back in both cases.
4158 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4159 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4160 use login_cap.h, so on its own it isn't the distinguishing feature. The
4161 new name refers directly to the setclassresources() function.
4163 PH/65 Added configuration files for NetBSD3.
4165 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4167 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4168 is preferred over IPv4.
4170 PH/68 The bounce_return_message and bounce_return_body options were not being
4171 honoured for bounces generated during the reception of non-SMTP messages.
4172 In particular, this applied to messages rejected by the ACL. This bug has
4173 been fixed. However, if bounce_return_message is true and bounce_return_
4174 body is false, the headers that are returned for a non-SMTP message
4175 include only those that have been read before the error was detected.
4176 (In the case of an ACL rejection, they have all been read.)
4178 PH/69 The HTML version of the specification is now built in a directory called
4179 spec_html instead of spec.html, because the latter looks like a path with
4180 a MIME-type, and this confuses some software.
4182 PH/70 Catch two compiler warnings in sieve.c.
4184 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4185 function verify_get_ident() calls ip_connect() to connect a socket, but
4186 if the "connect()" function timed out, ip_connect() used to close the
4187 socket. However, verify_get_ident() also closes the socket later, and in
4188 between Exim writes to the log, which may get opened at this point. When
4189 the socket was closed in ip_connect(), the log could get the same file
4190 descriptor number as the socket. This naturally causes chaos. The fix is
4191 not to close the socket in ip_connect(); the socket should be closed by
4192 the function that creates it. There was only one place in the code where
4193 this was missing, in the iplookup router, which I don't think anybody now
4194 uses, but I've fixed it anyway.
4196 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4197 well as to direct DNS lookups. Otherwise the handling of names in host
4198 lists is inconsistent and therefore confusing.
4204 PH/01 Two changes to the default runtime configuration:
4206 (1) Move the checks for relay_from_hosts and authenticated clients from
4207 after to before the (commented out) DNS black list checks.
4209 (2) Add control=submission to the relay_from_hosts and authenticated
4210 clients checks, on the grounds that messages accepted by these
4211 statements are most likely to be submissions.
4213 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4215 (1) Generate an error if the third argument for the ${prvs expansion is
4218 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4221 (3) Reset the variables that are obtained from the first argument of
4222 ${prvscheck and used in the second argument before leaving the code,
4223 because their memory is reclaimed, so using them afterwards may do
4226 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4227 one (it's much easier than Tom thought :-).
4229 (5) Because of (4), we can now allow for the use of $prvscheck_result
4230 inside the third argument.
4232 PH/03 For some reason, the default setting of PATH when running a command from
4233 a pipe transport was just "/usr/bin". I have changed it to
4236 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4237 anything to be listed in the output from -bV.
4239 PH/05 When a filter generated an autoreply, the entire To: header line was
4240 quoted in the delivery log line, like this:
4242 => >A.N.Other <ano@some.domain> <original@ddress> ...
4244 This has been changed so that it extracts the operative address. There
4245 may be more than one such address. If so, they are comma-separated, like
4248 => >ano@some.domain,ona@other.domain <original@ddress> ...
4250 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4251 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4252 not being looked up in the rDNS to get a host name, Exim was showing the
4253 IP address twice in Received: lines, even though the IP addresses were
4254 identical. For example:
4256 Received: from [1.2.3.4] (helo=[1.2.3.4])
4258 However, if the real host name was known, it was omitting the HELO data
4259 if it matched the actual IP address. This has been tidied up so that it
4260 doesn't show the same IP address twice.
4262 PH/07 When both +timestamp and +memory debugging was on, the value given by
4263 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4264 called by the debug printing, thereby overwriting the timestamp buffer.
4265 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4267 PH/08 When the original message was included in an autoreply transport, it
4268 always said "this is a copy of the message, including all the headers",
4269 even if body_only or headers_only was set. It now gives an appropriate
4272 PH/09 Applied a patch from the Sieve maintainer which:
4274 o fixes some comments
4275 o adds the (disabled) notify extension core
4276 o adds some debug output for the result of if/elsif tests
4277 o points to the current vacation draft in the documentation
4278 and documents the missing references header update
4282 o fixes a bug in processing the envelope test (when testing
4283 multiple envelope elements, the last element determined the
4286 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4287 Electronic Mail") by including:
4289 Auto-submitted: auto-generated
4291 in the messages that it generates (bounce messages and others, such as
4292 warnings). In the case of bounce messages for non-SMTP messages, there was
4293 also a typo: it was using "Auto_submitted" (underscore instead of
4294 hyphen). Since every message generated by Exim is necessarily in response
4295 to another message, thes have all been changed to:
4297 Auto-Submitted: auto-replied
4299 in accordance with these statements in the RFC:
4301 The auto-replied keyword:
4303 - SHOULD be used on messages sent in direct response to another
4304 message by an automatic process,
4306 - MUST NOT be used on manually-generated messages,
4308 - MAY be used on Delivery Status Notifications (DSNs) and Message
4309 Disposition Notifications (MDNs),
4311 - MUST NOT be used on messages generated by automatic or periodic
4312 processes, except for messages which are automatic responses to
4315 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4316 to the default Received: header definition.
4318 PH/12 Added log selector acl_warn_skipped (default on).
4320 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4321 variables because (a) they are in the wrong storage pool and (b) even if
4322 they were copied, it wouldn't work properly because of the caching.
4324 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4325 checking when decoding. Apparently there are clients that generate
4326 overlong encoded strings. Why am I not surprised?
4328 PH/15 If the first argument of "${if match_address" was not empty, but did not
4329 contain an "@" character, Exim crashed. Now it writes a panic log message
4330 and treats the condition as false.
4332 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4334 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4335 extension "envelope-auth". The code is finished and in agreement with
4336 other implementations, but there is no documentation so far and in fact,
4337 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4338 not changing the active code.
4340 Print executed "if" and "elsif" statements when debugging is used. This
4341 helps a great deal to understand what a filter does.
4343 Document more things not specified clearly in RFC3028. I had all this
4344 sorted out, when out of a sudden new issues came to my mind. Oops."
4346 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4349 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4350 canonical form (as documented). However, after a host name lookup from
4351 the IP address, check_host() was doing a simple string comparison with
4352 addresses acquired from the DNS when checking that the found name did
4353 have the original IP as one of its addresses. Since any found IPv6
4354 addresses are likely to be in abbreviated form, the comparison could
4355 fail. Luckily, there already exists a function for doing the comparison
4356 by converting both addresses to binary, so now that is used instead of
4357 the text comparison.
4359 PH/20 There was another similar case to PH/19, when a complete host name was
4360 given in a host list; looking up its IP address could give an abbreviated
4361 form, whereas the current host's name might or might not be abbreviated.
4362 The same fix has been applied.
4368 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4369 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4372 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4373 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4375 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4377 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4378 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4379 Also, for "Auto-Submitted", treat anything other than "no" as
4380 non-personal, in accordance with RFC 3834. (Previously it treated
4381 anything starting "auto-" as non-personal.)
4383 TF/01 The control=submission/name=... option had a problem with syntax
4384 errors if the name included a slash character. The /name= option
4385 now slurps the rest of the string, so it can include any characters
4386 but it must come last in the list of options (after /sender_retain
4389 PH/05 Some modifications to the interface to the fake nameserver for the new
4397 TK/01 Added the "success_on_redirect" address verification option. See
4398 NewStuff for rationale and an example.
4400 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4402 PH/02 Patch to exigrep to allow it to work on syslog lines.
4404 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4405 fread() to skip over the body file's header line, because in Cygwin the
4406 header line is locked and is inaccessible.
4408 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4409 co-exist for some time) to make it clear that it is the Exim ID that is
4410 referenced, not the Message-ID: header line.
4412 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4413 string_format() function, because snprintf() does not exist on all
4416 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4417 ${stat: expansion item.
4419 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4420 protocol synchronization error", to keep the pedants happy.
4422 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4423 well as for IRIX systems, when gcc is being used. See the host.c source
4426 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4428 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4431 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4432 transport and to the smtp transport in LMTP mode.
4434 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4436 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4437 run a filter in a subprocess. This could lead to confusion in subsequent
4438 lookups in the parent process. There should also be a search_tidyup() at
4439 the end of the subprocess.
4441 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4442 only if the host matched helo_try_verify_hosts, which caused the
4443 verification to occur when the EHLO/HELO command was issued. The ACL just
4444 tested the remembered result. Now, if a previous verification attempt has
4445 not happened, "verify = helo" does it there and then.
4447 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4449 TK/03 Fix log output including CR from clamd.
4451 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4452 exist provoked a memory error which could cause a segfault.
4454 PH/15 Installed PCRE 6.2
4456 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4458 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4459 of the problem. Specifically, suggested +O2 rather than +O1 for the
4462 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4464 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4465 record, the port was not correctly logged when the outgoing_port log
4466 selector was set (it logged the transort's default port).
4468 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4469 fallback_hosts, and "hosts" in the smtp transport.
4471 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4472 host errors such as "Connection refused".
4474 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4475 authentication with radiusclient 0.4.9:
4477 - Error returned from rc_read_config was caught wrongly
4478 - Username/password not passed on to radius server due to wrong length.
4480 The presumption is that some radiusclient API changes for 4.51/PH/17
4481 were not taken care of correctly. The code is still untested by me (my
4482 Linux distribution still has 0.3.2 of radiusclient), but it was
4483 contributed by a Radius user.
4485 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4486 expanding the "port" option of the smtp transport.
4488 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4489 while reading a MIME header. Thanks to Tom Hughes for a patch.
4491 PH/24 Include config.h inside local_scan.h so that configuration settings are
4494 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4495 that in an ACL for RCPT (for example), you can examine exactly what was
4498 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4499 commands, but it was not correctly comparing the address with the actual
4500 client host address. Thus, it would show the EHLO address in Received:
4501 header lines when this was not necessary.
4503 PH/27 Added the % operator to ${eval:}.
4505 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4506 it should be ignoring failures (because with -C, for example, it has lost
4507 privilege). It wasn't ignoring creation failures other than "already
4510 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4513 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4514 because an input line was too long, either on its own, or by virtue of
4515 too many continuations, the temporary file was not being removed, and the
4516 return code was incorrect.
4518 PH/31 Missing "BOOL" in function definition in filtertest.c.
4520 PH/32 Applied Sieve patches from the maintainer.
4522 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4524 PH/33 Added "verify = not_blind".
4526 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4527 Local/Makefile (with some defaults set). These are used in built scripts
4528 such as exicyclog, but they have never been used in the exim_install
4529 script (though there are many overriding facilities there). I have
4530 arranged that the exim_install script now takes note of these two
4533 PH/35 Installed configuration files for Dragonfly.
4535 PH/36 When a locally submitted message by a trusted user did not contain a
4536 From: header, and the sender address was obtained from -f or from an SMTP
4537 MAIL command, and the trusted user did not use -F to supply a sender
4538 name, $originator_name was incorrectly used when constructing a From:
4539 header. Furthermore, $originator_name was used for submission mode
4540 messages from external hosts without From: headers in a similar way,
4541 which is clearly wrong.
4543 PH/37 Added control=suppress_local_fixups.
4545 PH/38 When log_selector = +received_sender was set, and the addition of the
4546 sender made the log line's construction buffer exactly full, or one byte
4547 less than full, an overflow happened when the terminating "\n" was
4550 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4551 when the result of a list match is failure because a DNS lookup failed.
4553 PH/40 RM_COMMAND is now used in the building process.
4555 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4556 the "build-* directories that it finds.
4558 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4559 address in a domain literal was a prefix of an interface address.
4561 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4562 when verifying a sender address, unless rewrite_headers is false.
4564 PH/44 Wrote a long comment about why errors_to addresses are verified as
4565 recipients, not senders.
4567 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4568 the ratelimit ACL was added.
4570 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4572 PH/46 Added extra information about PostgreSQL errors to the error string.
4574 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4575 avoiding the need to install special zones in a real server. This is
4576 backwards compatible; if it can't find the fake resolver, it drops back.
4577 Thus, both old and new test suites can be run.
4579 TF/02 Added util/ratelimit.pl
4581 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4582 clock is set back in time.
4584 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4585 Candler <B.Candler@pobox.com>.
4587 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4588 OK for addresses that are the result of redirections.
4590 PH/48 A number of further additions for the benefit of the new test suite,
4591 including a fake gethostbyname() that interfaces to the fake DNS resolver
4594 TF/06 The fix for widen_domains has also been applied to qualify_single and
4595 search_parents which are the other dnslookup options that can cause
4598 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4601 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4603 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4604 service" if it encountered an MX record with an empty target hostname.
4605 The message is now "an MX or SRV record indicated no SMTP service".
4607 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4608 if the DNS of the sending site is misconfigured. This is quite a
4609 common situation. This change restores the behaviour of treating a
4610 helo verification defer as a failure.
4612 PH/51 If self=fail was set on a router, the bounce message did not include the
4613 actual error message.
4619 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4621 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4622 command itself ended in error, the underlying message about the transport
4623 filter timeout was being overwritten with the pipe command error. Now the
4624 underlying error message should be appended to the second error message.
4626 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4628 PH/02 Reduce the amount of output that "make" produces by default. Full output
4629 can still be requested.
4631 PH/03 The warning log line about a condition test deferring for a "warn" verb
4632 was being output only once per connection, rather than after each
4633 occurrence (because it was using the same function as for successful
4634 "warn" verbs). This seems wrong, so I have changed it.
4636 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4637 it should not have, which might have caused a crash in the right
4638 circumstances, but probably never did.
4640 PH/04 Installed a modified version of Tony Finch's patch to make submission
4641 mode fix the return path as well as the Sender: header line, and to
4642 add a /name= option so that you can make the user's friendly name appear
4645 TF/03 Added the control = fakedefer ACL modifier.
4647 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4648 Mark Lowes for thorough testing.
4650 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4652 TK/03 Merged latest SRS patch from Miles Wilton.
4654 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4655 with the definition in sysexits.h (which is #included earlier).
4656 Fortunately, Exim does not actually use EX_OK. The code used to try to
4657 preserve the sysexits.h value, by assuming that macro definitions were
4658 scanned for macro replacements. I have been disabused of this notion,
4659 so now the code just undefines EX_OK before #including unistd.h.
4661 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4662 in the smtp transport. When a block could not be written in a single
4663 write() function, the timeout was being re-applied to each part-write.
4664 This seems wrong - if the receiver was accepting one byte at a time it
4665 would take for ever. The timeout is now adjusted when this happens. It
4666 doesn't have to be particularly precise.
4668 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4669 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4671 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4672 without a domain if the check to <postmaster@domain> fails.
4674 SC/01 Eximstats: added -xls and the ability to specify output files
4675 (patch written by Frank Heydlauf).
4677 SC/02 Eximstats: use FileHandles for outputting results.
4679 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4681 SC/04 Eximstats: fixed display of large numbers with -nvr option
4683 SC/05 Eximstats: fixed merging of reports with empty tables.
4685 SC/06 Eximstats: added the -include_original_destination flag
4687 SC/07 Eximstats: removed tabs and trailing whitespace.
4689 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4691 TK/06 MBOX spool code: Add real "From " MBOX separator line
4692 so the .eml file is really in mbox format (even though
4693 most programs do not really care). Patch from Alex Miller.
4695 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4696 The latter is generated from $received_to and is only set if the
4697 message has one envelope recipient. SA can use these headers,
4698 obviously out-of-the-box. Patch from Alex Miller.
4700 PH/08 The ${def test on a variable was returning false if the variable's
4701 value was "0", contrary to what the specification has always said!
4702 The result should be true unless the variable is empty.
4704 PH/09 The syntax error of a character other than { following "${if
4705 def:variable_name" (after optional whitespace) was not being diagnosed.
4706 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4707 accidental colon was present, for example, could give incorrect results.
4709 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4710 result is used (not including appendfile, where other changes are about
4713 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4714 This involved changing a lot of size variables from int to off_t. It
4715 should work with maildirs and everything.
4717 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4718 spamd dying while we are connected to it.
4720 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4723 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4724 function for BDB 4.3.
4726 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4728 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4729 the sources for PCRE have been re-arranged and more files are now
4732 PH/15 The code I had for printing potentially long long variables in PH/11
4733 above was not the best (it lost precision). The length of off_t variables
4734 is now inspected at build time, and an appropriate printing format (%ld
4735 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4736 to be "long long int" or "long int". This is needed for the internal
4737 formatting function string_vformat().
4739 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4740 the configuration file to be ":syslog", then the script "guesses" where
4741 the logs files are, rather than using the compiled in default. In our
4742 case the guess is not the same as the compiled default, so the script
4743 suddenly stopped working when I started to use syslog. The patch checks
4744 to see if log_file_path is "". If so, it attempts to read it from exim
4745 with no configuration file to get the compiled in version, before it
4746 falls back to the previous guessing code."
4748 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4749 implementing BATV in an Exim configuration. See NewStuff for the gory
4752 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4753 Makefile that are specific to HP-UX.
4755 PH/18 If the "use_postmaster" option was set for a recipient callout together
4756 with the "random" option, the postmaster address was used as the MAIL
4757 FROM address for the random test, but not for the subsequent recipient
4758 test. It is now used for both.
4760 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4761 patch removes a few documentation additions to RFC 3028, because the
4762 latest draft now contains them. It adds the new en;ascii-case comparator
4763 and a new error check for 8bit text in MIME parts. Comparator and
4764 require names are now matched exactly. I enabled the subaddress
4765 extension, but it is not well tested yet (read: it works for me)."
4767 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4768 rework some of the code of TK/09 above to avoid the hardwired use of
4769 "%lld" and "long long". Replaced the call to snprintf() with a call to
4772 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4773 records point to non-existent hosts", "retry timeout exceeded", and
4774 "retry time not reached for any host after a long failure period".
4776 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4777 experimental DomainKeys support:
4779 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4780 (2) On an error such as an illegally used "control", the wrong name for
4781 the control was given.
4783 These problems did NOT occur unless DomainKeys support was compiled.
4785 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4787 PH/24 Added ${if match_ip condition.
4789 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4790 databases so that it will be absolutely obvious if a crash occurs in the
4791 DB library. This is a regular occurrence (often caused by mis-matched
4794 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4795 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4796 picked up on a user's system that detects such things. There doesn't seem
4797 to be a gcc warning option for this - only an attribute that has to be
4798 put on the function's prototype. It seems that in Fedora Core 4 they have
4799 set this on a number of new functions. No doubt there will be more in due
4802 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4803 specify a transport. However, if an address that was verified by such a
4804 router was the subject of a callout, Exim crashed because it tried to
4805 read the rcpt_include_affixes from the non-existent transport. Now it
4806 just assumes that the setting of that option is false. This bug was
4807 introduced by 4.51/PH/31.
4809 PH/28 Changed -d+all to exclude +memory, because that information is very
4810 rarely of interest, but it makes the output a lot bigger. People tend to
4811 do -d+all out of habit.
4813 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4814 code in os-type was giving problems when libc.so lives in lib64, like on
4817 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4818 aren't the modern standard, and it seems that some systems' include files
4819 don't always have them. Exim was already checking for some of the newer
4820 ones like T_AAAA, and defining it itself. I've added checks for all the
4821 record types that Exim uses.
4823 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4824 not automatically generating a new one, as it is supposed to. This
4825 prevented TLS from working. If the file did exist, but contained invalid
4826 data, a new version was generated, as expected. It was only the case of a
4827 non-existent file that was broken.
4829 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4830 with a change in libdomainkeys > 0.64.
4832 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4833 from DNS. If the selector record carries the flag, it now has
4834 precedence over the domain-wide flag.
4836 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4838 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4839 the use of a port name that isn't defined in /etc/services) occurred, the
4840 message was deferred as in a normal delivery, and thus remained on the
4841 spool, instead of being failed because of the mua_wrapper setting. This
4842 is now fixed, and I tidied up some of the mua_wrapper messages at the
4845 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4846 the messages in a hash of arrays rather than using individual hashes.
4847 This is a bit cleaner and results in dramatic memory savings, albeit
4848 at a slight CPU cost.
4850 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4851 as requested by Marc Sherman.
4853 SC/10 Eximstats: added histograms for user specified patterns as requested
4856 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4858 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4859 fopen() in the content-scanning modules that did not already have it.
4865 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4866 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4868 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4870 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4872 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4873 to test Sieve filters that use "vacation".
4875 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4876 that changes the way the GnuTLS parameters are stored in the cache file.
4877 The new format can be generated externally. For backward compatibility,
4878 if the data in the cache doesn't make sense, Exim assumes it has read an
4879 old-format file, and it generates new data and writes a new file. This
4880 means that you can't go back to an older release without removing the
4883 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4884 work if there are any delivery delays because "one_time" forces the
4885 parent to be marked "delivered", so its unseen clone is never tried
4886 again. For this reason, Exim now forbids the simultaneous setting of
4889 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4890 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4891 entry if you want to know the details. The fix, however, neglected to
4892 consider the case when local delivery batching is involved. The test for
4893 "previously delivered" was not happening when checking to see if an
4894 address could be batched with a previous (undelivered) one; under
4895 certain circumstances this could lead to multiple deliveries to the same
4898 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4899 in its include files, and this causes problems building Exim.
4901 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4902 header_syntax) but Exim was just ignoring anything given after a slash.
4903 In particular, this caused confusion with an attempt to use "verify =
4904 reverse_host_lookup/defer_ok". An error is now given when options are
4905 supplied for verify items that do not have them. (Maybe reverse_host_
4906 lookup should have a defer_ok option, but that's a different point.)
4908 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4909 defined by RFC 821) to 2048, because there were problems with some AUTH
4910 commands, and RFC 1869 says the size should be increased for extended
4911 SMTP commands that take arguments.
4913 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4916 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4917 "unknown" error; now it says that the functionality isn't in the binary.
4919 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4920 an address' error message when a string expansion fails (syntax or
4921 whatever). Otherwise the password may appear in the log. Following change
4922 PH/42 below, there is no longer a chance of it appearing in a bounce
4925 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4927 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4928 of its addresses was ever tried. (Bugzilla bug #2.)
4930 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4931 the result incorrectly in the debug output. (It correctly added a newline
4932 to what was transported.)
4934 TF/01 Added $received_time.
4936 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4937 commented out examples of how to interface to a virus scanner and to
4938 SpamAssassin. Also added commented examples of av_scanner and
4939 spamd_address settings.
4941 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4942 and controls are allowed in which ACLs. There were a couple of minor
4943 errors. Some of the entries in the conditions table (which is a table of
4944 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4945 negation of where the condition IS allowed.
4947 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4949 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4950 header file does not have a version number, so I've had to invent a new
4951 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4952 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4953 radiusclient), but it was contributed by a Radius user.
4955 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4956 files or directories, for OpenSSL.
4958 PH/19 When an Exim process that is running as root has to create an Exim log
4959 file, it does so in a subprocess that runs as exim:exim so as to get the
4960 ownership right at creation (otherwise, other Exim processes might see
4961 the file with the wrong ownership). There was no test for failure of this
4962 fork() call, which would lead to the process getting stuck as it waited
4963 for a non-existent subprocess. Forks do occasionally fail when resources
4964 run out. I reviewed all the other calls to fork(); they all seem to check
4967 PH/20 When checking for unexpected SMTP input at connect time (before writing
4968 the banner), Exim was not dealing correctly with a non-positive return
4969 from the read() function. If the client had disconnected by this time,
4970 the result was a log entry for a synchronization error with an empty
4971 string after "input=" when read() returned zero. If read() returned -1
4972 (an event I could not check), uninitialized data bytes were printed.
4973 There were reports of junk text (parts of files, etc) appearing after
4976 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4978 PH/22 Added support for macro redefinition, and (re)definition in between
4979 driver and ACL definitions.
4981 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4982 forgetting to use the resulting value; it was using the unexpanded value.
4984 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4985 hadn't been configured. The fix is from Juergen Kreileder, who
4986 understands it better than I do:
4988 "Here's what I see happening with three configured cyrus_sasl
4989 authenticators configured (plain, login, cram-md5):
4991 On startup auth_cyrus_sasl_init() gets called for each of these.
4992 This means three calls to sasl_listmech() without a specified mech_list.
4993 => SASL tests which mechs of all available mechs actually work
4994 => three warnings about OTP not working
4995 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4997 With the patch, sasl_listmech() also gets called three times. But now
4998 SASL's mech_list option is set to the server_mech specified in the the
4999 authenticator. Or in other words, the answer from sasl_listmech()
5000 gets limited to just the mech you're testing for (which is different
5002 => the return list contains just 'plain' or 'login', 'cram-md5' or
5003 nothing depending on the value of ob->server_mech.
5005 I've just tested the patch: Authentication still works fine,
5006 unavailable mechs specified in the exim configuration are still
5007 caught, and the auth.log warnings about OTP are gone."
5009 PH/25 When debugging is enabled, the contents of the command line are added
5010 to the debugging output, even when log_selector=+arguments is not
5013 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5014 answer is "GNU", and only if the return is "GNU/something" is the answer
5017 PH/27 $acl_verify_message is now set immediately after the failure of a
5018 verification in an ACL, and so is available in subsequent modifiers. In
5019 particular, the message can be preserved by coding like this:
5021 warn !verify = sender
5022 set acl_m0 = $acl_verify_message
5024 Previously, $acl_verify_message was set only while expanding "message"
5025 and "log_message" when a very denied access.
5027 PH/28 Modified OS/os.c-Linux with
5029 -#ifndef OS_LOAD_AVERAGE
5030 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5032 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5033 nomenclature these days.)
5035 PH/29 Installed patch from the Sieve maintainer that adds the options
5036 sieve_useraddress and sieve_subaddress to the redirect router.
5038 PH/30 In these circumstances:
5039 . Two addresses routed to the same list of hosts;
5040 . First host does not offer TLS;
5041 . First host accepts first address;
5042 . First host gives temporary error to second address;
5043 . Second host offers TLS and a TLS session is established;
5044 . Second host accepts second address.
5045 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5046 and peerdn, if requested) that were in fact used only for the second
5049 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5050 attention to any local part prefix or suffix that was matched by the
5051 router that accepted the address. It now behaves in the same way as it
5052 does for delivery: the affixes are removed from the local part unless
5053 rcpt_include_affixes is set on the transport.
5055 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5056 timeout during the DATA phase of an incoming message.
5058 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5059 applied a patch sanctioned by the Sieve maintainer.
5061 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5062 the uid or gid is negative. A case of a negative gid caused this to be
5063 noticed. The fix allows for either to be negative.
5065 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5066 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5069 PH/36 The change PH/12 above was broken. Fixed it.
5071 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5072 the grounds that routing the same address twice would always produce the
5073 same answer. This might have been true once, but it is certainly no
5074 longer true now. Routing a child address may depend on the previous
5075 routing that produced that child. Some complicated redirection strategies
5076 went wrong when messages had multiple recipients, and made Exim's
5077 behaviour dependent on the order in which the addresses were given.
5079 I have moved the duplicate checking until after the routing is complete.
5080 Exim scans the addresses that are assigned to local and remote
5081 transports, and removes any duplicates. This means that more work will be
5082 done, as duplicates will always all be routed, but duplicates are
5083 presumably rare, so I don't expect this is of any significance.
5085 For deliveries to pipes, files, and autoreplies, the duplicate checking
5086 still happens during the routing process, since they are not going to be
5089 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5090 It corrects a timeout issue with spamd. This is Ian's comment: "The
5091 background is that sometimes spamd either never reads data from a
5092 connection it has accepted, or it never writes response data. The exiscan
5093 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5094 blindly assumes that writes won't block so it may never time out."
5096 PH/39 Allow G after quota size as well as K and M.
5098 PH/40 The value set for $authenticated_id in an authenticator may not contain
5099 binary zeroes or newlines because the value is written to log lines and
5100 to spool files. There was no check on this. Now the value is run through
5101 the string_printing() function so that such characters are converted to
5102 printable escape sequences.
5104 PH/41 $message_linecount is a new variable that contains the total number of
5105 lines in the message. Compare $body_linecount, which is the count for the
5108 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5109 bounce and delay warning messages, except in certain special cases, which
5112 (a) An SMTP error message from a remote host;
5113 (b) A message specified in a :fail: redirection;
5114 (c) A message specified in a "fail" command in a system filter;
5115 (d) A message specified in a FAIL return from the queryprogram router;
5116 (e) A message specified by the cannot_route_message router option.
5118 In these cases only, Exim does include the error details in bounce and
5119 warning messages. There are also a few cases where bland messages such
5120 as "unrouteable address" or "local delivery error" are given.
5122 PH/43 $value is now also set for the "else" part of a ${run expansion.
5124 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5125 being worked on, but at least Exim now implements the latest version to
5128 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5129 process to complete was treated as a delivery failure, a timeout while
5130 writing the message to the pipe was logged, but erroneously treated as a
5131 successful delivery. Such timeouts include transport filter timeouts. For
5132 consistency with the overall process timeout, these timeouts are now
5133 treated as errors, giving rise to delivery failures by default. However,
5134 there is now a new Boolean option for the pipe transport called
5135 timeout_defer, which, if set TRUE, converts the failures into defers for
5136 both kinds of timeout. A transport filter timeout is now identified in
5139 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5140 systems where "make" and "gmake" are different, calling "gmake" at top
5141 level broke things. I've arranged for the value of $(MAKE) to be passed
5142 from the Makefile to this script so that it can call the same version of
5146 A note about Exim versions 4.44 and 4.50
5147 ----------------------------------------
5149 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5150 changes of various kinds. As a consequence, a big documentation update was
5151 needed. This delayed the release for rather longer than seemed good, especially
5152 in the light of a couple of (minor) security issues. Therefore, the changes
5153 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5154 release. So 4.44 and 4.50 are in effect two different branches that both start
5157 I have left the 4.50 change log unchanged; it contains all the changes since
5158 4.43. The change log for 4.44 is below; many of its items are identical to
5159 those for 4.50. This seems to be the most sensible way to preserve the
5160 historical information.
5166 1. Minor wording change to the doc/README.SIEVE file.
5168 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5169 computation of the current number of files was incorrect.
5171 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5172 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5175 4. Give more explanation in the error message when the command for a transport
5176 filter fails to execute.
5178 5. There are several places where Exim runs a non-Exim command in a
5179 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5180 was being done only for the command run by the queryprogram router. It is
5181 now done for all such subprocesses. The other cases are: ${run, transport
5182 filters, and the commands run by the lmtp and pipe transports.
5184 6. Added CONFIGURE_GROUP build-time option.
5186 7. Some older OS have a limit of 256 on the maximum number of file
5187 descriptors. Exim was using setrlimit() to set 1000 as a large value
5188 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5189 systems. I've change it so that if it can't get 1000, it tries for 256.
5191 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5192 was an oversight, and furthermore, ever since the addition of extra
5193 controls (e.g. 4.43/32), the checks on when to allow different forms of
5194 "control" were broken. There should now be diagnostics for all cases when a
5195 control that does not make sense is encountered.
5197 9. Added the /retain_sender option to "control=submission".
5199 10. $recipients is now available in the predata ACL (oversight).
5201 11. Tidy the search cache before the fork to do a delivery from a message
5202 received from the command line. Otherwise the child will trigger a lookup
5203 failure and thereby defer the delivery if it tries to use (for example) a
5204 cached ldap connection that the parent has called unbind on.
5206 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5207 of $address_data from the recipient verification was clobbered by the
5208 sender verification.
5210 13. The value of address_data from a sender verification is now available in
5211 $sender_address_data in subsequent conditions in the ACL statement.
5213 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5215 15. Added a new option "connect=<time>" to callout options, to set a different
5218 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5219 was its contents. (It was OK if the option was not defined at all.)
5221 17. A "Completed" log line is now written for messages that are removed from
5222 the spool by the -Mrm option.
5224 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5225 information about exactly what failed.
5227 19. Added -dd to debug only the daemon process.
5229 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5230 handles timeouts, both on the server side and network timeouts. Renamed the
5231 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5233 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5234 It is now set to "smtps".
5236 22. $host_address is now set to the target address during the checking of
5237 ignore_target_hosts.
5239 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5240 being passed; this would have caused $sender_host_name to have been used if
5241 matching the list had actually called for a host name (not very likely,
5242 since this list is usually IP addresses). A host name is now passed as
5245 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5246 code that specifies a non-restarting handler (typically sigaction() in
5247 modern systems) in an attempt to fix a rare and obscure crash bug.
5249 25. Narrowed the window for a race in the daemon that could cause it to ignore
5250 SIGCHLD signals. This is not a major problem, because they are used only to
5251 wake it up if nothing else does.
5253 26. A malformed maildirsize file could cause Exim to calculate negative values
5254 for the mailbox size or file count. Odd effects could occur as a result.
5255 The maildirsize information is now recalculated if the size or filecount
5258 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5259 support for a long time. Removed HAVE_SYS_VFS_H.
5261 28. Installed the latest version of exipick from John Jetmore.
5263 29. In an address list, if the pattern was not a regular expression, an empty
5264 subject address (from a bounce message) matched only if the pattern was an
5265 empty string. Non-empty patterns were not even tested. This was the wrong
5266 because it is perfectly reasonable to use an empty address as part of a
5267 database query. An empty address is now tested by patterns that are
5268 lookups. However, all the other forms of pattern expect the subject to
5269 contain a local part and a domain, and therefore, for them, an empty
5270 address still always fails if the pattern is not itself empty.
5272 30. Exim went into a mad DNS loop when attempting to do a callout where the
5273 host was specified on an smtp transport, and looking it up yielded more
5274 than one IP address.
5276 31. Re-factored the code for checking spool and log partition space into a
5277 function that finds that data and another that does the check. The former
5278 is then used to implement four new variables: $spool_space, $log_space,
5279 $spool_inodes, and $log_inodes.
5281 32. The RFC2047 encoding function was originally intended for short strings
5282 such as real names; it was not keeping to the 75-character limit for
5283 encoded words that the RFC imposes. It now respects the limit, and
5284 generates multiple encoded words if necessary. To be on the safe side, I
5285 have increased the buffer size for the ${rfc2047: expansion operator from
5288 33. It is now permitted to omit both strings after an "if" condition; if the
5289 condition is true, the result is "true". As before, when the second string
5290 is omitted, a false condition yields an empty string. This makes it less
5291 cumbersome to write custom ACL and router conditions.
5293 34. Failure to deliver a bounce message always caused it to be frozen, even if
5294 there was an errors_to setting on the router. The errors_to setting is now
5297 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5298 canonical form (fully expanded) before being placed in
5299 $sender_host_address.
5301 36. The table in the code that translates DNS record types into text (T_A to
5302 "A" for instance) was missing entries for NS and CNAME. It is just possible
5303 that this could have caused confusion if both these types were looked up
5304 for the same domain, because the text type is used as part of Exim's
5305 per-process caching. But the chance of anyone hitting this buglet seems
5308 37. The dnsdb lookup has been extended in a number of ways.
5310 (1) There is a new type, "zns", which walks up the domain tree until it
5311 finds some nameserver records. It should be used with care.
5313 (2) There is a new type, "mxh", which is like "mx" except that it returns
5314 just the host names, not the priorities.
5316 (3) It is now possible to give a list of domains (or IP addresses) to be
5317 looked up. The behaviour when one of the lookups defers can be
5318 controlled by a keyword.
5320 (4) It is now possible to specify the separator character for use when
5321 multiple records are returned.
5323 38. The dnslists ACL condition has been extended: it is now possible to supply
5324 a list of IP addresses and/or domains to be looked up in a particular DNS
5327 39. Added log_selector=+queue_time_overall.
5329 40. When running the queue in the test harness, wait just a tad after forking a
5330 delivery process, to get repeatability of debugging output.
5332 41. Include certificate and key file names in error message when GnuTLS fails
5333 to set them up, because the GnuTLS error message doesn't include the name
5334 of the failing file when there is a problem reading it.
5336 42. Allow both -bf and -bF in the same test run.
5338 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5340 44. The "Exiscan patch" is now merged into the mainline Exim source.
5342 45. Sometimes the final signoff response after QUIT could fail to get
5343 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5344 < 0 before doing a fflush(). This bug looks as though it goes back to the
5345 introduction of TLS in release 3.20, but "sometimes" must have been rare
5346 because the tests only now provoked it.
5348 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5349 (this can affect the format of dates).
5351 47. exim_tidydb, when checking for the continued existence of a message for
5352 which it has found a message-specific retry record, was not finding
5353 messages that were in split spool directories. Consequently, it was
5354 deleting retry records that should have stayed in existence.
5356 48. Steve fixed some bugs in eximstats.
5358 49. The SPA authentication driver was not abandoning authentication and moving
5359 on to the next authenticator when an expansion was forced to fail,
5360 contradicting the general specification for all authenticators. Instead it
5361 was generating a temporary error. It now behaves as specified.
5363 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5364 (the order specifies the preference for clients). The order is now AES256,
5365 AES128, 3DES, ARCFOUR128.
5367 51. Small patch to Sieve code - explicitly set From: when generating an
5370 52. Exim crashed if a remote delivery caused a very long error message to be
5371 recorded - for instance if somebody sent an entire SpamAssassin report back
5372 as a large number of 550 error lines. This bug was coincidentally fixed by
5373 increasing the size of one of Exim's internal buffers (big_buffer) that
5374 happened as part of the Exiscan merge. However, to be on the safe side, I
5375 have made the code more robust (and fixed the comments that describe what
5378 53. Now that there can be additional text after "Completed" in log lines (if
5379 the queue_time_overall log selector is set), a one-byte patch to exigrep
5380 was needed to allow it to recognize "Completed" as not the last thing in
5383 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5384 patch that reportedly fixes this has been added. I am not expert enough to
5385 create a test for it. This is what the patch creator wrote:
5387 "I found a little strange behaviour of ldap code when working with
5388 Windows 2003 AD Domain, where users was placed in more than one
5389 Organization Units. When I tried to give exim partial DN, the exit code
5390 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5391 But simultaneously result of request was absolutely normal ldap result,
5392 so I produce this patch..."
5394 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5395 REFERENCE, so I have modified the code to exclude the patch when that macro
5398 55. Some experimental protocols are using DNS PTR records for new purposes. The
5399 keys for these records are domain names, not reversed IP addresses. The
5400 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5401 leaves it alone. Component reversal etc. now happens only for IP addresses.
5404 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5406 57. Double the size of the debug message buffer (to 2048) so that more of very
5407 long debug lines gets shown.
5409 58. The exicyclog utility now does better if the number of log files to keep
5410 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5412 59. Two changes related to the smtp_active_hostname option:
5414 (1) $smtp_active_hostname is now available as a variable.
5415 (2) The default for smtp_banner uses $smtp_active_hostname instead
5416 of $primary_hostname.
5418 60. The host_aton() function is supposed to be passed a string that is known
5419 to be a valid IP address. However, in the case of IPv6 addresses, it was
5420 not checking this. This is a hostage to fortune. Exim now panics and dies
5421 if the condition is not met. A case was found where this could be provoked
5422 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5423 components; fortuitously, this particular loophole had already been fixed
5424 by change 4.50/55 above.
5426 If there are any other similar loopholes, the new check in host_aton()
5427 itself should stop them being exploited. The report I received stated that
5428 data on the command line could provoke the exploit when Exim was running as
5429 exim, but did not say which command line option was involved. All I could
5430 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5431 running as the user.
5434 61. There was a buffer overflow vulnerability in the SPA authentication code
5435 (which came originally from the Samba project). I have added a test to the
5436 spa_base64_to_bits() function which I hope fixes it.
5439 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5440 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5442 63. The daemon start-up calls getloadavg() while still root for those OS that
5443 need the first call to be done as root, but it missed one case: when
5444 deliver_queue_load_max is set with deliver_drop_privilege. This is
5445 necessary for the benefit of the queue runner, because there is no re-exec
5446 when deliver_drop_privilege is set.
5448 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5449 This has been fixed.
5451 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5452 was in use, was not putting the data itself into the right store pool;
5453 consequently, it could be overwritten for a subsequent message in the same
5454 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5457 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5459 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5460 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5461 the function were treating the return as a boolean value, which happened to
5462 work because 0=false and not-0=true, but is not correct code.
5464 68. The host_aton() function was not handling scoped IPv6 addresses (those
5465 with, for example, "%eth0" on the end) correctly.
5467 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5468 negated items (that is, ~something) in unsigned ints. Some compilers
5469 apparently mutter when there is no cast.
5471 70. If an address verification called from an ACL failed, and did not produce a
5472 user-specific message (i.e. there was only a "system" message), nothing was
5473 put in $acl_verify_message. In this situation, it now puts the system
5476 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5477 session; change 4.31/43 added the unwanted input to the log line - except
5478 that it did not do this in the start of session case. It now does.
5480 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5481 This is wrong and can cause the other end to generate a synchronization
5482 error if it is another Exim or anything else that does the synchronization
5483 check. A QUIT command is no longer sent after a timeout.
5485 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5486 during host lookups.
5488 74. The defer_ok option of callout verification was not working if it was used
5489 when verifying addresses in header lines, that is, for this case:
5491 verify = header_sender/callout=defer_ok
5493 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5494 those file descriptors could be used for SMTP connections. If anything
5495 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5496 could be sent to the SMTP client, causing chaos. The daemon now opens
5497 stdin, stdout, and stderr to /dev/null when it puts itself into the
5500 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5501 log by default. The user can override this with suitable Perl magic.
5503 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5504 the log message when discard triggers, was not working for the DATA ACL or
5505 for the non-SMTP ACL.
5507 78. Error message wording change in sieve.c.
5509 79. If smtp_accept_max_per_host was set, the number of connections could be
5510 restricted to fewer than expected, because the daemon was trying to set up
5511 a new connection before checking whether the processes handling previous
5512 connections had finished. The check for completed processes is now done
5513 earlier. On busy systems, this bug wouldn't be noticed because something
5514 else would have woken the daemon, and it would have reaped the completed
5517 80. If a message was submitted locally by a user whose login name contained one
5518 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5519 It caused a spool format error. I have fixed the spool reading code. A
5520 related problem was that the "from" clause in the Received: line became
5521 illegal because of the space(s). It is now covered by ${quote_local_part.
5523 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5526 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5528 83. Patch from Sieve maintainer to fix unterminated string problem in
5529 "vacation" handling.
5531 84. Some minor changes to the Linux configuration files to help with other
5532 OS variants using glibc.
5534 85. One more patch for Sieve to update vacation handling to latest spec.
5537 ----------------------------------------------------
5538 See the note above about the 4.44 and 4.50 releases.
5539 ----------------------------------------------------
5545 1. Change 4.43/35 introduced a bug that caused file counts to be
5546 incorrectly computed when quota_filecount was set in an appendfile
5549 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5550 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5553 3. Give more explanation in the error message when the command for a transport
5554 filter fails to execute.
5556 4. There are several places where Exim runs a non-Exim command in a
5557 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5558 was being done only for the command run by the queryprogram router. It is
5559 now done for all such subprocesses. The other cases are: ${run, transport
5560 filters, and the commands run by the lmtp and pipe transports.
5562 5. Some older OS have a limit of 256 on the maximum number of file
5563 descriptors. Exim was using setrlimit() to set 1000 as a large value
5564 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5565 systems. I've change it so that if it can't get 1000, it tries for 256.
5567 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5568 was an oversight, and furthermore, ever since the addition of extra
5569 controls (e.g. 4.43/32), the checks on when to allow different forms of
5570 "control" were broken. There should now be diagnostics for all cases when a
5571 control that does not make sense is encountered.
5573 7. $recipients is now available in the predata ACL (oversight).
5575 8. Tidy the search cache before the fork to do a delivery from a message
5576 received from the command line. Otherwise the child will trigger a lookup
5577 failure and thereby defer the delivery if it tries to use (for example) a
5578 cached ldap connection that the parent has called unbind on.
5580 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5581 of $address_data from the recipient verification was clobbered by the
5582 sender verification.
5584 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5585 was its contents. (It was OK if the option was not defined at all.)
5587 11. A "Completed" log line is now written for messages that are removed from
5588 the spool by the -Mrm option.
5590 12. $host_address is now set to the target address during the checking of
5591 ignore_target_hosts.
5593 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5594 being passed; this would have caused $sender_host_name to have been used if
5595 matching the list had actually called for a host name (not very likely,
5596 since this list is usually IP addresses). A host name is now passed as
5599 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5600 code that specifies a non-restarting handler (typically sigaction() in
5601 modern systems) in an attempt to fix a rare and obscure crash bug.
5603 15. Narrowed the window for a race in the daemon that could cause it to ignore
5604 SIGCHLD signals. This is not a major problem, because they are used only to
5605 wake it up if nothing else does.
5607 16. A malformed maildirsize file could cause Exim to calculate negative values
5608 for the mailbox size or file count. Odd effects could occur as a result.
5609 The maildirsize information is now recalculated if the size or filecount
5612 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5613 support for a long time. Removed HAVE_SYS_VFS_H.
5615 18. Updated exipick to current release from John Jetmore.
5617 19. Allow an empty sender to be matched against a lookup in an address list.
5618 Previously the only cases considered were a regular expression, or an
5621 20. Exim went into a mad DNS lookup loop when doing a callout where the
5622 host was specified on the transport, if the DNS lookup yielded more than
5625 21. The RFC2047 encoding function was originally intended for short strings
5626 such as real names; it was not keeping to the 75-character limit for
5627 encoded words that the RFC imposes. It now respects the limit, and
5628 generates multiple encoded words if necessary. To be on the safe side, I
5629 have increased the buffer size for the ${rfc2047: expansion operator from
5632 22. Failure to deliver a bounce message always caused it to be frozen, even if
5633 there was an errors_to setting on the router. The errors_to setting is now
5636 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5637 canonical form (fully expanded) before being placed in
5638 $sender_host_address.
5640 24. Updated eximstats to version 1.33
5642 25. Include certificate and key file names in error message when GnuTLS fails
5643 to set them up, because the GnuTLS error message doesn't include the name
5644 of the failing file when there is a problem reading it.
5646 26. Expand error message when OpenSSL has problems setting up cert/key files.
5649 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5650 (this can affect the format of dates).
5652 28. exim_tidydb, when checking for the continued existence of a message for
5653 which it has found a message-specific retry record, was not finding
5654 messages that were in split spool directories. Consequently, it was
5655 deleting retry records that should have stayed in existence.
5657 29. eximstats updated to version 1.35
5658 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5659 1.35 - bugfix such that pie charts by volume are generated correctly
5661 30. The SPA authentication driver was not abandoning authentication and moving
5662 on to the next authenticator when an expansion was forced to fail,
5663 contradicting the general specification for all authenticators. Instead it
5664 was generating a temporary error. It now behaves as specified.
5666 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5667 (the order specifies the preference for clients). The order is now AES256,
5668 AES128, 3DES, ARCFOUR128.
5670 31. Small patch to Sieve code - explicitly set From: when generating an
5673 32. Exim crashed if a remote delivery caused a very long error message to be
5674 recorded - for instance if somebody sent an entire SpamAssassin report back
5675 as a large number of 550 error lines. This bug was coincidentally fixed by
5676 increasing the size of one of Exim's internal buffers (big_buffer) that
5677 happened as part of the Exiscan merge. However, to be on the safe side, I
5678 have made the code more robust (and fixed the comments that describe what
5681 33. Some experimental protocols are using DNS PTR records for new purposes. The
5682 keys for these records are domain names, not reversed IP addresses. The
5683 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5684 leaves it alone. Component reversal etc. now happens only for IP addresses.
5687 34. The host_aton() function is supposed to be passed a string that is known
5688 to be a valid IP address. However, in the case of IPv6 addresses, it was
5689 not checking this. This is a hostage to fortune. Exim now panics and dies
5690 if the condition is not met. A case was found where this could be provoked
5691 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5692 components; fortuitously, this particular loophole had already been fixed
5693 by change 4.50/55 or 4.44/33 above.
5695 If there are any other similar loopholes, the new check in host_aton()
5696 itself should stop them being exploited. The report I received stated that
5697 data on the command line could provoke the exploit when Exim was running as
5698 exim, but did not say which command line option was involved. All I could
5699 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5700 running as the user.
5703 35. There was a buffer overflow vulnerability in the SPA authentication code
5704 (which came originally from the Samba project). I have added a test to the
5705 spa_base64_to_bits() function which I hope fixes it.
5708 36. The daemon start-up calls getloadavg() while still root for those OS that
5709 need the first call to be done as root, but it missed one case: when
5710 deliver_queue_load_max is set with deliver_drop_privilege. This is
5711 necessary for the benefit of the queue runner, because there is no re-exec
5712 when deliver_drop_privilege is set.
5714 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5715 was in use, was not putting the data itself into the right store pool;
5716 consequently, it could be overwritten for a subsequent message in the same
5717 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5720 38. Sometimes the final signoff response after QUIT could fail to get
5721 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5722 < 0 before doing a fflush(). This bug looks as though it goes back to the
5723 introduction of TLS in release 3.20, but "sometimes" must have been rare
5724 because the tests only now provoked it.
5730 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5731 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5732 Now it returns an integer. A number of calls were still expecting a T/F
5733 return. Fortuitously, in all cases, the tests worked in OK situations,
5734 which is the norm. However, things would have gone wrong on any write
5735 failures on the smtp file descriptor. This function is used when sending
5736 messages over SMTP and also when doing verify callouts.
5738 2. When Exim is called to do synchronous delivery of a locally submitted
5739 message (the -odf or -odi options), it no longer closes stderr before doing
5742 3. Implemented the mua_wrapper option.
5744 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5746 5. Implemented the functions header_remove(), header_testname(),
5747 header_add_at_position(), and receive_remove_recipient(), and exported them
5750 6. If an ACL "warn" statement specified the addition of headers, Exim already
5751 inserted X-ACL-Warn: at the start if there was no header name. However, it
5752 was not making this test for the second and subsequent header lines if
5753 there were newlines in the string. This meant that an invalid header could
5754 be inserted if Exim was badly configured.
5756 7. Allow an ACL "warn" statement to add header lines at the start or after all
5757 the Received: headers, as well as at the end.
5759 8. Added the rcpt_4xx retry error code.
5761 9. Added postmaster_mailfrom=xxx to callout verification option.
5763 10. Added mailfrom=xxxx to the callout verification option, for verify=
5766 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5767 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5769 12. Inserted some casts to stop certain compilers warning when using pointer
5770 differences as field lengths or precisions in printf-type calls (mostly
5771 affecting debugging statements).
5773 13. Added optional readline() support for -be (dynamically loaded).
5775 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5776 same clock tick as a message's arrival, so that its received time was the
5777 same as the "first fail" time on the retry record, and that message
5778 remained on the queue past the ultimate address timeout, every queue runner
5779 would try a delivery (because it was past the ultimate address timeout) but
5780 after another failure, the ultimate address timeout, which should have then
5781 bounced the address, did not kick in. This was a "< instead of <=" error;
5782 in most cases the first failure would have been in the next clock tick
5783 after the received time, and all would be well.
5785 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5786 being recognized when the domain list was tested by the match_domain
5787 condition in an expansion string.
5789 16. Added the ${str2b64: operator.
5791 17. Exim was always calling setrlimit() to set a large limit for the number of
5792 processes, without checking whether the existing limit was already
5793 adequate. (It did check for the limit on file descriptors.) Furthermore,
5794 errors from getrlimit() and setrlimit() were being ignored. Now they are
5795 logged to the main and panic logs, but Exim does carry on, to try to do its
5796 job under whatever limits there are.
5798 18. Imported PCRE 5.0.
5800 19. Trivial typo in log message " temporarily refused connection" (the leading
5803 20. If the log selector return_path_on_delivery was set and an address was
5804 redirected to /dev/null, the delivery process crashed because it assumed
5805 that a return path would always be set for a "successful" delivery. In this
5806 case, the whole delivery is bypassed as an optimization, and therefore no
5809 21. Internal re-arrangement: the function for sending a challenge and reading
5810 a response while authentication was assuming a zero-terminated challenge
5811 string. It's now changed to take a pointer and a length, to allow for
5812 binary data in such strings.
5814 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5816 23. Exim was not respecting finduser_retries when seeking the login of the
5817 uid under which it was called; it was always trying 10 times. (The default
5818 setting of finduser_retries is zero.) Also, it was sleeping after the final
5819 failure, which is pointless.
5821 24. Implemented tls_on_connect_ports.
5823 25. Implemented acl_smtp_predata.
5825 26. If the domain in control=submission is set empty, Exim assumes that the
5826 authenticated id is a complete email address when it generates From: or
5827 Sender: header lines.
5829 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5830 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5831 chown and chgrp in /bin and hostname in /usr/bin.
5833 28. Exim was keeping the "process log" file open after each use, just as it
5834 does for the main log. This opens the possibility of it remaining open for
5835 long periods when the USR1 signal hits a daemon. Occasional processlog
5836 errors were reported, that could have been caused by this. Anyway, it seems
5837 much more sensible not to leave this file open at all, so that is what now
5840 29. The long-running daemon process does not normally write to the log once it
5841 has entered its main loop, and it closes the log before doing so. This is
5842 so that log files can straightforwardly be renamed and moved. However,
5843 there are a couple of unusual error situations where the daemon does write
5844 log entries, and I had neglected to close the log afterwards.
5846 30. The text of an SMTP error response that was received during a remote
5847 delivery was being truncated at 512 bytes. This is too short for some of
5848 the long messages that one sometimes sees. I've increased the limit to
5851 31. It is now possible to make retry rules that apply only when a message has a
5852 specific sender, in particular, an empty sender.
5854 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5855 it possible to be selective about when SMTP synchronization is enforced.
5857 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5859 32. Implemented hosts_connection_nolog.
5861 33. Added an ACL for QUIT.
5863 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5866 35. Added mailbox_size and mailbox_filecount to appendfile.
5868 36. Added control = no_multiline_responses to ACLs.
5870 37. There was a bug in the logic of the code that waits for the clock to tick
5871 in the case where the clock went backwards by a substantial amount such
5872 that the microsecond fraction of "now" was more than the microsecond
5873 fraction of "then" (but the whole seconds number was less).
5875 38. Added support for the libradius Radius client library this is found on
5876 FreeBSD (previously only the radiusclient library was supported).
5882 1. When certain lookups returned multiple values in the form name=value, the
5883 quoting of the values was not always being done properly. Specifically:
5884 (a) If the value started with a double quote, but contained no whitespace,
5886 (b) If the value contained whitespace other than a space character (i.e.
5887 tabs or newlines or carriage returns) it was not quoted.
5888 This fix has been applied to the mysql and pgsql lookups by writing a
5889 separate quoting function and calling it from the lookup code. The fix
5890 should probably also be applied to nisplus, ibase and oracle lookups, but
5891 since I cannot test any of those, I have not disturbed their existing code.
5893 2. A hit in the callout cache for a specific address caused a log line with no
5894 reason for rejecting RCPT. Now it says "Previous (cached) callout
5895 verification failure".
5897 3. There was an off-by-one bug in the queryprogram router. An over-long
5898 return line was truncated at 256 instead of 255 characters, thereby
5899 overflowing its buffer with the terminating zero. As well as fixing this, I
5900 have increased the buffer size to 1024 (and made a note to document this).
5902 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5903 when Exim is waiting for an SMTP response from a remote server, Exim
5904 restarts its select() call on the socket, thereby resetting its timeout.
5905 This is not a problem when such interrupts are rare. Somebody set up a cron
5906 job to run exiwhat every 2 minutes, which is less than the normal select()
5907 timeout (5 or 10 minutes). This meant that the select() timeout never
5908 kicked in because it was always reset. I have fixed this by comparing the
5909 time when an interrupt arrives with the time at the start of the first call
5910 to select(). If more time than the timeout has elapsed, the interrupt is
5911 treated as a timeout.
5913 5. Some internal re-factoring in preparation for the addition of Sieve
5914 extensions (by MH). In particular, the "personal" test is moved to a
5915 separate function, and given an option for scanning Cc: and Bcc: (which is
5916 not set for Exim filters).
5918 6. When Exim created an email address using the login of the caller as the
5919 local part (e.g. when creating a From: or Sender: header line), it was not
5920 quoting the local part when it contained special characters such as @.
5922 7. Installed new OpenBSD configuration files.
5924 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5925 try to make them clearer.
5927 9. Callout options, other than the timeout value, were being ignored when
5928 verifying sender addresses in header lines. For example, when using
5930 verify = header_sender/callout=no_cache
5932 the cache was (incorrectly) being used.
5934 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5935 only the Cygwin environment.
5937 11. When return_path_on_delivery was set as a log selector, if different remote
5938 addresses in the same message used different return paths and parallel
5939 remote delivery occurred, the wrong values would sometimes be logged.
5940 (Whenever a remote delivery process finished, the return path value from
5941 the most recently started remote delivery process was logged.)
5943 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5944 lines when AUTH and/or TLS are in use. This is the "received protocol"
5945 field. Exim used to use "asmtp" for authenticated SMTP, without any
5946 indication (in the protocol name) for TLS use. Now it follows the RFC and
5947 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5948 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5949 names appear in log lines as well as in Received: header lines.
5951 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5952 extensions, and comparison tests, and to fix some bugs.
5954 14. Changes to the "personal" filter test:
5956 (1) The test was buggy in that it was just doing the equivalent of
5957 "contains" tests on header lines. For example, if a user's address was
5958 anne@some.where, the "personal" test would incorrectly be true for
5960 To: susanne@some.where
5962 This test is now done by extracting each address from the header in turn,
5963 and checking the entire address. Other tests that are part of "personal"
5964 are now done using regular expressions (for example, to check local parts
5965 of addresses in From: header lines).
5967 (2) The list of non-personal local parts in From: addresses has been
5968 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5969 taken from the Sieve specification recommendations.
5971 (3) If the message contains any header line starting with "List-" it is
5972 treated as non-personal.
5974 (4) The test for "circular" in the Subject: header line has been removed
5975 because it now seems ill-conceived.
5977 15. Minor typos in src/EDITME comments corrected.
5979 16. Installed latest exipick from John Jetmore.
5981 17. If headers_add on a router specified a text string that was too long for
5982 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5983 of string_sprintf() is now avoided.
5985 18. $message_body_size was not set (it was always zero) when running the DATA
5986 ACL and the local_scan() function.
5988 19. For the "mail" command in an Exim filter, no default was being set for
5989 the once_repeat time, causing a random time value to be used if "once" was
5990 specified. (If the value happened to be <= 0, no repeat happened.) The
5991 default is now 0s, meaning "never repeat". The "vacation" command was OK
5992 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5993 (I found it when inspecting the code).
5995 20. There is now an overall timeout for performing a callout verification. It
5996 defaults to 4 times the callout timeout, which applies to individual SMTP
5997 commands during the callout. The overall timeout applies when there is more
5998 than one host that can be tried. The timeout is checked before trying the
5999 next host. This prevents very long delays if there are a large number of
6000 hosts and all are timing out (e.g. when the network connections are timing
6001 out). The value of the overall timeout can be changed by specifying an
6002 additional sub-option for "callout", called "maxwait". For example:
6004 verify = sender/callout=5s,maxwait=20s
6006 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6007 to the end before writing, but this should make it even safer).
6009 22. Exim was forgetting that it had advertised PIPELINING for the second and
6010 subsequent messages on an SMTP connection. It was also not resetting its
6011 memory on STARTTLS and an internal HELO.
6013 23. When Exim logs an SMTP synchronization error within a session, it now
6014 records whether PIPELINING has been advertised or not.
6016 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6017 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6018 rather than long int.
6020 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6022 26. Added the never_mail option to autoreply.
6028 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6029 crash if the getsockname() call failed; this can happen if a connection is
6030 closed very soon after it is established. The problem was simply in the
6031 order in which certain operations were done, causing Exim to try to write
6032 to the SMTP stream before it had set up the file descriptor. The bug has
6033 been fixed by making things happen in the correct order.
6039 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6040 before the connection was closed, thus losing the rejection response.
6042 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6043 some early Solaris releases, but causes trouble in current releases where
6044 socklen_t is defined.
6046 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6049 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6052 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6053 that releasing the top of it at the end releases what was used for sub-
6054 expansions (unless the block got too big). However, discard this block if
6055 the first thing is a variable or header, so that we can use its block when
6056 it is dynamic (useful for very large $message_headers, for example).
6058 6. Lookups now cache *every* query, not just the most recent. A new, separate
6059 store pool is used for this. It can be recovered when all lookup caches are
6060 flushed. Lookups now release memory at the end of their result strings.
6061 This has involved some general refactoring of the lookup sources.
6063 7. Some code has been added to the store_xxx() functions to reduce the amount
6064 of flapping under certain conditions.
6066 8. log_incoming_interface used to affect only the <= reception log lines. Now
6067 it causes the local interface and port to be added to several more SMTP log
6068 lines, for example "SMTP connection from", and rejection lines.
6070 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6072 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6074 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6075 could be overwritten at the end of the current message (or the start of a
6076 new message if it was set in a HELO ACL). The value is now preserved for
6077 the duration of the SMTP connection.
6079 12. If a transport had a headers_rewrite setting, and a matching header line
6080 contained an unqualified address, that address was qualified, even if it
6081 did not match any rewriting rules. The underlying bug was that the values
6082 of the flags that permit the existence of unqualified sender and recipient
6083 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6084 non-local messages, and by -bnq for local messages) were not being
6085 preserved with the message after it was received.
6087 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6088 "next input=" as part of the text comprising the host identity instead of
6089 the correct text. The code was using the same buffer for two different
6090 strings. However, depending on which order the printing function evaluated
6091 its arguments, the bug did not always show up. Under Linux, for example, my
6092 test suite worked just fine.
6094 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6095 doesn't work with some older versions of Perl. It has been changed to "my",
6096 which in any case is probably the better facility to use.
6098 15. A really picky compiler found some instances of statements for creating
6099 error messages that either had too many or two few arguments for the format
6102 16. The size of the buffer for calls to the DNS resolver has been increased
6103 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6104 for addresses that have a lot of PTR records. This alleviates a problem; it
6105 does not fully solve it.
6107 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6108 buffer now truncates the list and logs the incident, which is the same
6109 action as happens when Exim is looking up a host name and its aliases.
6110 Previously in this situation something unpredictable would happen;
6111 sometimes it was "internal error: store_reset failed".
6113 18. If a server dropped the connection unexpectedly when an Exim client was
6114 using GnuTLS and trying to read a response, the client delivery process
6115 crashed while trying to generate an error log message.
6117 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6118 string, for example:
6120 warn message = H1: something\nH2: something
6122 the text was added as a single header line from Exim's point of view
6123 though it ended up OK in the delivered message. However, searching for the
6124 second and subsequent header lines using $h_h2: did not work. This has been
6125 fixed. Similarly, if a system filter added multiple headers in this way,
6126 the routers could not see them.
6128 20. Expanded the error message when iplsearch is called with an invalid key to
6129 suggest using net-iplsearch in a host list.
6131 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6132 ACLs are no longer actually imposed (and a message to that effect is
6135 22. If a "gecos" field in a passwd entry contained escaped characters, in
6136 particular, if it contained a \" sequence, Exim got it wrong when building
6137 a From: or a Sender: header from that name. A second bug also caused
6138 incorrect handling when an unquoted " was present following a character
6139 that needed quoting.
6141 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6142 was not being matched caselessly.
6144 24. Arranged for all hyphens in the exim.8 source to be escaped with
6147 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6148 properly. Recipient callout cache records were still being keyed to include
6149 the sender, even when use_sender was set false. This led to far more
6150 callouts that were necessary. The sender is no longer included in the key
6151 when use_sender is false.
6153 26. Added "control = submission" modifier to ACLs.
6155 27. Added the ${base62d: operator to decode base 62 numbers.
6157 28. dnsdb lookups can now access SRV records.
6159 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6160 the configuration file.
6162 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6163 (-v) mode. This makes the output for a verbose queue run more intelligible.
6165 31. Added a use_postmaster feature to recipient callouts.
6167 32. Added the $body_zerocount variable, containing the number of binary zero
6168 bytes in the message body.
6170 33. The time of last modification of the "new" subdirectory is now used as the
6171 "mailbox time last read" when there is a quota error for a maildir
6174 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6176 35. Added +ignore_unknown as a special item in host lists.
6178 36. Code for decoding IPv6 addresses in host lists is now included, even if
6179 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6180 address was recognized as an IP address, but was then not correctly decoded
6181 into binary, causing unexpected and incorrect effects when compared with
6188 1. Very minor rewording of debugging text in manualroute to say "list of
6189 hosts" instead of "hostlist".
6191 2. If verify=header_syntax was set, and a header line with an unqualified
6192 address (no domain) and a large number of spaces between the end of the
6193 name and the colon was received, the reception process suffered a buffer
6194 overflow, and (when I tested it) crashed. This was caused by some obsolete
6195 code that should have been removed. The fix is to remove it!
6197 3. When running in the test harness, delay a bit after writing a bounce
6198 message to get a bit more predictability in the log output.
6200 4. Added a call to search_tidyup() just before forking a reception process. In
6201 theory, someone could use a lookup in the expansion of smtp_accept_max_
6202 per_host which, without the tidyup, could leave open a database connection.
6204 5. Added the variables $recipient_data and $sender_data which get set from a
6205 lookup success in an ACL "recipients" or "senders" condition, or a router
6206 "senders" option, similar to $domain_data and $local_part_data.
6208 6. Moved the writing of debug_print from before to after the "senders" test
6211 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6212 problems for message scanning, either using a data ACL, or using
6213 local_scan() because the Received: header was not generated till after they
6214 were called (in order to set the time as the time of reception completion).
6215 I have revised the way this works. The header is now generated after the
6216 body is received, but before the ACL or local_scan() are called. After they
6217 are run, the timestamp in the header is updated.
6223 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6224 before starting a queue runner without re-exec. This happened only when
6225 deliver_drop_privilege was set or when the Exim user was set to root. The
6226 effect of the bug was that timeouts during subsequent deliveries caused
6227 crashes instead of being properly handled. The handler is now left at its
6228 default (and expected) setting.
6230 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6231 message, again when deliver_drop_privilege is set or Exim is run as root.
6232 The bug described in (1) was not present in this case, but the tidying up
6233 of the other signals was missing. I have made the two cases consistent.
6235 3. The ignore_target_hosts setting on a manualroute router was being ignored
6236 for hosts that were looked up using the /MX notation.
6238 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6241 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6242 operated on the sender address. After changing the $sender_address to <>
6243 for the sender address verify, Exim was re-instated it as the original
6244 (before rewriting) address, but remembering that it had rewritten it, so it
6245 wasn't rewriting it again. This bug also had the effect of breaking the
6246 sender address verification caching when the sender address was rewritten.
6248 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6249 This has been changed so that if the ip literal address matches
6250 ignore_target_hosts, the router declines.
6252 7. Added expansion conditions match_domain, match_address, and match_local_
6253 part (NOT match_host).
6255 8. The placeholder for the Received: header didn't have a length field set.
6257 9. Added code to Exim itself and to exim_lock to test for a specific race
6258 condition that could lead to file corruption when using MBX delivery. The
6259 issue is with the lockfile that is created in /tmp. If this file is removed
6260 after a process has opened it but before that process has acquired a lock,
6261 there is the potential for a second process to recreate the file and also
6262 acquire a lock. This could lead to two Exim processes writing to the file
6263 at the same time. The added code performs the same test as UW imapd; it
6264 checks after acquiring the lock that its file descriptor still refers to
6265 the same named file.
6267 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6268 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6271 11. Added the smtp_active_hostname option. If used, this will typically be made
6272 to depend on the incoming interface address. Because $interface_address is
6273 not set up until the daemon has forked a reception process, error responses
6274 that can happen earlier (such as "too many connections") no longer contain
6277 12. If an expansion in a condition on a "warn" statement fails because a lookup
6278 defers, the "warn" statement is abandoned, and the next ACL statement is
6279 processed. Previously this caused the whole ACL to be aborted.
6281 13. Added the iplsearch lookup type.
6283 14. Added ident_timeout as a log selector.
6285 15. Added tls_certificate_verified as a log selector.
6287 16. Added a global option tls_require_ciphers (compare the smtp transport
6288 option of the same name). This controls incoming TLS connections.
6290 17. I finally figured out how to make tls_require_ciphers do a similar thing
6291 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6292 before starting the TLS session.
6294 18. Tabs are now shown as \t in -bP output.
6296 19. If the log selector return_path_on_delivery was set, Exim crashed when
6297 bouncing a message because it had too many Received: header lines.
6299 20. If two routers both had headers_remove settings, and the first one included
6300 a superfluous trailing colon, the final name in the first list and the
6301 first name in the second list were incorrectly joined into one item (with a
6302 colon in the middle).
6308 1. Added -C and -D options to the exinext utility, mainly to make it easier
6309 to include in the automated testing, but these could be helpful when
6310 multiple configurations are in use.
6312 2. The exinext utility was not formatting the output nicely when there was
6313 an alternate port involved in the retry record key, nor when there was a
6314 message id as well (for retries that were specific to a specific message
6315 and a specific host). It was also confused by IPv6 addresses, because of
6316 the additional colons they contain. I have fixed the IPv4 problem, and
6317 patched it up to do a reasonable job for IPv6.
6319 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6320 delivery, the log line now contains "pipelined" if PIPELINING was used.
6322 4. An SMTP transport process used to panic and die if the bind() call to set
6323 an explicit outgoing interface failed. This has been changed; it is now
6324 treated in the same way as a connect() failure.
6326 5. A reference to $sender_host_name in the part of a conditional expansion
6327 that was being skipped was still causing a DNS lookup. This no longer
6330 6. The def: expansion condition was not recognizing references to header lines
6331 that used bh_ and bheader_.
6333 7. Added the _cache feature to named lists.
6335 8. The code for checking quota_filecount in the appendfile transport was
6336 allowing one more file than it should have been.
6338 9. For compatibility with Sendmail, the command line option
6346 and sets the incoming protocol and host name (for trusted callers). The
6347 host name and its colon can be omitted when only the protocol is to be set.
6348 Note the Exim already has two private options, -pd and -ps, that refer to
6349 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6350 "s", but I don't think that's a major issue.
6352 10. A number of refactoring changes to the code, none of which should affect
6355 (a) The number of logging options was getting close to filling up the
6356 32-bit word that was used as a bit map. I have split them into two classes:
6357 those that are passed in the argument to log_write(), and those that are
6358 only ever tested independently outside of that function. These are now in
6359 separate 32-bit words, so there is plenty of room for expansion again.
6360 There is no change in the user interface or the logging behaviour.
6362 (b) When building, for example, log lines, the code previously used a
6363 macro that called string_cat() twice, in order to add two strings. This is
6364 not really sufficiently general. Furthermore, there was one instance where
6365 it was actually wrong because one of the argument was used twice, and in
6366 one call a function was used. (As it happened, calling the function twice
6367 did not affect the overall behaviour.) The macro has been replaced by a
6368 function that can join an arbitrary number of extra strings onto a growing
6371 (c) The code for expansion conditions now uses a table and a binary chop
6372 instead of a serial search (which was left over from when there were very
6373 few conditions). Also, it now recognizes conditions like "pam" even when
6374 the relevant support is not compiled in: a suitably worded error message is
6375 given if an attempt is made to use such a condition.
6377 11. Added ${time_interval:xxxxx}.
6379 12. A bug was causing one of the ddress fields not to be passed back correctly
6380 from remote delivery subprocesses. The field in question was not being
6381 subsequently used, so this caused to problems in practice.
6383 13. Added new log selectors queue_time and deliver_time.
6385 14. Might have fixed a bug in maildirsizefile handling that threw up
6386 "unexpected character" debug warnings, and recalculated the data
6387 unnecessarily. In any case, I expanded the warning message to give more
6390 15. Added the message "Restricted characters in address" to the statements in
6391 the default ACL that block characters like @ and % in local parts.
6393 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6394 Three changes have been made:
6396 (a) There was a serious bug; a negative response to MAIL caused the whole
6397 recipient domain to be cached as invalid, thereby blocking all messages
6398 to all local parts at the same domain, from all senders. This bug has
6399 been fixed. The domain is no longer cached after a negative response to
6400 MAIL if the sender used is not empty.
6402 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6405 (c) A new callout option, "use_sender" has been added for people who want
6406 the modified behaviour.
6412 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6415 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6416 indeed breaks things for older releases.
6418 3. Added additional logging to the case where there is a problem reading data
6419 from a filter that is running in a subprocess using a pipe, in order to
6420 try to track down a specific problem.
6422 4. Testing facility fudge: when running in the test harness and attempting
6423 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6424 getting "No route to host". Convert this to a timeout.
6426 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6429 6. Some OS don't have socklen_t but use size_t instead. This affects the
6430 fifth argument of getsockopt() amongst other things. This is now
6431 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6432 can be set for individual OS. I have set it for SunOS5, OSF1, and
6433 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6434 some earlier ones do not.
6436 7. Change 4.30/15 was not doing the test caselessly.
6438 8. The standard form for an IPv6 address literal was being rejected by address
6439 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6440 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6441 this, as well as the form without the "IPv6" on the front (but only when
6442 address literals are enabled, of course).
6444 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6446 10. Exim crashed if a message with an empty sender address specified by -f
6447 encountered a router with an errors_to setting. This could be provoked only
6448 by a command such as
6452 where an empty string was supplied; "<>" did not hit this bug.
6454 11. Installed PCRE release 4.5.
6456 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6457 remained set. It is now erased.
6459 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6460 times from message ids (which are base 36 rather than the normal 62).
6462 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6463 were being counted as actual protocol errors, and logged if the log
6464 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6465 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6466 and DATA following a set of rejected RCPTs do not count as protocol errors.
6467 In other words, Exim assumes they were pipelined, though this may not
6468 actually be the case. Of course, in all cases the client gets an
6469 appropriate error code.
6471 15. If a lookup fails in an ACL condition, a message about the failure may
6472 be available; it is used if testing the ACL cannot continue, because most
6473 such messages specify what the cause of the deferral is. However, some
6474 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6475 that caused an old message to be retained and used if a later statement
6476 caused a defer, replacing the real cause of the deferral.
6478 16. If an IP address had so many PTR records that the DNS lookup buffer
6479 was not large enough to hold them, Exim could crash while trying to process
6480 the truncated data. It now detects and logs this case.
6482 17. Further to 4.21/58, another change has been made: if (and only if) the
6483 first line of a message (the first header line) ends with CRLF, a bare LF
6484 in a subsequent header line has a space inserted after it, so as not to
6485 terminate the header.
6487 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6488 unnecessarily, used atoi() instead of strtol(), and didn't check the
6489 termination when getting file sizes from file names by regex.
6491 19. Completely re-implemented the support for maildirsize files, in the light
6492 of a number of problems with the previous contributed implementation
6493 (4.30/29). In particular:
6495 . If the quota is zero, the maildirsize file is maintained, but no quota is
6498 . If the maildir directory does not exist, it is created before any attempt
6499 to write a maildirsize file.
6501 . The quota value in the file is just a cache; if the quota is changed in
6502 the transport, the new value overrides.
6504 . A regular expression is available for excluding directories from the
6507 20. The autoreply transport checks the characters in options that define the
6508 message's headers; it allows continued headers, but it was checking with
6509 isspace() after an embedded newline instead of explicitly looking for a
6512 21. If all the "regular" hosts to which an address was routed had passed their
6513 expiry times, and had not reached their retry times, the address was
6514 bounced, even if fallback hosts were defined. Now Exim should go on to try
6517 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6518 equivalent code in the SMTP transport. Some hosts send humungous responses
6519 to HELO/EHLO, more than 1024 it seems.
6521 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6522 gives compiler warnings in some environments. I've now done it "properly",
6525 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6526 (because of problems with the built-in one) was declared to return uschar *
6527 instead of char *, causing compiler failure.
6529 25. Fixed a file descriptor leak when processing alias/forward files.
6531 26. Fixed a minor format string issue in dbfn.c.
6533 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6535 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6536 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6539 29. When a sender address is verified, it is cached, to save repeating the test
6540 when there is more than one recipient in a message. However, when the
6541 verification involves a callout, it is possible for different callout
6542 options to be set for different recipients. It is too complicated to keep
6543 track of this in the cache, so now Exim always runs a verification when a
6544 callout is required, relying on the callout cache for the optimization.
6545 The overhead is duplication of the address routing, but this should not be
6548 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6549 to be verified with callout=postmaster, and the main callout worked but the
6550 postmaster check failed, the verification correctly failed. However, if a
6551 subsequent RCPT command asked for sender verification *without* the
6552 postmaster check, incorrect caching caused this verification also to fail,
6555 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6556 it was not caching the DNS options (qualify_single, search_parents) that
6557 were used when the lookup failed. A subsequent lookup with different
6558 options therefore always gave the same answer, though there were cases
6559 where it should not have. (Example: a "domains = !$mx_any" option on a
6560 dnslookup router: the "domains" option is always processed without any
6561 widening, but the router might have qualify_single set.) Now Exim uses the
6562 cached value only when the same options are set.
6564 32. Added John Jetmore's "exipick" utility to the distribution.
6566 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6567 than a timeout (e.g. a certificate is required, and is not provided), an
6568 Exim server now closes the connection immediately. Previously it waited for
6569 the client to close - but if the client is SSL, it seems that they each
6570 wait for each other, leading to a delay before one of them times out.
6572 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6573 maintained 0.8.x compatibility because I don't think many are using it, and
6574 it is clearly obsolete.
6576 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6579 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6580 client certificate was expired. A simple patch fixes this, though I don't
6581 understand the full logic of why the verify callback is called multiple
6584 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6585 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6586 which causes problems with some clients (such as the Certicom SSL Plus
6587 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6588 disables the coutermeasure allowing Eudora to connect."
6590 38. Exim was not checking that a write() to a log file succeeded. This could
6591 lead to Bad Things if a log got too big, in particular if it hit a file
6592 size limit. Exim now panics and dies if it cannot write to a log file, just
6593 as it does if it cannot open a log file.
6595 39. Modified OS/Makefile-Linux so that it now contains
6597 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6599 The two -D definitions ensure that Exim is compiled with large file
6600 support, which makes it possible to handle log files that are bigger than
6603 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6604 instance) a domain was checked against a named list that involved a lookup,
6605 causing $domain_data to be set, then another domain was checked against the
6606 same list, then the first domain was re-checked, the value of $domain_data
6607 after the final check could be wrong. In particular, if the second check
6608 failed, it could be set empty. This bug probably also applied to
6611 41. The strip_trailing_dot option was not being applied to the address given
6612 with the -f command-line option.
6614 42. The code for reading a message's header from the spool was incrementing
6615 $received_count, but never initializing it. This meant that the value was
6616 incorrect (doubled) while delivering a message in the same process in which
6617 it was received. In the most common configuration of Exim, this never
6618 happens - a fresh exec is done - but it can happen when
6619 deliver_drop_privilege is set.
6621 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6622 it now includes up to 150 characters of the unexpected data in the log
6625 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6626 and building data strings. The size of both of these buffers was 10 000
6627 bytes - far larger than anybody would *ever* want, thought I. Needless to
6628 say, somebody hit the limit. I have increased the maximum line length to
6629 20 000 and the maximum data length of concatenated lines to 100 000. I have
6630 also fixed two bugs, because there was no checking on these buffers. Tsk,
6631 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6632 buffer is too small.
6634 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6635 lsearch lookups. Now it does.
6637 46. When parsing a route_list item in a manualroute router, a fixed-length
6638 buffer was used for the list of hosts. I made this 1024 bytes long,
6639 thinking that nobody would ever have a list of hosts that long. Wrong.
6640 Somebody had a whole pile of complicated expansion conditions, and the
6641 string was silently truncated, leading to an expansion error. It turns out
6642 that it is easier to change to an unlimited length (owing to other changes
6643 that have happened since this code was originally written) than to build
6644 structure for giving a limitation error. The length of the item that
6645 expands into the list of hosts is now unlimited.
6647 47. The lsearch lookup could not handle data where the length of text line was
6648 more than 4095 characters. Such lines were truncated, leading to shortened
6649 data being returned. It should now handle lines of any length.
6651 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6652 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6655 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6656 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6657 no longer happens in comment lines. A list of replacements is now placed
6658 at the head of all of the source files, except those whose only change is
6659 to replace PERL_COMMAND in the very first #! line.
6661 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6662 messages on the queue, with a bottom-up merge sort, using code contributed
6663 by Michael Haardt. This should make operations like -bp somewhat faster on
6664 large queues. It won't affect queue runners, except when queue_run_in_order
6667 51. Installed eximstats 1.31 in the distribution.
6669 52. Added support for SRV lookups to the dnslookup router.
6671 53. If an ACL referred to $message_body or $message_body_end, the value was not
6672 reset for any messages that followed in the same SMTP session.
6674 54. The store-handling optimization for building very long strings was not
6675 differentiating between the different store pools. I don't think this
6676 actually made any difference in practice, but I've tidied it.
6678 55. While running the routers to verify a sender address, $sender_address
6679 was still set to the sender address. This is wrong, because when routing to
6680 send a bounce to the sender, it would be empty. Therefore, I have changed
6681 it so that, while verifying a sender address, $sender_address is set to <>.
6682 (There is no change to what happens when verifying a recipient address.)
6684 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6685 target A or AAAA records (if not already returned) without resetting the
6686 qualify_single or search_parents options of the DNS resolver. These are
6687 inappropriate in this case because the targets of MX and SRV records must
6688 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6689 target that, when qualified, matched something in the local domain. These
6690 two options are now turned off when doing these lookups.
6692 57. It seems that at least some releases of Reiserfs (which does not have the
6693 concept of a fixed number of inodes) returns zero and not -1 for the
6694 number of available inodes. This interacted badly with check_spool_inodes,
6695 which assumed that -1 was the "no such thing" setting. What I have done is
6696 to check that the total number of inodes is greater than zero before doing
6697 the test of how many are available.
6699 58. When a "warn" ACL statement has a log_message modifier, the message is
6700 remembered, and not repeated. This is to avoid a lot of repetition when a
6701 message has many recipients that cause the same warning to be written.
6702 However, Exim was preserving the list of already written lines for an
6703 entire SMTP session, which doesn't seem right. The memory is now reset if a
6704 new message is started.
6706 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6707 debugging output unless log_rewrite was also set.
6709 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6710 of (int)(handle) when we know that handle contains (void *)(-1).
6712 61. The Exim daemon panic-logs an error return when it closes the incoming
6713 connection. However "connection reset by peer" seems to be common, and
6714 isn't really an error worthy of noting specially, so that particular error
6717 62. When Exim is trying to find all the local interfaces, it used to panic and
6718 die if the ioctl to get the interface flags failed. However, it seems that
6719 on at least one OS (Solaris 9) it is possible to have an interface that is
6720 included in the list of interfaces, but for which you get a failure error
6721 for this call. This happens when the interface is not "plumbed" into a
6722 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6723 failure of the "get flags" call assumes that the interface is down.
6725 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6726 makes life easier for people who are doing arithmetic on fields extracted
6727 from dates, where you often get leading zeros that should not be
6728 interpreted as octal.
6730 64. Added qualify_domain to the redirect router, to override the global
6733 65. If a pathologically long header line contained very many addresses (the
6734 report of this problem mentioned 10 000) and each of them was rewritten,
6735 Exim could use up a very large amount of memory. (It kept on making new
6736 copies of the header line as it rewrote, and never released the old ones.)
6737 At the expense of a bit more processing, the header rewriting function has
6738 been changed so that it no longer eats memory in this way.
6740 66. The generation of the Received: header has been moved from the time that a
6741 message starts to be received, to the time that it finishes. The timestamp
6742 in the Received: header should now be very close to that of the <= log
6743 line. There are two side-effects of this change:
6745 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6746 logged header lines no longer include the local Received: line, because
6747 it has not yet been created. The same applies to a copy of the message
6748 that is returned to a non-SMTP sender when a message is rejected.
6750 (b) When a filter file is tested using -bf, no additional Received: header
6751 is added to the test message. After some thought, I decided that this
6754 This change does not affect the value of $received_for. It is still set
6755 after address rewriting, but before local_scan() is called.
6757 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6759 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6760 gave an unhelpful panic error message, and a defer error. I have managed to
6761 change this behaviour so that it now rejects any supplied certificate,
6762 which seems right, as the list of acceptable certificates is empty.
6764 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6765 gave an unhelpful defer error. I have not managed to make this reject any
6766 supplied certificates, but the error message it gives is "no certificate
6767 supplied", which is not helpful.
6769 70. exigrep's output now also includes lines that are not associated with any
6770 message, but which match the given pattern. Implemented by a patch from
6771 Martin Sluka, which also tidied up the Perl a bit.
6773 71. Recipient callout verification, like sender verification, was using <> in
6774 the MAIL FROM command. This isn't really the right thing, since the actual
6775 sender may affect whether the remote host accepts the recipient or not. I
6776 have changed it to use the actual sender in the callout; this means that
6777 the cache record is now keyed on a recipient/sender pair, not just the
6778 recipient address. There doesn't seem to be a real danger of callout loops,
6779 since a callout by the remote host to check the sender would use <>.
6780 [SEE ABOVE: changed after hitting problems.]
6782 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6783 temporary errors. However, in the case of such a code being given after
6784 the end of a data transmission (i.e. after ".") Exim was failing to write
6785 a retry record for the message. (Yes, there was some broken host that was
6786 actually sending 8xx at this point.)
6788 73. An unknown lookup type in a host list could cause Exim to panic-die when
6789 the list was checked. (An example that provoked this was putting <; in the
6790 middle of a list instead of at the start.) If this happened during a DATA
6791 ACL check, a -D file could be left lying around. This kind of configuration
6792 error no longer causes Exim to die; instead it causes a defer error. The
6793 incident is still logged to the main and panic logs.
6795 74. Buglet left over from Exim 3 conversion. The message "too many messages
6796 in one connection" was written to the rejectlog but not the mainlog, except
6797 when address rewriting (yes!) was being logged.
6799 75. Added write_rejectlog option.
6801 76. When a system filter was run not as root (that is, when system_filter_user
6802 was set), the values of the $n variables were not being returned to the
6803 main process; thus, they were not subsequently available in the $sn
6806 77. Added +return_path_on_delivery log selector.
6808 78. A connection timeout was being treated differently from recipients deferred
6809 when testing hosts_max_try with a message that was older than the host's
6810 retry timeout. (The host should not be counted, thus allowing all hosts to
6811 be tried at least once before bouncing.) This may have been the cause of an
6812 occasionally reported bug whereby a message would remain on the queue
6813 longer than the retry timeout, but would be bounced if a delivery was
6814 forced. I say "may" because I never totally pinned down the problem;
6815 setting up timeout/retry tests is difficult. See also the next item.
6817 79. The ultimate address timeout was not being applied to errors that involved
6818 a combination of host plus message (for example, a timeout on a MAIL
6819 command). When an address resolved to a number of possible hosts, and they
6820 were not all tried for each delivery (e.g. because of hosts_max_try), a
6821 message could remain on the queue longer than the retry timeout.
6823 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6826 81. Fixed an obscure SMTP outgoing bug which required at least the following
6827 conditions: (a) there was another message waiting for the same server;
6828 (b) the server returned 5xx to all RCPT commands in the first message so
6829 that the message was not completed; (c) the server dropped the connection
6830 or gave a negative response to the RSET that Exim sends to abort the
6831 transaction. The observed case was a dropped connection after DATA that had
6832 been sent in pipelining mode. That is, the server had advertised PIPELINING
6833 but was not implementing it correctly. The effect of the bug was incorrect
6834 behaviour, such as trying another host, and this could lead to a crash.
6840 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6841 and daemon.c were passed as pointers to ints; they should have been
6842 pointers to socklen_t variables (which are typically unsigned ints).
6844 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6847 3. Fixed a really odd bug that affected only the testing scheme; patching a
6848 certain fixed string in the binary changed the value of another string that
6849 happened to be identical to the end of the original first string.
6851 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6852 name", it returns that address as the IP address. On some operating
6853 systems (e.g. Solaris), it also passes back the IP address string as the
6854 "host name". However, on others (e.g. Linux), it passes back an empty
6855 string. Exim wasn't checking for this, and was changing the host name to an
6856 empty string, assuming it had been canonicalized.
6858 5. Although rare, it is permitted to have more than one PTR record for a given
6859 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6860 all the names associated with an address, because they do in Solaris.
6861 However, it seems that they do not in Linux for data that comes from the
6862 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6863 I found this out when I moved to a new Linux workstation and tried to run
6864 the Exim test suite.
6866 To get round this problem I have changed the code so that it now does its
6867 own call to the DNS to look up PTR records when searching for a host name.
6868 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6869 addresses that are only in /etc/hosts are still found.
6871 This behaviour is, however, controlled by an option called host_lookup_
6872 order, which defaults to "bydns:byaddr". If people want to use the other
6873 order, or indeed, just use one or the other means of lookup, they can
6874 specify it in this variable.
6876 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6877 some operating systems, this comes back from gethostbyaddr() as an empty
6878 string, and this is what Exim used to test for. However, it seems that in
6879 other systems, "." is yielded. Exim now tests for this case too.
6881 7. The values of check_spool_space and check_log_space are now held internally
6882 as a number of kilobytes instead of an absolute number of bytes. If a
6883 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6884 kilobyte. This means that much larger values can be stored.
6886 8. Exim monitor: an attempt to get the action menu when not actually pointing
6887 at a message produces an empty menu entitled "No message selected". This
6888 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6889 no entries in it ("Shell widget menu has zero width and/or height"). So I
6890 have added a single, blank menu entry in this case.
6892 9. Added ${quote_local_part.
6894 10. MIME decoding is now applied to the contents of Subject: header lines when
6897 11. Now that a reference to $sender_host_address automatically causes a reverse
6898 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6899 host lookup before query-style lookups in lists that might use this
6900 variable. This has therefore been abolished, and the "net-" prefix is no
6901 longer necessary for query-style lookups.
6903 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6904 have been a typo for LFLAGS, so it has been changed.
6906 13. The install script calls Exim with "-C /dev/null" in order to find the
6907 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6908 to be output. However, since Exim outputs its version number before the
6909 error, it didn't break the script. It just looked ugly. I fixed this by
6910 always allowing "-C /dev/null" if the caller is root.
6912 14. Ignore overlarge ACL variable number when reading spool file - insurance
6913 against a later release with more variables having written the file.
6915 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6916 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6917 this, as well as the form without the "IPv6" on the front.
6919 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6920 OS/Makefile-Darwin file.
6922 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6923 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6925 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6926 variable. However, some OS use ints for inodes. I've added cast to long int
6927 to get rid of the compiler warning.
6929 19. I had forgotten to lock out "/../" in configuration file names when
6930 ALT_CONFIG_PREFIX was set.
6932 20. Routers used for verification do not need to specify transports. However,
6933 if such a router generated a host list, and callout was configured, Exim
6934 crashed, because it could not find a port number from the (non-existent)
6935 transport. It now assumes port 25 in this circumstance.
6937 21. Added the -t option to exigrep.
6939 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6940 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6941 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6942 them is set, LOOKUP_LSEARCH is forced.
6944 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6946 24. Added sender and host information to the "rejected by local_scan()" log
6947 line; previously there was no indication of these.
6949 25. Added .include_if_exists.
6951 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6952 Linux. It turns out that not all file systems support this. Apparently some
6953 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6954 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6955 not supported on the file descriptor, is now ignored when Exim is trying to
6956 sync a directory. This applies only to Linux.
6958 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6960 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6961 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6964 29. Added support for maildirsize files from supplied patch (modified a bit).
6966 30. The use of :fail: followed by an empty string could lead Exim to respond to
6967 sender verification failures with (e.g.):
6969 550 Verification failed for <xxx>
6970 550 Sender verify failed
6972 where the first response line was missing the '-' that indicates it is not
6973 the final line of the response.
6975 31. The loop for finding the name of the user that called Exim had a hardwired
6976 limit of 10; it now uses the value of finduser_retries, which is used for
6977 all other user lookups.
6979 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6982 33. Exim was neglecting to zero errno before one call of strtol() when
6983 expanding a string and expecting an integer value. On some systems this
6984 resulted in spurious "integer overflow" errors. Also, it was casting the
6985 result into an int without checking.
6987 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6988 did not work. The code looks as if it has *never* worked, though it appears
6989 to have been documented since at least release 1.62. I have made it work.
6991 35. The "timeout_DNS" error in retry rules, also documented since at least
6992 1.62, also never worked. As it isn't clear exactly what this means, and
6993 clearly it isn't a major issue, I have abolished the feature by treating it
6994 as "timeout", and writing a warning to the main and panic logs.
6996 36. The display of retry rules for -brt wasn't always showing the error code
6999 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7000 timeout_connect_A, timeout_connect_MX.
7002 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7003 to the empty sender.
7005 39. The daemon was not analysing the content of -oX till after it had closed
7006 stderr and disconnected from the controlling terminal. This meant that any
7007 syntax errors were only noted on the panic log, and the return code from
7008 the command was 0. By re-arranging the code a little, I've made the
7009 decoding happen first, so such errors now appear on stderr, and the return
7010 code is 1. However, the actual setting up of the sockets still happens in
7011 the disconnected process, so errors there are still only recorded on the
7014 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7015 connections (as happens on some IP stacks) was logged at start up time as
7016 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7017 it from "IPv6 and IPv4", which means that two separate sockets are being
7020 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7021 says whether AF_INET or AF_INET6 was passed as an argument.
7023 42. Exiwhat output was messed up when time zones were included in log
7026 43. Exiwhat now gives more information about the daemon's listening ports,
7027 and whether -tls-on-connect was used.
7029 44. The "port" option of the smtp transport is now expanded.
7031 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7032 silently ignored. Now an error message is written to the main and panic
7035 46. There's a new ACL modifier called "logwrite" which writes to a log file
7036 as soon as it is encountered.
7038 47. Added $local_user_uid and $local_user_gid at routing time.
7040 48. Exim crashed when trying to verify a sender address that was being
7043 49. Exim was recognizing only a space character after ".include". It now also
7044 recognizes a tab character.
7046 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7047 extracting the relevant information from the specification. The man page no
7048 longer contains scrambled data for the -d option, and I've added a section
7049 at the front about calling Exim under different names.
7051 51. Added "extra_headers" argument to the "mail" command in filter files.
7053 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7056 53. Installed eximstats 1.29.
7058 54. Added transport_filter_timeout as a generic transport option.
7060 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7061 Cc: header lines. This was required by RFC 822, but it not required by RFC
7064 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7065 incoming messages that did not have them. Now it does so only if the
7066 message originates locally, that is, if there is no associated remote host
7067 address. When Resent- header lines are present, this applies to the Resent-
7068 lines rather than the non-Resent- lines.
7070 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7071 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7073 58. Messages for configuration errors now include the name of the main
7074 configuration file - useful now that there may be more than one file in a
7075 list (.included file names were always shown).
7077 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7078 for those rare installations that do not start the daemon as root or run it
7079 setuid root. I've cut out the call to initgroups() if the daemon is not
7082 60. The Exim user and group can now be bound into the binary as text strings
7083 that are looked up at the start of Exim's processing.
7085 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7087 62. Added $mailstore_basename variable.
7089 63. Installed patch to sieve.c from Michael Haardt.
7091 64. When Exim failed to open the panic log after failing to open the main log,
7092 the original message it was trying to log was written to stderr and debug
7093 output, but if they were not available (the usual case in production), it
7094 was lost. Now it is written to syslog before the two lines that record the
7095 failures to open the logs.
7097 65. Users' Exim filters run in subprocesses under the user's uid. It is
7098 possible for a "deliver" command or an alias in a "personal" command to
7099 provoke an address rewrite. If logging of address rewriting is configured,
7100 this fails because the process is not running as root or exim. There may be
7101 a better way of dealing with this, but for the moment (because 4.30 needs
7102 to be released), I have disabled address rewrite logging when running a
7103 filter in a non-root, non-exim process.
7109 1. The buildconfig auxiliary program wasn't quoting the value set for
7110 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7111 not defined. This bug was masked in 4.22 by the effect that was fixed in
7114 2. Some messages that were rejected after a message id was allocated were
7115 shown as "incomplete" by exigrep. It no longer does this for messages that
7116 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7118 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7119 have allow_domain_literals set, the ID did not get logged in the <= line.
7120 Domain literals are now always recognized in Message-ID: header lines.
7122 4. The first argument for a ${extract expansion item is the key name or field
7123 number. Leading and trailing spaces in this item were not being ignored,
7124 causing some misleading effects.
7126 5. When deliver_drop_privilege was set, single queue runner processes started
7127 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7128 same command in the process it spins off) were not dropping privilege.
7130 6. When the daemon running as "exim" started a queue runner, it always
7131 re-executed Exim in the spun-off process. This is a waste of effort when
7132 deliver_drop_privilege is set. The new process now just calls the
7133 queue-runner function directly.
7139 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7142 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7143 ignored. Though the use of -oP was forcing the writing of a pid file, it
7144 was always written to the default place.
7146 3. If the message "no IP address found for host xxxx" is generated during
7147 incoming verification, it is now followed by identification of the incoming
7148 connection (so you can more easily find what provoked it).
7150 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7152 5. Added some features to "harden" Exim a bit more against certain attacks:
7154 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7155 be put in Local/Makefile. This is like the never_users runtime option,
7156 but it cannot be overridden. The default setting is "root".
7158 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7159 prefix string with which any file named in a -C command line option
7162 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7163 is retained for -C and -D only if the caller of Exim is root. Without
7164 it, the exim user may also use -C and -D and retain privilege.
7166 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7167 command line option is disabled.
7169 6. Macro names set by the -D option must start with an upper case letter, just
7170 like macro names defined in the configuration file.
7172 7. Added "dereference=" facility to LDAP.
7174 8. Two instances of the typo "uknown" in the source files are fixed.
7176 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7177 the Configure-Makefile script screwed up while processing it.
7179 10. Incorporated PCRE 4.4.
7181 11. The SMTP synchronization check was not operating right at the start of an
7182 SMTP session. For example, it could not catch a HELO sent before the client
7183 waited for the greeting. There is now a check for outstanding input at the
7184 point when the greeting is written. Because of the duplex, asynchronous
7185 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7186 way, but not yet received, when the check is performed.
7188 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7189 on TCP/IP sockets, because this apparently causes some broken clients to
7192 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7193 unchanged) from the Cygwin maintainer.
7195 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7196 was supported instead of testing for mbx. Effectively a typo.
7198 15. The spa authenticator server code was not checking that the input it
7199 received was valid base64.
7201 16. The debug output line for the "set" modifier in ACLs was not showing the
7202 name of the variable that was being set.
7204 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7206 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7207 lookup on its own. Something else was needed to trigger the lookup. For
7208 example, a match in host_lookup or the need for a host name in a host list.
7209 Now, if $sender_host_name is referenced and the host name has not yet been
7210 looked up, a lookup is performed. If the lookup fails, the variable remains
7211 empty, and $host_lookup_failed is set to "1".
7213 19. Added "eqi" as a case-independent comparison operator.
7215 20. The saslauthd authentication condition could segfault if neither service
7216 nor realm was specified.
7218 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7219 error message that was logged was misleading, and incoming SMTP
7220 connections were dropped. The message is now more accurate, and temporary
7221 errors are given to SMTP connections.
7223 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7224 (except RSET) with a 421 error, until QUIT is received. However, it was
7225 failing to send a response to QUIT.
7227 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7228 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7229 ACL is now run after the helo_try_verify_hosts code.
7231 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7232 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7233 case-independent, so other case variants are also recognized). Apparently
7234 some systems use these upper case variants.
7236 25. If more than two messages were waiting for the same host, and a transport
7237 filter was specified for the transport, Exim sent two messages over the
7238 same TCP/IP connection, and then failed with "socket operation on non-
7239 socket" when it tried to send the third.
7241 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7243 27. The extern definition of crypt16() in expand.c was not being excluded when
7244 the OS had its own crypt16() function.
7246 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7247 as a preferred synonym for return_size_limit, both as an option and as an
7250 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7252 30. Changed the default configuration ACL to relax the local part checking rule
7253 for addresses that are not in any local domains. For these addresses,
7254 slashes and pipe symbols are allowed within local parts, but the sequence
7255 /../ is explicitly forbidden.
7257 31. SPA server authentication was not clearing the challenge buffer before
7260 32. log_message in a "warn" ACL statement was writing to the reject log as
7261 well as to the main log, which contradicts the documentation and doesn't
7262 seem right (because no rejection is happening). So I have stopped it.
7264 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7265 However, I am unable to do any testing of this.
7267 34. Fixed an infelicity in the appendfile transport. When checking directories
7268 for a mailbox, to see if any needed to be created, it was accidentally
7269 using path names with one or more superfluous leading slashes; tracing
7270 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7272 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7273 contents are added to the log line that is written for every discarded
7274 recipient. (Previously a log_message setting was ignored.)
7276 36. The ${quote: operator now quotes the string if it is empty.
7278 37. The install script runs exim in order to find its version number. If for
7279 some reason other than non-existence or emptiness, which it checks, it
7280 could not run './exim', it was installing it with an empty version number,
7281 i.e. as "exim-". This error state is now caught, and the installation is
7284 38. An argument was missing from the function that creates an error message
7285 when Exim fails to connect to the socket for saslauthd authentication.
7286 This could cause Exim to crash, or give a corrupted message.
7288 39. Added isip, isip4, and isip6 to ${if conditions.
7290 40. The ACL variables $acl_xx are now saved with the message, and can be
7291 accessed later in routers, transports, and filters.
7293 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7294 strings in the file are not string-expanded.
7296 42. If a MAIL command specified a SIZE value that was too large to fit into an
7297 int variable, the check against message_size_limit failed. Such values are
7298 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7299 day this will have to be increased, but I don't think I want to be around
7300 when emails are that large.
7307 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7308 iconv() is not standard in FreeBSD.
7310 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7311 IPv6 enabled. The observed symptom was a segmentation fault on return from
7312 the function os_common_find_running_interfaces() in src/os.c.
7314 3. In the check_special_case() function in daemon.c I had used "errno" as an
7315 argument name, which causes warnings on some systems. This was basically a
7316 typo, since it was named "eno" in the comments!
7318 4. The code that waits for the clock to tick (at a resolution of some fraction
7319 of a second) so as to ensure message-id uniqueness was always waiting for
7320 at least one whole tick, when it could have waited for less. [This is
7321 almost certainly not relevant at current processor speeds, where it is
7322 unlikely to ever wait at all. But we try to future-proof.]
7324 5. The function that sleeps for a time interval that includes fractions of a
7325 second contained a race. It did not block SIGALRM between setting the
7326 timer, and suspending (a couple of lines later). If the interval was short
7327 and the sigsuspend() was delayed until after it had expired, the suspension
7328 never ended. On busy systems this could lead to processes getting stuck for
7331 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7332 process, before it forks any delivery processes. The open lookup caching
7333 mechanism meant that the open file or database connection was passed into
7334 the delivery process. The problem was that delivery processes always tidy
7335 up cached lookup data. This could cause a problem for the next delivery
7336 process started by the queue runner, because the external queue runner
7337 process does not know about the closure. So the next delivery process
7338 still has data in the lookup cache. In the case of a file lookup, there was
7339 no problem because closing a file descriptor in a subprocess doesn't affect
7340 the parent. However, if the lookup was caching a connection to a database,
7341 the connection was closed, and the second delivery process was likely to
7342 see errors such as "PGSQL: query failed: server closed the connection
7343 unexpectedly". The problem has been fixed by closing all cached lookups
7344 in a queue runner before running a delivery process.
7346 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7347 seem to have the "const" qualifier which it has on other OS. I've
7350 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7351 *of the same type* (client or server) with the same public name that an
7352 error should be diagnosed.
7354 9. When Exim looked up a host name for an IP address, but failed to find the
7355 original IP address when looking up the host name (a safety check), it
7356 output the message "<ip address> does not match any IP for NULL", which was
7357 confusing, to say the least. The bug was that the host name should have
7358 appeared instead of "NULL".
7360 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7361 user that is built into the binary, and the -C or -D options is used, root
7362 privilege is dropped before the configuration file is read. In addition,
7363 logging is switched to stderr instead of the normal log files. If the
7364 configuration then re-defines the Exim user, the unprivileged environment
7365 is probably not what is expected, so Exim logs a panic warning message (but
7368 However, if deliver_drop_privilege is set, the unprivileged state may well
7369 be exactly what is intended, so the warning has been cut out in that case,
7370 and Exim is allowed to try to write to its normal log files.
7376 1. smtp_return_error_details was not giving details for temporary sender
7377 or receiver verification errors.
7379 2. Diagnose a configuration error if two authenticators have the same public
7382 3. Exim used not to create the message log file for a message until the first
7383 delivery attempt. This could be confusing when incoming messages were held
7384 for policy or load reasons. The message log file is now created at the time
7385 the message is received, and an initial "Received" line is written to it.
7387 4. The automatically generated man page for command line options had a minor
7388 bug that caused no ill effects; however, a more serious problem was that
7389 the procedure for building the man page automatically didn't always
7390 operate. Consequently, release 4.20 contains an out-of-date version. This
7391 shouldn't happen again.
7393 5. When building Exim with embedded Perl support, the script that builds the
7394 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7395 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7397 6. The freeze_tell option was not being used for messages that were frozen on
7398 arrival, either by an ACL or by local_scan().
7400 7. Added the smtp_incomplete_transaction log selector.
7402 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7403 was accepting AUTH without a new EHLO.
7405 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7406 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7409 10. Logging of TCP/IP connections (when configured) now happens in the main
7410 daemon process instead of the child process, so that the TCP/IP connection
7411 count is more accurate (but it can never be perfect).
7413 11. The use of "drop" in a nested ACL was not being handled correctly in the
7414 outer ACL. Now, if condition failure induced by the nested "drop" causes
7415 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7416 or "require"), the connection is dropped.
7418 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7419 that yield "discard" can now be used with an "accept" or a "discard" verb,
7420 but an error is generated for any others (because I can't see a useful way
7421 to define what should happen).
7423 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7424 are now processed in the same way as lines in the Exim configuration file.
7425 In particular, continuation lines are supported.
7427 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7429 15. Added -ti meaning -t -i.
7431 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7432 domains, and warn by logging if others are found.
7434 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7435 returned from GIFCONF when Exim is trying to find the list of interfaces on
7436 a host. The code in os.c has been modified to copy each ifreq to an aligned
7437 structure in all cases.
7439 Also, in some cases, the returned ifreq's were being copied to a 'struct
7440 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7441 means the last couple of bytes of an IPv6 address could be chopped if the
7442 ifreq contained only a normal sockaddr (14 bytes storage).
7444 18. Named domain lists were not supported in the hosts_treat_as_local option.
7445 An entry such as +xxxx was not recognized, and was treated as a literal
7448 19. Ensure that header lines added by a DATA ACL are included in the reject log
7449 if the ACL subsequently rejects the message.
7451 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7452 MD5 (which is deprecated).
7454 21. When testing a filter file using -bf, Exim was writing a message when it
7455 took the sender from a "From " line in the message, but it was not doing so
7456 when it took $return_path from a Return-Path: header line. It now does.
7458 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7459 with a valid header line field name (a series of printing characters
7460 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7462 23. Changed "disc" in the source to "disk" to conform to the documentation and
7463 the book and for uniformity.
7465 24. Ignore Sendmail's -Ooption=value command line item.
7467 25. When execve() failed while trying to run a command in a pipe transport,
7468 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7469 could be confused with a return value of 69 from the command itself. This
7470 has been changed to 127, the value the shell returns if it is asked to run
7471 a non-existent command. The wording for the related log line suggests a
7472 non-existent command as the problem.
7474 26. If received_header_text expands to an empty string, do not add a Received:
7475 header line to the message. (Well, it adds a token one on the spool, but
7476 marks it "old" so that it doesn't get used or transmitted.)
7478 27. Installed eximstats 1.28 (addition of -nt option).
7480 28. There was no check for failure on the call to getsockname() in the daemon
7481 code. This can fail if there is a shortage of resources on the system, with
7482 ENOMEM, for example. A temporary error is now given on failure.
7484 29. Contrary to the C standard, it seems that in some environments, the
7485 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7486 program. Exim now does this explicitly; it affects the formatting of
7487 timestamps using strftime().
7489 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7490 complaints. I've now initialized all the variables, to avoid this.
7492 32. Header lines added by a system filter were not being "seen" during
7493 transport-time rewrites.
7495 33. The info_callback() function passed to OpenSSL is set up with type void
7496 (*)(SSL *, int, int), as described somewhere. However, when calling the
7497 function (actually a macro) that sets it up, the type void(*)() is
7498 expected. I've put in a cast to prevent warnings from picky compilers.
7500 34. If a DNS black list lookup found a CNAME record, but there were no A
7501 records associated with the domain it pointed at, Exim crashed.
7503 35. If a DNS black list lookup returned more than one A record, Exim ignored
7504 all but the first. It now scans all returned addresses if a particular IP
7505 value is being sought. In this situation, the contents of the
7506 $dnslist_value variable are a list of all the addresses, separated by a
7509 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7510 to accept a host name and all its aliases if the forward lookup for any of
7511 them yielded the IP address of the incoming connection. Now it accepts only
7512 those names whose forward lookup yields the correct IP address. Any other
7513 names are discarded. This closes a loophole whereby a rogue DNS
7514 administrator could create reverse DNS records to break through a
7515 wildcarded host restriction in an ACL.
7517 37. If a user filter or a system filter that ran in a subprocess used any of
7518 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7519 the wrong values were passed to the pipe command ($thisaddress had the
7520 value of $0, $0 had the value of $1, etc). This bug was introduced by
7521 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7523 38. Improved the line breaking for long SMTP error messages from ACLs.
7524 Previously, if there was no break point between 40 and 75 characters, Exim
7525 left the rest of the message alone. Two changes have been made: (a) I've
7526 reduced the minimum length to 35 characters; (b) if it can't find a break
7527 point between 35 and 75 characters, it looks ahead and uses the first one
7528 that it finds. This may give the occasional overlong line, but at least the
7529 remaining text gets split now.
7531 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7532 file descriptors might be low, and that setting 1000 would always raise it.
7533 It turns out that in some environments, the limit is already over 1000 and
7534 that lowering it causes trouble. So now Exim takes care not to decrease it.
7536 40. When delivering a message, the value of $return_path is set to $sender_
7537 address at the start of routing (routers may change the value). By an
7538 oversight, this default was not being set up when an address was tested by
7539 -bt or -bv, which affected the outcome if any router or filter referred to
7542 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7543 log only when "message" or "log_message" are set. However, if one of the
7544 conditions was an address verification, or a call to a nested ACL, the
7545 messages generated by the underlying test were being passed through. This
7546 no longer happens. The underlying message is available in $acl_verify_
7547 message for both "message" and "log_message" expansions, so it can be
7548 passed through if needed.
7550 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7551 new expansion $bh_ to give the encoded byte string without charset
7552 translation. Translation happens only if iconv() is available; HAVE_ICONV
7553 indicates this at build time. HEADERS_CHARSET gives the charset to
7554 translate to; headers_charset can change it in the configuration, and
7555 "headers charset" can change it in an individual filter file.
7557 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7558 that creates RFC 2047 encoded "words" labels them as that charset instead
7559 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7560 expansion operator; (ii) when Exim creates a From: line for a local
7561 message; (iii) when a header line is rewritten to include a "phrase" part.
7563 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7564 buggy, causing it to skip the first lines of messages whose message ID
7565 ended in 'D'. This would not have bitten before Exim release 4.14, because
7566 message IDs were unlikely to end in 'D' before then. The effect was to have
7567 incorrect size information for certain domains.
7569 45. #include "config.h" was missing at the start of the crypt16.c module. This
7570 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7573 46. If there was a timeout during a "random" callout check, Exim treated it as
7574 a failure of the random address, and carried on sending RSET and the real
7575 address. If the delay was just some slowness somewhere, the response to the
7576 original RCPT would be taken as a response to RSET and so on, causing
7577 mayhem of various kinds.
7579 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7580 when I implemented it. It didn't allow for the fact that some option values
7581 may legitimately be negative (e.g. size_addition), and it didn't even do
7582 the right test for positive values.
7584 48. Domain names in DNS records are case-independent. Exim always looks them up
7585 in lower case. Some resolvers return domain names in exactly the case they
7586 appear in the zone file, that is, they may contain uppercase letters. Not
7587 all resolvers do this - some return always lower case. Exim was treating a
7588 change of case by a resolver as a change of domain, similar to a widening
7589 of a domain abbreviation. This triggered its re-routing code and so it was
7590 trying to route what was effectively the same domain again. This normally
7591 caused routing to fail (because the router wouldn't handle the domain
7592 twice). Now Exim checks for this case specially, and just changes the
7593 casing of the domain that it ultimately uses when it transmits the message
7596 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7599 50. If a filter generated a file delivery with a non-absolute name (possible if
7600 no home directory exists for the router), the forbid_file option was not
7603 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7604 the existing equality matching.
7606 52. Exim was using ints instead of ino_t variables in some places where it was
7607 dealing with inode numbers.
7609 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7610 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7611 called TMPDIR, and if it finds it is different, it changes its value.
7613 54. The smtp_printf() function is now made available to local_scan() so
7614 additional output lines can be written before returning. There is also an
7615 smtp_fflush() function to enable the detection of a dropped connection.
7616 The variables smtp_input and smtp_batched_input are exported to
7619 55. Changed the default runtime configuration: the message "Unknown user"
7620 has been removed from the ACL, and instead placed on the localuser router,
7621 using the cannot_route_message feature. This means that any verification
7622 failures that generate their own messages won't get overridden. Similarly,
7623 the "Unrouteable address" message that was in the ACL for unverifiable
7624 relay addresses has also been removed.
7626 56. Added hosts_avoid_esmtp to the smtp transport.
7628 57. The exicyclog script was not checking for the esoteric option
7629 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7630 will work only if exicyclog is run under the appropriate euid.
7632 58. Following a discussion on the list, the rules by which Exim recognises line
7633 endings on incoming messages have been changed. The -dropcr and drop_cr
7634 options are now no-ops, retained only for backwards compatibility. The
7635 following line terminators are recognized: LF CRLF CR. However, special
7636 processing applies to CR:
7638 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7639 nor a local message in the state where . is a terminator.
7641 (ii) If a bare CR is encountered in a header line, an extra space is added
7642 after the line terminator so as not to end the header. The reasoning
7643 behind this is that bare CRs in header lines are most likely either
7644 to be mistakes, or people trying to play silly games.
7646 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7647 was being incorrectly given as 18 bytes larger than it should have been.
7648 This is a VOB (very old bug).
7650 60. This may never have affected anything current, but just in case it has:
7651 When the local host is found other than at the start of a list of hosts,
7652 the local host, those with the same MX, and any that follow, are discarded.
7653 When the list in question was part of a longer list of hosts, the following
7654 hosts (not currently being processed) were also being discarded. This no
7655 longer happens. I'm not sure if this situation could ever has previously
7658 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7661 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7662 header. This is something that is recommended in a new Internet Draft, and
7663 is something that is documented as being done by Sendmail. There are two
7664 possible values. For messages generated by the autoreply transport, Exim
7667 Auto-Submitted: auto-replied
7669 whereas for all other generated messages (e.g. bounces) it adds
7671 Auto-Submitted: auto-generated
7673 63. The "personal" condition in filters now includes a test for the
7674 Auto-Submitted: header. If it contains the string "auto-" the message it
7675 not considered personal.
7677 64. Added rcpt_include_affixes as a generic transport option.
7679 65. Added queue_only_override (default true).
7681 66. Added the syslog_duplication option.
7683 67. If what should have been the first header line of a message consisted of
7684 a space followed by a colon, Exim was mis-interpreting it as a header line.
7685 It isn't of course - it is syntactically invalid and should therefore be
7686 treated as the start of the message body. The misbehaviour could have
7687 caused a number of strange effects, including loss of data in subsequent
7688 header lines, and spool format errors.
7690 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7691 client host had authenticated. This control can now be exercised by an ACL
7692 for more flexibility.
7694 69. By default, callouts do not happen when testing with -bh. There is now a
7695 variant, -bhc, which does actually run the callout code, including
7696 consulting and updating the callout cache.
7698 70. Added support for saslauthd authentication, courtesy of Alexander
7701 71. If statvfs() failed on the spool or log directories while checking their
7702 size for availability, Exim confusingly gave the error "space shortage".
7703 Furthermore, in debugging mode it crashed with a floating point exception.
7704 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7705 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7706 serious problem, Exim now writes to the main and panic logs when this
7707 happens, with details of the failure. It then refuses to accept the
7708 incoming message, giving the message "spool directory problem" or "log
7709 directory problem" with a 421 code for SMTP messages.
7711 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7712 0, 1, and 2 exist, because some OS complain for execs without them (see
7713 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7714 descriptors. However, the code omitted to check that the open succeeded,
7715 causing mysterious errors if for some reason the permissions on /dev/null
7716 got screwed. Now Exim writes a message to the main and panic logs, and
7717 bombs out if it can't open /dev/null.
7719 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7720 interact so that it is all more flexible. It is supposed to remain
7721 backwards compatible. Also added extra_local_interfaces.
7723 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7724 to bomb out with an assertion failure - to the client this appears as a
7725 connection drop. This problem occurs in the part of the code that was taken
7726 from the Samba project. Fortunately, the assertion is in a very simple
7727 function, so I have fixed this by reproducing the function inline in the
7728 one place where it is called, and arranging for authentication to fail
7729 instead of killing the process with assert().
7731 75. The SPA client code was not working when the server requested OEM rather
7732 than Unicode encoding.
7734 76. Added code to make require_files with a specific uid setting more usable in
7735 the case where statting the file as root fails - usually a non-root-mounted
7736 NFS file system. When this happens and the failure is EACCES, Exim now
7737 forks a subprocess and does the per-uid checking as the relevant uid.
7739 77. Added process_log_path.
7741 78. If log_file_path was not explicitly set, a setting of check_log_space or
7742 check_log_inodes was ignored.
7744 79. If a space check for the spool or log partitions fails, the incident is now
7745 logged. Of course, in the latter case the data may get lost...
7747 80. Added the %p formatting code to string_format() so that it can be used to
7748 print addresses in debug_print(). Adjusted all the address printing in the
7749 debugging in store.c to use %p rather than %d.
7751 81. There was a concern that a line of code in smtp_in.c could overflow a
7752 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7753 initially expressed, the concern was not well-founded, because trailing
7754 spaces are removed early. However, if the trailing spaces were followed by
7755 a NULL, they did not get removed, so the overflow was possible. Two fixes
7758 (a) I re-wrote the offending code in a cleaner fashion.
7759 (b) If an incoming SMTP command contains a NULL character, it is rejected
7762 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7763 runs initgroups(), so that if the Exim user is in any additional groups,
7764 they will be used during message reception.
7770 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob