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.
6 Since Exim version 4.94
7 -----------------------
9 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
10 as arguments, so an implementation trying to copy these into a local
11 buffer was taking a taint-enforcement trap. Fix by using dynamically
12 created buffers. Similar fix for radius expansion condition.
14 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
15 reasonable, eg. to count headers. Fix by using dynamically created
16 buffers rather than a local. Do similar fixes for ACL actions "dcc",
17 "log_reject_target", "malware" and "spam"; the arguments are expanded
18 so could be handling tainted values.
20 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
21 broken the (no-op) support for this sendmail command. Restore it
22 to doing nothing, silently, and returning good status.
24 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
25 record path was given (or the default used) without a leading directory
26 path, an error occurred on trying to open it. Use the transport's working
29 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
30 transport. Previously it was the name on the DNS A-record; use instead
31 the head of the CNAME chain leading there (if there is one). This seems
32 to align better with RFC 6125.
34 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
35 smtp_accept_max_per_host allocated resources which were not released
36 when the limit was exceeded. This eventually crashed the daemon. Fix
37 by adding a relase action in that path.
39 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
40 expanded; previously using tainted values was rejected. Fix by using
41 dynamically-created buffers.
43 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
44 Previously a macro used one argument twice; when called with the
45 argument as an expression having side-effects, incorrect operation
46 resulted. Use an inlineable function.
48 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
49 held open for a verify callout. Previously this wan not accounted for
50 and a corrupt onward SMTP conversation resulted.
52 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
53 excluded, not matching the documentation.
59 JH/01 Avoid costly startup code when not strictly needed. This reduces time
60 for some exim process initialisations. It does mean that the logging
61 of TLS configuration problems is only done for the daemon startup.
63 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
65 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
66 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
69 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
70 DKIM signing is being done. Previously a transport_filter would always
71 disable CHUNKING, falling back to traditional DATA.
73 JH/05 Regard command-line receipients as tainted.
75 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
77 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
78 PAM library frees one of the arguments given to it, despite the
79 documentation. Therefore a plain malloc must be used.
81 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
82 on-stack buffers were used, resulting in a taint trap when DSN information
83 copied from a received message was written into the buffer.
85 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
86 the ordering of its ARC headers. This caused a crash.
88 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
89 a new record was being constructed with information from the peer, a trap
92 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
93 installation would get error messages from DMARC verify, when it hit the
94 nonexistent file indicated by the default. Distros wanting DMARC enabled
95 should both provide the file and set the option.
96 Also enforce no DMARC verification for command-line sourced messages.
98 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
99 could, depending on the platform, hang at the STARTTLS response.
101 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
102 message on a connection. Previously if one message had ARC headers and
103 the following one did not, a crash could result when adding an
104 Authentication-Results: header.
106 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
107 Exim main code and Exim-related utities. The introduction of taint
108 tracking also did many adjustments to string handling. Since then, eximon
109 frequently terminated with an assert failure.
111 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
112 check for 452 responses. This slightly helps the inefficieny of doing
113 a large alias-expansion into a recipient-limited target. The max_rcpt
114 transport option still applies (and at the current default, will override
115 the new feature). The check is done for either cause of synch, and forces
116 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
117 connection. The new facility is not tunable at this time.
119 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
120 library live data was being used, so the results became garbage. Make
121 copies while it is still usable.
123 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
124 on delivery deferred (==) and failed (**) lines (if a delivery was
125 attemtped). Previously it was only on completion (=>) lines.
127 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
128 for the expansion of the server_scram_iter and server_scram_salt options.
130 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
131 are now specifically given a NO_DATA response without hitting the system
132 resolver. The library goes on to do the now-standard TXT lookup.
133 Use of dnsdb lookups is not affected.
135 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
136 only retrieve the errormessage once. Previously two calls to dlerror()
137 were used, and the second one (for mainlog/paniclog) retrieved null
140 JH/20 Taint checking: disallow use of tainted data for
141 - the appendfile transport file and directory options
142 - the pipe transport command
143 - the autoreply transport file, log and once options
144 - file names used by the redirect router (including filter files)
146 - paths used by single-key lookups
147 Previously this was permitted.
149 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
150 adjusted the size of a major service buffer; this failed because the
151 buffer was in use at the time. Change to a compile-time increase in the
152 buffer size, when this authenticator is compiled into exim.
154 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
155 previous fast-mode was untenable in the face of glibs using mmap to
156 support larger malloc requests.
158 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
159 New values supported, if defined on system where compiled:
160 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
161 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
163 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
164 running a limited number of proceses in parallel, a benefit is gained. The
165 amount varies with the platform hardware and load. The use of the option
166 queue_run_in_order means we cannot do this, as ordering becomes
169 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
170 had introduced a string-copy (for ensuring NUL-termination) which was not
171 appropriate for that case, which can include embedded NUL bytes in the
172 block of data. Investigation showed the copy to actually be needless, the
173 data being length-specified.
175 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
176 done during a receiving connection, and both used TLS, global info was
177 used rather than per-connection info for tracking the state of data
178 queued for transmission. This could result in a connection hang.
180 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
181 Previously, when delivering serveral messages down a single connection
182 only the first would provide a SIZE. This was due to the size information
183 not being properly tracked.
185 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
186 TAI (at 37 seconds currently), pretend to be in UTC for time-related
187 expansion and logging. Previously, spurious values such as a future
188 minute could be seen.
190 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
191 it could crash from a null-deref. This could also affect the
192 ${addresses: } operator and ${readsock } item.
194 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
195 message following a mime one, the variable was not reset.
197 JH/30 When an pipelined-connect fails at the first response, assume incorrect
198 cached capability (perhaps the peer reneged?) and immediately retry in
201 JH/31 Fix spurious detection of timeout while writing to transport filter.
203 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
204 an attempt to copy the string was made before checking it.
206 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
207 taint of the lookup key was maintained; we now regard the presence in the
208 filesystem as sufficient validation.
210 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
211 argument is supplied.
213 JH/35 The dsearch lookup now requires that the directory is an absolute path.
214 Previously this was not checked, and nonempty relative paths made an
215 access under Exim's current working directory.
217 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
218 Previously no event was raised.
220 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
221 parameter supplied by the sender MAIL FROM command. Previously it was
222 ignored, and only the check_spool_space option value for the required
225 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
226 the size of the signing public-key. Previously it was instead giving
227 the size of the signature hash.
229 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
230 the default. See the (new) dkim_verify_min_keysizes option.
232 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
233 and an ACL use a lookup for checking either the local_part or domain,
234 stale data could be accessed. Ensure that variable references are
235 dropped between messages.
237 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
238 by the client was not checked as pointing within response data before
239 being used. A malicious client could thus cause an out-of-bounds read and
240 possibly gain authentication. Fix by adding the check.
242 JH/42 Internationalisation: change the default for downconversion in the smtp
243 transport to be "if needed". Previously it was "as previously set" for
244 the message, which usually meant "if needed" for message-submission but
245 "no" for everything else. However, MTAs have been seen using SMTPUTF8
246 even when the envelope addresses did not need it, resulting in forwarding
247 failures to non-supporting MTAs. A downconvert in such cases will be
248 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
249 transport. The change does mean that addresses needing conversion will
250 be converted when previously a delivery failure would occur.
252 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
253 response was received it would be used unchecked in a fail-DSN, violating
254 standards on line-length limits. Truncate if needed.
256 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
257 doesn't work. (Additionally add a new main config option to configure the
264 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
265 side, to decode a TLS 1.3 packet capture.
267 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
268 Previously the default library behaviour applied, sending two, each in
271 JH/03 Debug output for ACL now gives the config file name and line number for
274 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
276 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
278 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
279 buffer overrun for (non-chunking) other transports.
281 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
282 TLS1.3, means that a server rejecting a client certificate is not visible
283 to the client until the first read of encrypted data (typically the
284 response to EHLO). Add detection for that case and treat it as a failed
285 TLS connection attempt, so that the normal retry-in-clear can work (if
286 suitably configured).
288 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
289 and/or domain. Found and fixed by Jason Betts.
291 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
292 configuration). If a CNAME target was not a wellformed name pattern, a
295 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
296 the OS reports them interleaved with other addresses.
298 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
299 used both for input and for a verify callout, both encrypted, SMTP
300 responses being sent by the server could be lost. This resulted in
301 dropped connections and sometimes bounces generated by a peer sending
304 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
305 string. Previously it was possible to cause undefined behaviour in a
306 library routine (usually a crash). Found by "zerons".
308 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
311 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
312 API was removed, so update to use the newer ones.
314 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
315 any timeout set, is taking a long time. Previously we would hang on to a
316 rotated logfile "forever" if the input was arriving with long gaps
317 (a previous attempt to fix addressed lack, for a long time, of initial
320 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
321 shared (NFS) environment. The length of the tempfile name is now
322 4 + 16 ("hdr.$message_exim_id") which might break on file
323 systems which restrict the file name length to lower values.
326 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
327 shared (NFS) environment.
329 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
330 did for all versions <4.90). Notably -M, -m, --invert, -I may be
333 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
334 on some platforms for bit 31.
336 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
337 to changes apparently associated with TLS1.3 handling some of the APIs
338 previously used were either nonfunctional or inappropriate. Strings
339 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
340 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
341 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
342 This affects log line X= elements, the $tls_{in,out}_cipher variables,
343 and the use of specific cipher names in the encrypted= ACL condition.
345 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
347 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
348 verification result was not updated unless hosts_require_ocsp applied.
350 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
351 queue_list_requires_admin set to false, non-admin users were denied the
354 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
355 directory-of-certs mode. Previously they were advertised despite the
358 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
359 A single TCP connection by a client will now hold a TLS connection open
360 for multiple message deliveries, by default. Previously the default was to
363 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
364 default. If built with the facility, DANE will be used. The facility
365 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
367 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
368 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
369 must be defined and you must still, unless you define DISABLE_TLS, manage
370 the the include-dir and library-file requirements that go with that
371 choice. Non-TLS builds are still supported.
373 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
376 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
377 default. If the platform supports and has the facility enabled, it will
378 be requested on all coneections.
380 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
381 controlled by the build-time option SUPPORT_PIPE_CONNECT.
383 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
385 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
386 success-DSN messages. Previously the From: header was always the default
387 one for these; the option was ignored.
389 JH/28 Fix the timeout on smtp response to apply to the whole response.
390 Previously it was reset for every read, so a teergrubing peer sending
391 single bytes within the time limit could extend the connection for a
392 long time. Credit to Qualsys Security Advisory Team for the discovery.
394 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
395 delivery address, which leaked information of the results of local
396 forwarding. Change to the original envelope recipient address, per
399 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
400 requested. Previously not bounce was generated and a log entry of
401 error ignored was made.
403 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
405 JH/32 Introduce a general tainting mechanism for values read from the input
406 channel, and values derived from them. Refuse to expand any tainted
407 values, to catch one form of exploit.
409 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
410 was unused and the unexpanded text used for the test. Found and
411 fixed by Ruben Jenster.
413 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
414 an attempt to use a TLS library read routine dereffed a nul pointer,
417 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
418 them in our smtp response.
420 JH/36 Have the general router option retry_use_local_part default to true when
421 any of the restrictive preconditions are set (to anything). Previously it
422 was only for check_local user. The change removes one item of manual
423 configuration which is required for proper retries when a remote router
424 handles a subset of addresses for a domain.
426 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
427 link count into consideration.
429 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
430 caused the extension of big_buffer, the following lines were ignored.
432 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
433 accordance with RFC 2308. Previously there was no expiry, so a longlived
434 receive process (eg. due to ACL delays) versus a short SOA value could
437 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
439 JH/39 Promote DMARC support to mainline.
441 JH/40 Bug 2452: Add a References: header to DSNs.
443 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
444 parameters. The relevant library call is documented as "Deprecated: This
445 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
446 3.6.0, DH parameters are negotiated following RFC7919."
448 HS/06 Change the default of dnssec_request_domains to "*"
450 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
451 carried on and emitted a BDAT command, even when PIPELINING was not
454 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
455 buffer was used for the filename, resulting in a trap when tainted
456 arguments (eg. $domain) were used.
458 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
459 recommended to avoid a possible server-load attack. The feature can be
460 re-enabled via the openssl_options main cofiguration option.
462 JH/45 local_scan API: documented the current smtp_printf() call. This changed
463 for version 4.90 - adding a "more data" boolean to the arguments.
464 Bumped the ABI version number also, this having been missed previously;
465 release versions 4.90 to 4.92.3 inclusive were effectively broken in
466 respect of usage of smtp_printf() by either local_scan code or libraries
467 accessed via the ${dlfunc } expansion item. Both will need coding
468 adjustment for any calls to smtp_printf() to match the new function
469 signature; a FALSE value for the new argument is always safe.
471 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
472 the file-offset (which the Linux syscall does, and exim expects); this
473 resulted in an indefinite loop.
475 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
476 to do ARC verification. The Authentication-Results: header line added
477 by the configuration then had no ARC item.
483 JH/01 Remove code calling the customisable local_scan function, unless a new
484 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
486 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
487 non-signal-safe functions being used.
489 JH/03 Bug 2269: When presented with a received message having a stupidly large
490 number of DKIM-Signature headers, disable DKIM verification to avoid
491 a resource-consumption attack. The limit is set at twenty.
493 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
494 report of oldest_pass in ${authres } in consequence, and separate out
495 some descriptions of reasons for verification fail.
497 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
498 files in the spool were present and unlocked. A queue-runner could spot
499 them, resulting in a duplicate delivery. Fix that by doing the unlock
500 after the unlink. Investigation by Tim Stewart. Take the opportunity to
501 add more error-checking on spoolfile handling while that code is being
504 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
505 No known attacks, no CVE, this is defensive hardening.
507 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
508 a queue-runner could start a delivery while other operations were ongoing.
509 Cutthrough delivery was a common victim, resulting in duplicate delivery.
510 Found and investigated by Tim Stewart. Fix by using the open message data
511 file handle rather than opening another, and not locally closing it (which
512 releases a lock) for that case, while creating the temporary .eml format
513 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
515 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
516 $sender_verify_failure/$recipient_verify_failure to "random".
518 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
521 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
522 Previously this would segfault.
524 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
527 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
528 like zero, since the resolver should be doing this for us, But we need one
529 as a CNAME but no MX presence gets the CNAME returned; we need to check
530 that doesn't point to an MX to declare it "no MX returned" rather than
531 "error, loop". A new main option is added so the older capability of
532 following some limited number of chain links is maintained.
534 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
536 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
537 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
538 now-deprecated earlier definitions used only specified the range up to TLS
539 1.2 (in the older-version library docs).
541 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
543 JH/15 Rework TLS client-side context management. Stop using a global, and
544 explicitly pass a context around. This enables future use of TLS for
545 connections to service-daemons (eg. malware scanning) while a client smtp
546 connection is using TLS; with cutthrough connections this is quite likely.
548 JH/16 Fix ARC verification to do AS checks in reverse order.
550 JH/17 Support a "tls" option on the ${readsocket } expansion item.
552 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
553 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
554 Previously the "utf8" would be re-prepended for every additional message.
556 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
557 Previously thery were accepted, resulting in issues when attempting to
558 forward messages to a non-supporting MTA.
560 PP/02 Let -n work with printing macros too, not just options.
562 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
563 one parent address was copied, and bogus data was used at delivery-logging
564 time. Either a crash (after delivery) or bogus log data could result.
565 Discovery and analysis by Tim Stewart.
567 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
568 Previously if the string ended mid-character, we did not insert the
569 promised '?' replacement.
571 PP/04 Documentation: current string operators work on bytes, not codepoints.
573 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
574 should pack well giving a smaller memory footprint so better caching and
575 therefore performance. Group the declarations where this can't be done so
576 that the byte-sized flag variables are not interspersed among pointer
577 variables, giving a better chance of good packing by the compiler.
579 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
580 non-null, to avoid issues with sites running BATV. Previously reports were
581 sent with an empty envelope sender so looked like bounces.
583 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
584 The ignore_error flag wasn't being returned from the filter subprocess so
585 was not set for later routers. Investigation and fix by Matthias Kurz.
587 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
588 and a msg:complete for the whole, when a message is manually removed using
589 -Mrm. Developement by Matthias Kurz, hacked on by JH.
591 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
592 a "Gnu special" function, asprintf() in the DB utility binary builds; I
593 hope that is portable enough.
595 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
596 requiring a known-CA anchor certificate; make it now rely entirely on the
597 TLSA as an anchor. Checking the name on the leaf cert against the name
598 on the A-record for the host is still done for TA (but not for EE mode).
600 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
601 included in delivery lines for non-proxied connections, when compiled with
602 SUPPORT_SOCKS and running with proxy logging enabled.
604 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
605 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
606 move the existing event to fire before the normal logging of message
607 failure so that custom logging is bracketed by normal logging.
609 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
610 msg:fail:internal event. Developement by Matthias Kurz.
612 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
613 far too small for todays use of crypto signatures stored there. Go all
614 the way to the max DNS message size of 64kB, even though this might be
615 overmuch for IOT constrained device use.
617 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
618 copy a string over itself. The library routine is documented as not
619 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
621 JH/32 For main options check_spool_space and check_inode_space, where the
622 platform supports 64b integers, support more than the previous 2^31 kB
623 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
624 the previous G, M, k.
626 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
627 $authenticated_fail_id variable on authentication failure. Previously
630 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
631 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
632 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
633 GNUTLS_SEC_PARAM_MEDIUM.
635 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
636 side. Previously we would continue as if no SNI had been received.
638 JH/36 Harden the handling of string-lists. When a list consisted of a sole
639 "<" character, which should be a list-separator specification, we walked
640 off past the nul-terimation.
642 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
643 causes) even when the retry time is not yet met. Previously they were
644 not, meaning that when (say) an account was over-quota and temp-rejecting,
645 and multiple senders' messages were queued, only one sender would get
646 notified on each configured delay_warning cycle.
648 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
650 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
651 cipher-suites, an error can be left on the stack even for a succeeding
652 accept; this results in impossible error messages when a later operation
655 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
656 return error codes indicating retry. Under TLS1.3 this becomes required.
658 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
659 it only wrote the new authenticators, resulting in a lack of tracking of
660 peer changes of ESMTP extensions until the next cache flush.
662 JH/41 Fix the loop reading a message header line to check for integer overflow,
663 and more-often against header_maxsize. Previously a crafted message could
664 induce a crash of the recive process; now the message is cleanly rejected.
666 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
667 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
673 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
674 When redis_servers is set to a list of > 1 element, and the Redis servers
675 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
676 case of MOVED into a DEFER case instead, thus moving the query onto the
677 next server in the list. For a cluster of N elements, all N servers must
678 be defined in redis_servers.
680 GF/02 Catch and remove uninitialized value warning in exiqsumm
681 Check for existence of @ARGV before looking at $ARGV[0]
683 JH/01 Replace the store_release() internal interface with store_newblock(),
684 which internalises the check required to safely use the old one, plus
685 the allocate and data copy operations duplicated in both (!) of the
686 extant use locations.
688 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
689 modifier. This matches the restriction on the commandline.
691 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
692 Previously only the last row was returned.
694 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
695 we assumed that tags in the header were well-formed, and parsed the
696 element content after inspecting only the first char of the tag.
697 Assumptions at that stage could crash the receive process on malformed
700 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
701 While running the DKIM ACL we operate on the Permanent memory pool so that
702 variables created with "set" persist to the DATA ACL. Also (at any time)
703 DNS lookups that fail create cache records using the Permanent pool. But
704 expansions release any allocations made on the current pool - so a dnsdb
705 lookup expansion done in the DKIM ACL releases the memory used for the
706 DNS negative-cache, and bad things result. Solution is to switch to the
707 Main pool for expansions.
708 While we're in that code, add checks on the DNS cache during store_reset,
709 active in the testsuite.
710 Problem spotted, and debugging aided, by Wolfgang Breyha.
712 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
713 When none of the hosts presented to a transport match an already-open
714 connection, close it and proceed with the list. Previously we would
715 queue the message. Spotted by Lena with Yahoo, probably involving
718 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
719 Previously a spurious "250 OK id=" response was appended to the proper
722 JH/08 The "support for" informational output now, which built with Content
723 Scanning support, has a line for the malware scanner interfaces compiled
724 in. Interface can be individually included or not at build time.
726 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
727 by the template makefile "src/EDITME". The "STREAM" support for an older
728 ClamAV interface method is removed.
730 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
731 rows affected is given instead).
733 JH/11 The runtime Berkeley DB library version is now additionally output by
734 "exim -d -bV". Previously only the compile-time version was shown.
736 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
737 SMTP connection. Previously, when one had more receipients than the
738 first, an abortive onward connection was made. Move to full support for
739 multiple onward connections in sequence, handling cutthrough connection
740 for all multi-message initiating connections.
742 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
743 routers. Previously, a multi-recipient message would fail to match the
744 onward-connection opened for the first recipient, and cause its closure.
746 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
747 a timeout on read on a GnuTLS initiating connection, resulting in the
748 initiating connection being dropped. This mattered most when the callout
749 was marked defer_ok. Fix to keep the two timeout-detection methods
752 JH/15 Relax results from ACL control request to enable cutthrough, in
753 unsupported situations, from error to silently (except under debug)
754 ignoring. This covers use with PRDR, frozen messages, queue-only and
757 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
759 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
760 metadata, resulting in a crash in free().
762 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
763 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
764 Broken also in d185889f4, with init system revamp.
766 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
767 Previously we abruptly closed the connection after reading a malware-
768 found indication; now we go on to read the "scan ok" response line,
771 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
772 ACL. Previously, a crash would result.
774 JH/19 Speed up macro lookups during configuration file read, by skipping non-
775 macro text after a replacement (previously it was only once per line) and
776 by skipping builtin macros when searching for an uppercase lead character.
778 JH/20 DANE support moved from Experimental to mainline. The Makefile control
779 for the build is renamed.
781 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
782 was allocated for every new TLS startup, meaning one per message. Fix
783 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
785 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
786 reported the original. Fix to report (as far as possible) the ACL
787 result replacing the original.
789 JH/23 Fix memory leak during multi-message connections using STARTTLS under
790 OpenSSL. Certificate information is loaded for every new TLS startup,
791 and the resources needed to be freed.
793 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
795 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
798 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
799 DATA response info to the (existing) per-recipient response info for
800 the "C=" log element. It can have useful tracking info from the
801 destination system. Patch from Simon Arlott.
803 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
804 length value. Previously this would segfault.
806 HS/02 Support Avast multiline protoocol, this allows passing flags to
807 newer versions of the scanner.
809 JH/28 Ensure that variables possibly set during message acceptance are marked
810 dead before release of memory in the daemon loop. This stops complaints
811 about them when the debug_store option is enabled. Discovered specifically
812 for sender_rate_period, but applies to a whole set of variables.
813 Do the same for the queue-runner and queue-list loops, for variables set
814 from spool message files. Do the same for the SMTP per-message loop, for
815 certain variables indirectly set in ACL operations.
817 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
818 as a multi-recipient message from a mailinglist manager). The coding had
819 an arbitrary cutoff number of characters while checking for more input;
820 enforced by writing a NUL into the buffer. This corrupted long / fast
821 input. The problem was exposed more widely when more pipelineing of SMTP
822 responses was introduced, and one Exim system was feeding another.
823 The symptom is log complaints of SMTP syntax error (NUL chars) on the
824 receiving system, and refused recipients seen by the sending system
825 (propating to people being dropped from mailing lists).
826 Discovered and pinpointed by David Carter.
828 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
829 replaced by the ${authresults } expansion.
831 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
833 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
834 allows proper process termination in container environments.
836 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
837 Previously the "final dot" had a newline after it; ensure it is CR,LF.
839 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
840 and "err_perm", deprecated since 4.83 when the RFC-defined words
841 "temperror" and "permerror" were introduced.
843 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
844 transport-filters or DKIM-signing. The restriction was lost in the
845 consolidation of verify-callout and delivery SMTP handling.
846 Extend the restriction to also cover ARC-signing.
848 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
849 in defer=pass mode supply a 450 to the initiator. Previously the message
852 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
853 tls_require_ciphers is used as before.
855 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
856 "pass_unscanned". Only tmpfails from the scanner are written to
857 the paniclog, as they may require admin intervention (permission
858 denied, license issues). Other scanner errors (like decompression
859 bombs) do not cause a paniclog entry.
861 JH/36 Fix reinitialisation of DKIM logging variable between messages.
862 Previously it was possible to log spurious information in receive log
865 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
866 triggered odd behaviour from Outlook Express clients.
868 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
871 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
872 since the IETF WG has not yet settled on that versus the original
873 "bare" representation.
875 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
876 Previously the millisecond value corrupted the output.
877 Fix also for syslog_pid=no and log_selector +pid, for which the pid
878 corrupted the output.
884 JH/01 Rework error string handling in TLS interface so that the caller in
885 more cases is responsible for logging. This permits library-sourced
886 string to be attached to addresses during delivery, and collapses
887 pairs of long lines into single ones.
889 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
890 during configuration. Wildcards are allowed and expanded.
892 JH/02 Rework error string handling in DKIM to pass more info back to callers.
893 This permits better logging.
895 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
896 do not close it down and have the child transport start it up again on
897 the passed-on TCP connection. Instead, proxy the child (and any
898 subsequent ones) for TLS via a unix-domain socket channel. Logging is
899 affected: the continued delivery log lines do not have any DNSSEC, TLS
900 Certificate or OCSP information. TLS cipher information is still logged.
902 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
903 identical IP addresses on different listening ports. Will also affect
906 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
907 add noisy ifdef guards to special-case this sillyness.
908 Patch from Bernd Kuhls.
910 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
911 than 255 are no longer allowed.
913 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
914 Disable the session-cache too, which might reduce our load. Since we
915 currrectly use a new context for every connection, both as server and
916 client, there is no benefit for these.
917 GnuTLS appears to not support tickets server-side by default (we don't
918 call gnutls_session_ticket_enable_server()) but client side is enabled
919 by default on recent versions (3.1.3 +) unless the PFS priority string
922 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
923 <https://reproducible-builds.org/specs/source-date-epoch/>.
925 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
926 the check for any unsuccessful recipients did not notice the limit, and
927 erroneously found still-pending ones.
929 JH/08 Pipeline CHUNKING command and data together, on kernels that support
930 MSG_MORE. Only in-clear (not on TLS connections).
932 JH/09 Avoid using a temporary file during transport using dkim. Unless a
933 transport-filter is involved we can buffer the headers in memory for
934 creating the signature, and read the spool data file once for the
935 signature and again for transmission.
937 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
938 4.77 as the kernel support then wasn't solid, having issues in 64bit
939 mode. Now, it's been long enough. Add support for FreeBSD also.
941 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
942 case where the routing stage had gathered several addresses to send to
943 a host before calling the transport for the first, we previously failed
944 to close down TLS in the old transport process before passing the TCP
945 connection to the new process. The new one sent a STARTTLS command
946 which naturally failed, giving a failed delivery and bloating the retry
947 database. Investigation and fix prototype from Wolfgang Breyha.
949 JH/12 Fix check on SMTP command input synchronisation. Previously there were
950 false-negatives in the check that the sender had not preempted a response
951 or prompt from Exim (running as a server), due to that code's lack of
952 awareness of the SMTP input buffering.
954 PP/04 Add commandline_checks_require_admin option.
955 Exim drops privileges sanely, various checks such as -be aren't a
956 security problem, as long as you trust local users with access to their
957 own account. When invoked by services which pass untrusted data to
958 Exim, this might be an issue. Set this option in main configuration
959 AND make fixes to the calling application, such as using `--` to stop
962 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
963 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
964 responses to those, into a single TLS record each way (this usually means
965 a single packet). As a side issue, smtp_enforce_sync now works on TLS
968 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
969 affects you only if you're dancing at the edge of the param size limits.
970 If you are, and this message makes sense to you, then: raise the
971 configured limit or use OpenSSL 1.1. Nothing we can do for older
974 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
975 cmdline element to get the documented default one. Previously it was
978 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
979 are unsuable for non-message-specific reasons - eg. connection timeout,
980 banner-time rejection.
982 JH/16 Fix logging of delivery remote port, when specified by router, under
985 PP/06 Repair manualroute's ability to take options in any order, even if one
986 is the name of a transport.
989 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
991 JH/17 Change the list-building routines interface to use the expanding-string
992 triplet model, for better allocation and copying behaviour.
994 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
995 Previously it was constructed the first time a possibly-matching string
996 was met in the configuration file input during startup; now it is done
999 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1000 compatible one, to avoid the (poorly documented) possibility of a config
1001 file in the working directory redirecting the DB files, possibly correpting
1002 some existing file. CVE-2017-10140 assigned for BDB.
1004 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1005 cache-hot. Previously, although the result was properly cached, the
1006 initial verify call returned a defer.
1008 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1009 the main verify for receipient in uncached-mode.
1011 JH/22 Retire historical build files to an "unsupported" subdir. These are
1012 defined as "ones for which we have no current evidence of testing".
1014 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1015 if present. Previously it was ignored.
1017 JH/24 Start using specified-initialisers in C structure init coding. This is
1018 a C99 feature (it's 2017, so now considered safe).
1020 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1021 if was a fixed-sized field and bitmask ops via macros; it is now more
1024 PP/07 GitHub PR 56: Apply MariaDB build fix.
1025 Patch provided by Jaroslav Škarvada.
1027 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1028 during Coverity cleanups [4.87 JH/47]
1029 Diagnosis and fix provided by Michael Fischer v. Mollard.
1031 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1032 the right size to place the terminating semicolon on its own folded
1033 line, the header hash was calculated to an incorrect value thanks to
1034 the (relaxed) space the fold became.
1036 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1037 and confused the parent.
1039 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1040 which could crash as a result. This could lead to undeliverable messages.
1042 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1045 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1046 matters on fast-turnover and PID-randomising systems, which were getting
1047 out-of-order delivery.
1049 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1050 a possibly-overlapping copy. The symptom was that "Remote host closed
1051 connection in response to HELO" was logged instead of the actual 4xx
1054 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1055 Previously only that bufferd was discarded, resulting in SYMTP command
1058 JH/32 DKIM: when a message has multiple signatures matching an identity given
1059 in dkim_verify_signers, run the dkim acl once for each. Previously only
1060 one run was done. Bug 2189.
1062 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1063 "panic the current process" to "deliberately defer". The panic log is
1064 still written with the problem list name; the mail and reject logs now
1065 get a temp-reject line for the message that was being handled, saying
1066 something like "domains check lookup or other defer". The SMTP 451
1067 message is still "Temporary local problem".
1069 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1070 A crafted sequence of BDAT commands could result in in-use memory beeing
1071 freed. CVE-2017-16943.
1073 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1074 from SMTP input. Previously it was always done; now only done for DATA
1075 and not BDAT commands. CVE-2017-16944.
1077 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1078 to the message (such as an overlong header line). Previously this was
1079 not done and we did not exit BDAT mode. Followon from the previous item
1080 though a different problem.
1086 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1087 than -2003 did; needs libidn2 in addition to libidn.
1089 JH/02 The path option on a pipe transport is now expanded before use.
1091 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1092 Patch provided by "Björn", documentation fix added too.
1094 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1095 missing a wire-to-host endian conversion.
1097 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1098 close after a BDAT command line could be taken as a following command,
1099 giving a synch failure. Fix by only checking for synch immediately
1100 before acknowledging the chunk.
1102 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1103 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1104 macro. Patches provided by Josh Soref.
1106 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1107 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1108 by IANA as a service extension. However, John Klensin suggests that we
1111 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1112 the dkim code may be unix-mode line endings rather than smtp wire-format
1113 CRLF, so prepend a CR to any bare LF.
1115 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1116 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1118 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1119 the first were themselves being wrongly included in the feed into dkim
1120 processing; with most chunk sizes in use this resulted in an incorrect
1121 body hash calculated value.
1123 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1124 DKIM signature block, for verification. Although advised against by
1125 standards it is specifically not ruled illegal.
1127 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1129 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1130 missing a body hash (the bh= tag).
1132 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1133 It seems that HAProxy sends the Proxy Protocol information in clear and
1134 only then does a TLS startup, so do the same.
1136 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1137 TCP connections (such as for Spamd) unless the daemon successfully set
1138 Fast Open mode on its listening sockets. This fixes breakage seen on
1139 too-old kernels or those not configured for Fast Open, at the cost of
1140 requiring both directions being enabled for TFO, and TFO never being used
1141 by non-daemon-related Exim processes.
1143 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1144 endings, at least on the first header line. Try to canonify any that get
1145 past that check, despite the cost.
1147 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1148 now limited to an arbitrary five deep, while parsing addresses with the
1149 strip_excess_angle_brackets option enabled.
1151 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1152 instead leave the unprompted TLS handshake in socket buffer for the
1153 TLS library to consume.
1155 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1157 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1159 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1160 region in one operation, for speed, and this leaves assigned pointers
1161 dangling. Add checks run only under the testsuite which checks all
1162 variables at a store-reset and panics on a dangling pointer; add code
1163 explicitly nulling out all the variables discovered. Fixes one known
1164 bug: a transport crash, where a dangling pointer for $sending_ip_address
1165 originally assigned in a verify callout, is re-used.
1167 PP/06 Drop '.' from @INC in various Perl scripts.
1169 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1171 PP/08 Reduce a number of compilation warnings under clang; building with
1172 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1173 should be warning-free.
1175 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1177 HS/01 Fix portability problems introduced by PP/08 for platforms where
1178 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1180 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1181 chunk. This allows us to accept broken chunked messages. We need a more
1182 general solution here.
1184 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1185 already-broken messages in the queue.
1187 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1189 JH/19 Fix reference counting bug in routing-generated-address tracking.
1195 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1196 supports it and a size is available (ie. the sending peer gave us one).
1198 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1199 years of being deprecated). The replacements are the ACLs
1200 acl_smtp_mime and acl_not_smtp_mime.
1202 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1203 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1204 or even an in-clear connection were permitted. Now, if the host lookup
1205 was dnssec and dane was requested then the host is only used if the
1206 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1207 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1208 if one fails this test.
1209 This means that a poorly-configured remote DNS will make it incommunicado;
1210 but it protects against a DNS-interception attack on it.
1212 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1213 non-noisy when a race steals the message being considered.
1215 JH/05 If main configuration option tls_certificate is unset, generate a
1216 self-signed certificate for inbound TLS connections.
1218 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1219 in rewrites and routers.
1221 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1222 and logged a warning sing 4.83; now they are a configuration file error.
1224 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1225 (lacking @domain). Apply the same qualification processing as RCPT.
1227 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1229 JH/10 Support ${sha256:} applied to a string (as well as the previous
1232 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1233 a cutthrough deliver is pending, as we always want to make a connection.
1234 This also avoids re-routing the message when later placing the cutthrough
1235 connection after a verify cache hit.
1236 Do not update it with the verify result either.
1238 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1239 when routing results in more than one destination address.
1241 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1242 signing (which inhibits the cutthrough capability). Previously only
1243 the presence of an option was tested; now an expansion evaluating as
1244 empty is permissible (obviously it should depend only on data available
1245 when the cutthrough connection is made).
1247 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1248 the relevant preceding SMTP command did not note the pipelining mode.
1250 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1251 Previously they were not counted.
1253 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1254 as one having no matching records. Previously we deferred the message
1255 that needed the lookup.
1257 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1258 distinguished as "(=".
1260 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1261 for missing MX records. Previously it only worked for missing A records.
1263 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1265 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1266 after the data-go-ahead and data-ack. Patch from Jason Betts.
1268 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1269 even for a "none" policy. Patch from Tony Meyer.
1271 JH/22 Fix continued use of a connection for further deliveries. If a port was
1272 specified by a router, it must also match for the delivery to be
1275 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1276 When one of the recipients of a message was unsuitable for the connection
1277 (has no matching addresses), we lost track of needing to mark it
1278 deferred. As a result mail would be lost.
1280 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1282 JH/25 Decoding ACL controls is now done using a binary search; the source code
1283 takes up less space and should be simpler to maintain. Merge the ACL
1284 condition decode tables also, with similar effect.
1286 JH/26 Fix problem with one_time used on a redirect router which returned the
1287 parent address unchanged. A retry would see the parent address marked as
1288 delivered, so not attempt the (identical) child. As a result mail would
1291 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1292 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1293 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1296 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1297 Main config options check_{log,spool}_{inodes,space} are now
1298 100 inodes, 10MB unless set otherwise in the configuration.
1300 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1301 Previously it only applied to the main-section connection policy
1304 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1306 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1307 by me. Added RFC7919 DH primes as an alternative.
1309 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1310 are not in the system include path.
1312 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1313 GnuTLS, when a session startup failed (eg because the client disconnected)
1314 Exim did stdio operations after fclose. This was exposed by a recent
1315 change which nulled out the file handle after the fclose.
1317 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1318 signed directly by the cert-signing cert, rather than an intermediate
1319 OCSP-signing cert. This is the model used by LetsEncrypt.
1321 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1323 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1324 an incoming connection.
1326 HS/02 Bug 1802: Do not half-close the connection after sending a request
1329 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1330 fallback to "prime256v1".
1332 JH/34 SECURITY: Use proper copy of DATA command in error message.
1333 Could leak key material. Remotely exploitable. CVE-2016-9963.
1339 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1340 and 3.4.4 - once the server is enabled to respond to an OCSP request
1341 it does even when not requested, resulting in a stapling non-aware
1342 client dropping the TLS connection.
1344 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1345 support variable-length bit vectors. No functional change.
1347 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1348 The I= interface field on outgoing lines is now after the H= remote
1349 host field, same as incoming lines. There is a separate
1350 outgoing_interface log selector which allows you to disable the
1353 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1354 If not running log_selector +smtp_connection the mainlog would be held
1355 open indefinitely after a "too many connections" event, including to a
1356 deleted file after a log rotate. Leave the per net connection logging
1357 leaving it open for efficiency as that will be quickly detected by the
1358 check on the next write.
1360 HS/01 Bug 1671: Fix post transport crash.
1361 Processing the wait-<transport> messages could crash the delivery
1362 process if the message IDs didn't exist for some reason. When
1363 using 'split_spool_directory=yes' the construction of the spool
1364 file name failed already, exposing the same netto behaviour.
1366 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1367 mime_regex ACL conditions.
1369 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1370 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1371 to HELO, local diagnostic string.
1373 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1374 log line to debug. Even when configured with a tls authenticator many
1375 client connections are expected to not authenticate in this way, so
1376 an authenticate fail is not an error.
1378 HS/02 Add the Exim version string to the process info. This way exiwhat
1379 gives some more detail about the running daemon.
1381 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1382 matter for fast-change records such as DNSBLs.
1384 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1385 retry record, even if constant. There may be multiple transports with
1386 different interface settings and the retry behaviour needs to be kept
1389 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1391 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1393 JH/10 Bug 840: fix log_defer_output option of pipe transport
1395 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1396 hosts_randomize. This matters a lot when combined with mua_wrapper.
1398 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1399 ${quote_pgsql:<string>} operator.
1401 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1403 JH/14 Change header file name for freeradius-client. Relevant if compiling
1404 with Radius support; from the Gentoo tree and checked under Fedora.
1406 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1408 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1409 option transport_filter, meaning no filtering.
1411 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1413 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1414 defaults to "*" (all hosts). The variable is now available when not built
1415 with TLS, default unset, mainly to enable keeping the testsuite sane.
1416 If a server certificate is not supplied (via tls_certificate) an error is
1417 logged, and clients will find TLS connections fail on startup. Presumably
1418 they will retry in-clear.
1419 Packagers of Exim are strongly encouraged to create a server certificate
1420 at installation time.
1422 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1423 with the $config_file variable.
1425 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1426 in transport context, after the attempt, and per-recipient. The latter type
1427 is per host attempted. The event data is the error message, and the errno
1428 information encodes the lookup type (A vs. MX) used for the (first) host,
1429 and the trailing two digits of the smtp 4xx response.
1431 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1432 to write to mainlog (or rejectlog, paniclog) in the window between file
1433 creation and permissions/ownership being changed. Particularly affects
1434 installations where exicyclog is run as root, rather than exim user;
1435 result is that the running daemon panics and dies.
1437 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1439 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1440 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1441 "pri" and "weight". Note that the previous implicit priority given by the
1442 list order is no longer honoured.
1444 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1445 for DKIM processing.
1447 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1448 by defining SUPPORT_SOCKS.
1450 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1451 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1452 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1453 variables are renamed to proxy_{local,external}_{address,port}.
1455 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1456 for a build by defining SUPPORT_I18N
1458 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1459 of the query string, and make ${quote_redis:} do that quoting.
1461 JH/29 Move Events support from Experimental to mainline, enabled by default
1462 and removable for a build by defining DISABLE_EVENT.
1464 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1466 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1467 cached by the daemon.
1469 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1470 by defining LOOKUP_REDIS. The libhiredis library is required.
1472 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1473 keys are given for lookup.
1475 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1476 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1477 only supported when built with TLS support. The PolarSSL SHA routines
1478 are still used when the TLS library is too old for convenient support.
1480 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1481 openssl_options), for security. OpenSSL forces this from version 1.1.0
1482 server-side so match that on older versions.
1484 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1485 allocation for $value could be released as the expansion processing
1486 concluded, but leaving the global pointer active for it.
1488 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1489 and to use the domains and local_parts ACL conditions.
1491 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1492 incorrectly not doubled on cutthrough transmission, hence seen as a
1493 body-termination at the receiving system - resulting in truncated mails.
1494 Commonly the sender saw a TCP-level error, and retransmitted the message
1495 via the normal store-and-forward channel. This could result in duplicates
1496 received - but deduplicating mailstores were liable to retain only the
1497 initial truncated version.
1499 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1501 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1503 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1504 we're in there, support oversigning also; bug 1309.
1506 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1508 HS/04 Add support for keep_environment and add_environment options.
1510 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1511 either intentional arithmetic overflow during PRNG, or testing config-
1514 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1515 delivery resulted in actual delivery. Cancel cutthrough before DATA
1518 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1519 rejects a recipient: pass the reject to the originator.
1521 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1522 Many were false-positives and ignorable, but it's worth fixing the
1525 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1526 for the new environment-manipulation done at startup. Move the routines
1527 from being local to tls.c to being global via the os.c file.
1529 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1530 an extract embedded as result-arg for a map, the first arg for extract
1531 is unavailable so we cannot tell if this is a numbered or keyed
1532 extraction. Accept either.
1538 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1541 JH/02 The smtp transport option "multi_domain" is now expanded.
1543 JH/03 The smtp transport now requests PRDR by default, if the server offers
1546 JH/04 Certificate name checking on server certificates, when exim is a client,
1547 is now done by default. The transport option tls_verify_cert_hostnames
1548 can be used to disable this per-host. The build option
1549 EXPERIMENTAL_CERTNAMES is withdrawn.
1551 JH/05 The value of the tls_verify_certificates smtp transport and main options
1552 default to the word "system" to access the system default CA bundle.
1553 For GnuTLS, only version 3.0.20 or later.
1555 JH/06 Verification of the server certificate for a TLS connection is now tried
1556 (but not required) by default. The verification status is now logged by
1557 default, for both outbound TLS and client-certificate supplying inbound
1560 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1563 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1564 Status Notification (bounce) messages are now MIME format per RFC 3464.
1565 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1566 under the control of the dsn_advertise_hosts option, and routers may
1567 have a dsn_lasthop option.
1569 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1570 default, modifiable by a malware= option. The list separator for
1571 the options can now be changed in the usual way. Bug 68.
1573 JH/10 The smtp_receive_timeout main option is now expanded before use.
1575 JH/11 The incoming_interface log option now also enables logging of the
1576 local interface on delivery outgoing connections.
1578 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1579 if the interface and destination host and port all match.
1581 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1584 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1585 Patch from Andrew Lewis.
1587 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1588 now supports optional time-restrictions, weighting, and priority
1589 modifiers per server. Patch originally by <rommer@active.by>.
1591 JH/16 The spamd_address main option now supports a mixed list of local
1592 and remote servers. Remote servers can be IPv6 addresses, and
1593 specify a port-range.
1595 JH/17 Bug 68: The spamd_address main option now supports an optional
1596 timeout value per server.
1598 JH/18 Bug 1581: Router and transport options headers_add/remove can
1599 now have the list separator specified.
1601 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1604 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1607 JH/21 Support for the A6 type of dns record is withdrawn.
1609 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1610 rather than the verbs used.
1612 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1613 from 255 to 1024 chars.
1615 JH/24 Verification callouts now attempt to use TLS by default.
1617 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1618 are generic router options now. The defaults didn't change.
1620 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1621 Original patch from Alexander Shikoff, worked over by JH.
1623 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1624 files if ZCAT_COMMAND is not executable.
1626 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1628 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1630 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1631 Normally benign, it bites when the pair was led to by a CNAME;
1632 modern usage is to not canonicalize the domain to a CNAME target
1633 (and we were inconsistent anyway for A-only vs AAAA+A).
1635 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1637 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1638 when evaluating $sender_host_dnssec.
1640 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1641 $sender_helo_dnssec variable.
1643 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1645 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1647 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1649 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1650 documented as working, but never had. Support all but $spam_report.
1652 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1653 added for tls authenticator.
1655 HS/03 Add perl_taintmode main config option
1661 TL/01 When running the test suite, the README says that variables such as
1662 no_msglog_check are global and can be placed anywhere in a specific
1663 test's script, however it was observed that placement needed to be near
1664 the beginning for it to behave that way. Changed the runtest perl
1665 script to read through the entire script once to detect and set these
1666 variables, reset to the beginning of the script, and then run through
1667 the script parsing/test process like normal.
1669 TL/02 The BSD's have an arc4random API. One of the functions to induce
1670 adding randomness was arc4random_stir(), but it has been removed in
1671 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1672 function when detected.
1674 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1675 cause callback expansion.
1677 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1678 syntax errors in an expansion can be treated as a string instead of
1679 logging or causing an error, due to the internal use of bool_lax
1680 instead of bool when processing it.
1682 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1683 server certificates when making smtp deliveries.
1685 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1687 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1689 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1691 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1692 Merged patch from Sebastian Wiedenroth.
1694 JH/05 Fix results-pipe from transport process. Several recipients, combined
1695 with certificate use, exposed issues where response data items split
1696 over buffer boundaries were not parsed properly. This eventually
1697 resulted in duplicates being sent. This issue only became common enough
1698 to notice due to the introduction of connection certificate information,
1699 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1701 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1702 size buffer was used, resulting in syntax errors when an expansion
1705 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1706 version 3.3.6 or later.
1708 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1709 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1710 both become "event_action", the variables become $event_name, $event_data
1711 and $event_defer_errno. There is a new variable $verify_mode, usable in
1712 routers, transports and related events. The tls:cert event is now also
1713 raised for inbound connections, if the main configuration event_action
1716 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1717 early OCSP support, but no stapling (appears to be less than 1.0.0).
1719 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1720 server certificate names available under the smtp transport option
1721 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1724 JH/10 Time-related extraction expansions from certificates now use the main
1725 option "timezone" setting for output formatting, and are consistent
1726 between OpenSSL and GnuTLS compilations. Bug 1541.
1728 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1729 encoded parameter in the incoming message. Bug 1558.
1731 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1732 include certificate info, eximon was claiming there were spoolfile
1735 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1737 JH/14 Log delivery-related information more consistently, using the sequence
1738 "H=<name> [<ip>]" wherever possible.
1740 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1741 are problematic for Debian distribution, omit them from the release
1744 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1746 JH/16 Fix string representation of time values on 64bit time_t architectures.
1749 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1750 output list separator was used.
1755 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1756 checkers that were complaining about end of non-void function with no
1759 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1760 This was a regression introduced in 4.83 by another bugfix.
1762 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1764 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1765 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1771 TF/01 Correctly close the server side of TLS when forking for delivery.
1773 When a message was received over SMTP with TLS, Exim failed to clear up
1774 the incoming connection properly after forking off the child process to
1775 deliver the message. In some situations the subsequent outgoing
1776 delivery connection happened to have the same fd number as the incoming
1777 connection previously had. Exim would try to use TLS and fail, logging
1778 a "Bad file descriptor" error.
1780 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1781 utilities have not been installed.
1783 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1784 temporary space as the ACL may create new global variables.
1786 TL/01 LDAP support uses per connection or global context settings, depending
1787 upon the detected version of the libraries at build time.
1789 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1790 to extract and use the src ip:port in logging and expansions as if it
1791 were a direct connection from the outside internet. PPv2 support was
1792 updated based on HAProxy spec change in May 2014.
1794 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1796 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1797 Properly escape header and check for NULL return.
1799 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1802 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1804 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1805 characters in header names, implemented as a verify condition.
1806 Contributed by Michael Fischer v. Mollard.
1808 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1809 results permerror and temperror. Previous values are deprecated but
1810 still accepted. In a future release, err_perm and err_temp will be
1811 completely removed, which will be a backward incompatibility if the
1812 ACL tests for either of these two old results. Patch contributed by
1813 user bes-internal on the mailing list.
1815 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1817 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1818 selectors, in both main and reject logs.
1820 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1823 JH/07 Add malware type "sock" for talking to simple daemon.
1825 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1827 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1828 routers/transports under cutthrough routing.
1830 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1831 numbers. Touch up "bool" conditional to keep the same definition.
1833 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1835 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1837 JH/12 Expand items in router/transport headers_add or headers_remove lists
1838 individually rather than the list as a whole. Bug 1452.
1840 Required for reasonable handling of multiple headers_ options when
1841 they may be empty; requires that headers_remove items with embedded
1842 colons must have them doubled (or the list-separator changed).
1844 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1845 view the policy declared in the DMARC record. Currently, $dmarc_status
1846 is a combined value of both the record presence and the result of the
1849 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1851 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1852 dnslookup router and the smtp transport (applying to the forward
1855 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1856 of ldap servers used for a specific lookup. Patch provided by Heiko
1859 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1860 New variable $lookup_dnssec_authenticated for observability.
1862 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1863 Patch submitted by Lars Timman.
1865 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1867 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1868 Requires trusted mode and valid format message id, aborts otherwise.
1869 Patch contributed by Heiko Schlichting.
1871 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1872 certextract with support for various fields. Bug 1358.
1874 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1875 is requested by default, modifiable by smtp transport option
1878 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1879 operate on certificate variables to give certificate fingerprints
1880 Also new ${sha256:cert_variable}.
1882 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1884 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1885 Christian Aistleitner.
1887 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1889 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1890 file. Patch from Wolfgang Breyha.
1892 JH/25 Expand the coverage of the delivery $host and $host_address to
1893 client authenticators run in verify callout. Bug 1476.
1895 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1896 align with daemon_smtp_ports. Bug 72.
1898 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1899 support and error reporting did not work properly.
1901 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1902 and is readable. Patch from Andrew Colin Kissa.
1904 TL/14 Enhance documentation of ${run expansion and how it parses the
1905 commandline after expansion, particularly in the case when an
1906 unquoted variable expansion results in an empty value.
1908 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1910 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1911 by renaming away from it. Addresses GH issue 15, reported by
1914 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1916 TL/15 SECURITY: prevent double expansion in math comparison functions
1917 (can expand unsanitized data). Not remotely exploitable.
1924 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1926 PP/02 Make -n do something, by making it not do something.
1927 When combined with -bP, the name of an option is not output.
1929 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1932 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1933 $sender_host_name and config options to manage this, and basic check
1936 PP/05 DSCP support for outbound connections and control modifier for inbound.
1938 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1939 (Only plugin which currently uses this is kerberos4, which nobody should
1940 be using, but we should make it available and other future plugins might
1941 conceivably use it, even though it would break NAT; stuff *should* be
1942 using channel bindings instead).
1944 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1945 name; added for Sendmail compatibility; requires admin caller.
1946 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1947 just ignore it); requires trusted caller.
1948 Also parse but ignore: -Ac -Am -X<logfile>
1951 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1953 TL/02 Add +smtp_confirmation as a default logging option.
1955 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1956 Patch by Magnus Holmgren from 2007-02-20.
1958 TL/04 Bugzilla 1281 - Spec typo.
1959 Bugzilla 1283 - Spec typo.
1960 Bugzilla 1290 - Spec grammar fixes.
1962 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1964 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1966 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1967 mailing list by by Dmitry Isaikin.
1969 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1971 JH/02 Support "G" suffix to numbers in ${if comparisons.
1973 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1975 NM/01 Bugzilla 1197 - Spec typo
1976 Bugzilla 1196 - Spec examples corrections
1978 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1980 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1981 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1984 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1985 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1987 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1988 "acl {{name}{arg}...}", and optional args on acl condition
1991 JH/05 Permit multiple router/transport headers_add/remove lines.
1993 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1995 JH/07 Avoid using a waiting database for a single-message-only transport.
1996 Performance patch from Paul Fisher. Bugzilla 1262.
1998 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2001 JH/09 Add $headers_added variable, with content from use of ACL modifier
2002 add_header (but not yet added to the message). Bugzilla 199.
2004 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2005 Pulled from Bugzilla 817 by Wolfgang Breyha.
2007 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2009 (nb: this is the same fix as in Exim 4.80.1)
2011 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2014 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2017 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2019 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2020 advertise SMTP AUTH mechanism to us, instead of a generic
2021 protocol violation error. Also, make Exim more robust to bad
2022 data from the Dovecot auth socket.
2024 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2026 When a queue runner is handling a message, Exim first routes the
2027 recipient addresses, during which it prunes them based on the retry
2028 hints database. After that it attempts to deliver the message to
2029 any remaining recipients. It then updates the hints database using
2032 So if a recipient address works intermittently, it can get repeatedly
2033 deferred at routing time. The retry hints record remains fresh so the
2034 address never reaches the final cutoff time.
2036 This is a fairly common occurrence when a user is bumping up against
2037 their storage quota. Exim had some logic in its local delivery code
2038 to deal with this. However it did not apply to per-recipient defers
2039 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2041 This change adds a proper retry rule check during routing so that the
2042 final cutoff time is checked against the message's age. We only do
2043 this check if there is an address retry record and there is not a
2044 domain retry record; this implies that previous attempts to handle
2045 the address had the retry_use_local_parts option turned on. We use
2046 this as an approximation for the destination being like a local
2047 delivery, as in LMTP.
2049 I suspect this new check makes the old local delivery cutoff check
2050 redundant, but I have not verified this so I left the code in place.
2052 TF/02 Correct gecos expansion when From: is a prefix of the username.
2054 Test 0254 submits a message to Exim with the header
2058 When I ran the test suite under the user fanf2, Exim expanded
2059 the header to contain my full name, whereas it should have added
2060 a Resent-Sender: header. It erroneously treats any prefix of the
2061 username as equal to the username.
2063 This change corrects that bug.
2065 GF/01 DCC debug and logging tidyup
2066 Error conditions log to paniclog rather than rejectlog.
2067 Debug lines prefixed by "DCC: " to remove any ambiguity.
2069 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2071 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2072 Bug spotted by Jeremy Harris; was flawed since initial commit.
2073 Would have resulted in OCSP responses post-SNI triggering an Exim
2074 NULL dereference and crash.
2076 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2078 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2079 Bug detection, analysis and fix by Samuel Thibault.
2080 Bugzilla 1331, Debian bug #698092.
2082 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2084 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2085 Server implementation by Todd Lyons, client by JH.
2086 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2087 config variable "prdr_enable" controls whether the server
2088 advertises the facility. If the client requests PRDR a new
2089 acl_data_smtp_prdr ACL is called once for each recipient, after
2090 the body content is received and before the acl_smtp_data ACL.
2091 The client is controlled by both of: a hosts_try_prdr option
2092 on the smtp transport, and the server advertisement.
2093 Default client logging of deliveries and rejections involving
2094 PRDR are flagged with the string "PRDR".
2096 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2097 fclose(). Diagnosis by Todd Lyons.
2099 PP/17 Update configure.default to handle IPv6 localhost better.
2100 Patch by Alain Williams (plus minor tweaks).
2103 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2104 This is now consistent with GnuTLS, and is now documented: the
2105 previous undocumented portable approach to treating the option as
2106 unset was to force an expansion failure. That still works, and
2107 an empty string is now equivalent.
2109 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2110 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2111 not performing validation itself.
2113 PP/20 Added force_command boolean option to pipe transport.
2114 Patch from Nick Koston, of cPanel Inc.
2116 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2119 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2121 PP/21 Fix eximon continuous updating with timestamped log-files.
2122 Broken in a format-string cleanup in 4.80, missed when I repaired the
2123 other false fix of the same issue.
2124 Report and fix from Heiko Schlichting.
2127 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2128 Report from Prashanth Katuri.
2130 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2131 It's SecureTransport, so affects any MacOS clients which use the
2132 system-integrated TLS libraries, including email clients.
2134 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2135 using a MIME ACL for non-SMTP local injection.
2136 Report and assistance in diagnosis by Warren Baker.
2138 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2140 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2142 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2143 last id that failed so it may be referenced in subsequent ACL's.
2145 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2148 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2149 ldap library initialization, allowing self-signed CA's to be
2150 used. Also properly sets require_cert option later in code by
2151 using NULL (global ldap config) instead of ldap handle (per
2152 session). Bug diagnosis and testing by alxgomz.
2154 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2155 the src/util/ subdirectory.
2157 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2158 renamed to Transport Post Delivery Action by Jeremy Harris, as
2161 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2162 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2163 redis_servers = needs to be configured which will be used by the redis
2164 lookup. Patch from Warren Baker, of The Packet Hub.
2166 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2168 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2169 hostname or reverse DNS when processing a host list. Used suggestions
2170 from multiple comments on this bug.
2172 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2174 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2175 Missed a few lines, added it to make the runtest require no keyboard
2178 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2179 contains upper case chars. Make router use caseful_local_part.
2181 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2182 support when GnuTLS has been built with p11-kit.
2188 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2190 This, or similar/improved, will also be change PP/11 of 4.82.
2196 PP/01 Handle short writes when writing local log-files.
2197 In practice, only affects FreeBSD (8 onwards).
2198 Bugzilla 1053, with thanks to Dmitry Isaikin.
2200 NM/01 Bugzilla 949 - Documentation tweak
2202 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2205 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2207 PP/02 Implemented gsasl authenticator.
2209 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2211 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2212 `pkg-config foo` for cflags/libs.
2214 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2215 with rest of GSASL and with heimdal_gssapi.
2217 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2218 `pkg-config foo` for cflags/libs for the TLS implementation.
2220 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2221 properties get this fed in as external SSF. A number of robustness
2222 and debugging improvements to the cyrus_sasl authenticator.
2224 PP/08 cyrus_sasl server now expands the server_realm option.
2226 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2227 Patch by Jeremy Harris.
2229 PP/10 Added dbmjz lookup type.
2231 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2233 PP/12 MAIL args handles TAB as well as SP, for better interop with
2234 non-compliant senders.
2235 Analysis and variant patch by Todd Lyons.
2237 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2238 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2239 Patch from Dirk Mueller <dmueller@suse.com>
2241 PP/13 tls_peerdn now print-escaped for spool files.
2242 Observed some $tls_peerdn in wild which contained \n, which resulted
2243 in spool file corruption.
2245 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2246 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2247 or write after TLS renegotiation, which otherwise led to messages
2250 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2251 as a tracking header (ie: a signed header comes before the signature).
2252 Patch from Wolfgang Breyha.
2254 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2255 comma-sep list; embedded commas doubled.
2257 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2259 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2261 Report and patch from Dmitry Banschikov.
2263 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2264 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2265 failures. We appear to now support TLS1.1+ with Exim.
2267 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2268 lets Exim select keys and certificates based upon TLS SNI from client.
2269 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2270 before an outbound SMTP session. New log_selector, +tls_sni.
2272 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2273 NULL dereference. Report and patch from Alun Jones.
2275 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2276 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2277 Not seeing resolver debug output on NetBSD, but suspect this is a
2278 resolver implementation change.
2280 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2281 Left warnings. Added "eximon gdb" invocation mode.
2283 PP/21 Defaulting "accept_8bitmime" to true, not false.
2285 PP/22 Added -bw for inetd wait mode support.
2287 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2288 locate the relevant includes and libraries. Made this the default.
2290 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2291 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2293 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2294 This may cause build issues on older platforms.
2296 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2297 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2298 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2299 Added SNI support via GnuTLS too.
2300 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2302 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2304 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2305 Applied second patch from Janne, implementing suggestion to default
2306 multiple-strings-in-record handling to match SPF spec.
2308 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2310 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2311 read-only, out of scope).
2312 Patch from Wolfgang Breyha, report from Stuart Northfield.
2314 PP/29 Fix three issues highlighted by clang analyser static analysis.
2315 Only crash-plausible issue would require the Cambridge-specific
2316 iplookup router and a misconfiguration.
2317 Report from Marcin Mirosław.
2319 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2321 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2322 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2323 As part of this, removing so much warning spew let me fix some minor
2324 real issues in debug logging.
2326 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2327 assignment on my part. Fixed.
2329 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2330 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2331 Janne Snabb (who went above and beyond: thank you).
2333 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2334 string otherwise requires a connection and a bunch more work and it's
2335 relatively easy to get wrong. Should also expose TLS library linkage
2338 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2339 64-bit ${eval} (JH/03).
2341 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2342 GNU libc to support some of the 64-bit stuff, should not lead to
2343 conflicts. Defined before os.h is pulled in, so if a given platform
2344 needs to override this, it can.
2346 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2347 protection layer was required, which is not implemented.
2348 Bugzilla 1254, patch from Wolfgang Breyha.
2350 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2351 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2352 tls_dhparam take prime identifiers. Also unbreak combination of
2353 OpenSSL+DH_params+TLSSNI.
2355 PP/39 Disable SSLv2 by default in OpenSSL support.
2361 PP/01 Solaris build fix for Oracle's LDAP libraries.
2362 Bugzilla 1109, patch from Stephen Usher.
2364 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2366 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2369 TF/02 Fix a couple more cases where we did not log the error message
2370 when unlink() failed. See also change 4.74-TF/03.
2372 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2373 lock up or crash if it happened to be inside a call to libc when it
2374 got a SIGUSR1 from exiwhat.
2376 The SIGUSR1 handler appends the current process status to the process
2377 log which is later printed by exiwhat. It used to use the general
2378 purpose logging code to do this, but several functions it calls are
2379 not safe for signals.
2381 The new output code in the SIGUSR1 handler is specific to the process
2382 log, and simple enough that it's easy to inspect for signal safety.
2383 Removing some special cases also simplifies the general logging code.
2384 Removing the spurious timestamps from the process log simplifies
2387 TF/04 Improved ratelimit ACL condition.
2389 The /noupdate option has been deprecated in favour of /readonly which
2390 has clearer semantics. The /leaky, /strict, and /readonly update modes
2391 are mutually exclusive. The update mode is no longer included in the
2392 database key; it just determines when the database is updated. (This
2393 means that when you upgrade Exim will forget old rate measurements.)
2395 Exim now checks that the per_* options are used with an update mode that
2396 makes sense for the current ACL. For example, when Exim is processing a
2397 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2398 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2399 must specify per_mail/readonly. If you omit the update mode it defaults to
2400 /leaky where that makes sense (as before) or /readonly where required.
2402 The /noupdate option is now undocumented but still supported for
2403 backwards compatibility. It is equivalent to /readonly except that in
2404 ACLs where /readonly is required you may specify /leaky/noupdate or
2405 /strict/noupdate which are treated the same as /readonly.
2407 A useful new feature is the /count= option. This is a generalization
2408 of the per_byte option, so that you can measure the throughput of other
2409 aggregate values. For example, the per_byte option is now equivalent
2410 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2412 The per_rcpt option has been generalized using the /count= mechanism
2413 (though it's more complicated than the per_byte equivalence). When it is
2414 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2415 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2416 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2417 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2418 Note that using per_rcpt with a non-readonly update mode in more than
2419 one ACL will cause the recipients to be double-counted. (The per_mail
2420 and per_byte options don't have this problem.)
2422 The handling of very low rates has changed slightly. If the computed rate
2423 is less than the event's count (usually one) then this event is the first
2424 after a long gap. In this case the rate is set to the same as this event's
2425 count, so that the first message of a spam run is counted properly.
2427 The major new feature is a mechanism for counting the rate of unique
2428 events. The new per_addr option counts the number of different
2429 recipients that someone has sent messages to in the last time period. It
2430 behaves like per_rcpt if all the recipient addresses are different, but
2431 duplicate recipient addresses do not increase the measured rate. Like
2432 the /count= option this is a general mechanism, so the per_addr option
2433 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2434 example, measure the rate that a client uses different sender addresses
2435 with the options per_mail/unique=$sender_address. There are further
2436 details in the main documentation.
2438 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2440 TF/06 Removed a few PCRE remnants.
2442 TF/07 Automatically extract Exim's version number from tags in the git
2443 repository when doing development or release builds.
2445 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2446 Bugzilla 879. Patch from Paul Fisher.
2448 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2449 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2452 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2454 PP/05 Variable $av_failed, true if the AV scanner deferred.
2455 Bugzilla 1078. Patch from John Horne.
2457 PP/06 Stop make process more reliably on build failure.
2458 Bugzilla 1087. Patch from Heiko Schlittermann.
2460 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2461 Bugzilla 1089. Patch from Heiko Schlittermann.
2463 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2464 Bugzilla 1131. Patch from Holger Weiß.
2466 PP/09 Handle IPv6 addresses with SPF.
2467 Bugzilla 860. Patch from Wolfgang Breyha.
2469 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2471 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2474 PP/11 match_* no longer expand right-hand-side by default.
2475 New compile-time build option, EXPAND_LISTMATCH_RHS.
2476 New expansion conditions, "inlist", "inlisti".
2478 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2480 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2482 PP/14 fix log_write() format string regression from TF/03.
2483 Bugzilla 1152. Patch from Dmitry Isaikin.
2489 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2491 PP/02 Harmonised TLS library version reporting; only show if debugging.
2492 Layout now matches that introduced for other libraries in 4.74 PP/03.
2494 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2496 PP/04 New "dns_use_edns0" global option.
2498 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2501 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2502 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2504 TK/01 Updated PolarSSL code to 0.14.2.
2505 Bugzilla 1097. Patch from Andreas Metzler.
2507 PP/07 Catch divide-by-zero in ${eval:...}.
2508 Fixes bugzilla 1102.
2510 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2513 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2514 format-string attack -- SECURITY: remote arbitrary code execution.
2516 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2517 time unintentionally subject to list matching rules, letting the header
2518 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2519 arbitrary string expansion). This allowed for information disclosure.
2521 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2522 INT_MIN/-1 -- value coerced to INT_MAX.
2528 NM/01 Workaround for PCRE version dependency in version reporting
2531 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2532 This fixes portability to compilers other than gcc, notably
2533 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2535 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2536 makefiles for portability to HP-UX and POSIX correctness.
2538 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2539 Also via indented variable definition in the Makefile.
2540 (Debugging by Oliver Heesakkers).
2542 PP/02 Restore caching of spamd results with expanded spamd_address.
2543 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2545 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2546 Improves build reliability. Fix from: Frank Elsner
2548 NM/02 Fix wide character breakage in the rfc2047 coding
2549 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2551 NM/03 Allow underscore in dnslist lookups
2552 Fixes bug 1026. Patch from Graeme Fowler
2554 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2555 Code patches from Adam Ciarcinski of NetBSD.
2557 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2560 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2561 is logged, to avoid truncation. Patch from John Horne.
2563 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2564 Patch from Jakob Hirsch.
2566 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2567 SQL string expansion failure details.
2568 Patch from Andrey Oktyabrski.
2570 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2571 Patch from Simon Arlott.
2573 PP/09 New lookups functionality failed to compile on old gcc which rejects
2574 extern declarations in function scope.
2575 Patch from Oliver Fleischmann
2577 PP/10 Use sig_atomic_t for flags set from signal handlers.
2578 Check getgroups() return and improve debugging.
2579 Fixed developed for diagnosis in bug 927 (which turned out to be
2582 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2583 Patch from Mark Zealey.
2585 PP/12 Bugzilla 1056: Improved spamd server selection.
2586 Patch from Mark Zealey.
2588 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2589 Based on patch from Heiko Schlittermann.
2591 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2592 Patch from Uwe Doering, sign-off by Michael Haardt.
2594 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2595 variable declaration deep within a block. Bug and patch from
2598 PP/15 lookups-Makefile IRIX compatibility coercion.
2600 PP/16 Make DISABLE_DKIM build knob functional.
2602 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2603 Patch by Simon Arlott
2605 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2606 variable argument macros. Our copy now differs from upstream.
2612 TF/01 Failure to get a lock on a hints database can have serious
2613 consequences so log it to the panic log.
2615 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2616 controlled using the smtp_confirmation log selector.
2618 TF/03 Include the error message when we fail to unlink a spool file.
2620 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2621 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2622 for maintaining out-of-tree patches for some time.
2624 PP/01 Bugzilla 139: Documentation and portability issues.
2625 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2626 Handle per-OS dynamic-module compilation flags.
2628 PP/02 Let /dev/null have normal permissions.
2629 The 4.73 fixes were a little too stringent and complained about the
2630 permissions on /dev/null. Exempt it from some checks.
2631 Reported by Andreas M. Kirchwitz.
2633 PP/03 Report version information for many libraries, including
2634 Exim version information for dynamically loaded libraries. Created
2635 version.h, now support a version extension string for distributors
2636 who patch heavily. Dynamic module ABI change.
2638 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2639 privilege escalation vulnerability whereby the Exim run-time user
2640 can cause root to append content of the attacker's choosing to
2643 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2646 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2647 If dropping privileges for untrusted macros, we disabled normal logging
2648 on the basis that it would fail; for the Exim run-time user, this is not
2649 the case, and it resulted in successful deliveries going unlogged.
2650 Fixed. Reported by Andreas Metzler.
2656 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2657 only prepend for the Resent-* case. Fixes regression introduced in
2658 Exim 4.70 by NM/22 for Bugzilla 607.
2660 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2661 increasing numbers of administrators be bitten by this.
2663 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2665 PP/03 Bugzilla 994: added openssl_options main configuration option.
2667 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2669 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2671 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2673 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2674 without a peer certificate, leading to a segfault because of an
2675 assumption that peers always have certificates. Be a little more
2676 paranoid. Problem reported by Martin Tscholak.
2678 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2679 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2680 NB: ClamAV planning to remove STREAM in "middle of 2010".
2681 CL also introduces -bmalware, various -d+acl logging additions and
2682 more caution in buffer sizes.
2684 PP/09 Implemented reverse_ip expansion operator.
2686 PP/10 Bugzilla 937: provide a "debug" ACL control.
2688 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2690 PP/12 Bugzilla 973: Implement --version.
2692 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2694 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2696 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2698 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2699 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2700 ignore trailing whitespace.
2702 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2704 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2707 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2708 Notification from Dr Andrew Aitchison.
2710 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2711 ExtendedDetectionInfo response format.
2712 Notification from John Horne.
2714 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2717 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2718 XSL and documented dependency on system catalogs, with examples of how
2721 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2724 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2725 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2726 configuration file which is writeable by the Exim user or group.
2728 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2729 of configuration files to cover files specified with the -C option if
2730 they are going to be used with root privileges, not just the default
2733 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2734 option (effectively making it always true).
2736 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2737 files to be used while preserving root privileges.
2739 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2740 that rogue child processes cannot use them.
2742 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2743 run-time user, instead of root.
2745 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2746 Exim run-time user without dropping privileges.
2748 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2749 result string, instead of calling string_vformat() twice with the same
2752 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2753 for other users. Others should always drop root privileges if they use
2754 -C on the command line, even for a whitelisted configure file.
2756 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2758 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2764 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2765 $data_path, and $header_path variables; fixed documentation bugs and
2768 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2769 exipick to access non-standard spools, including the "frozen" queue
2772 NM/01 Bugzilla 965: Support mysql stored procedures.
2773 Patch from Alain Williams
2775 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2777 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2778 Patch from Andreas Metzler
2780 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2781 Patch from Kirill Miazine
2783 NM/05 Bugzilla 671: Added umask to procmail example.
2785 JJ/03 installed exipick 20100323.0, fixing doc bug
2787 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2788 directory. Notification and patch from Dan Rosenberg.
2790 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2792 TK/02 Improve log output when DKIM signing operation fails.
2794 MH/01 Treat the transport option dkim_domain as a colon separated
2795 list, not as a single string, and sign the message with each element,
2796 omitting multiple occurences of the same signer.
2798 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2799 Bugzilla 985, 986. Patch by Simon Arlott
2801 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2802 Patch by Simon Arlott
2804 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2805 MBX locking. Notification from Dan Rosenberg.
2811 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2813 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2815 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2817 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2819 NM/04 Bugzilla 915: Flag broken perl installation during build.
2825 TK/01 Added patch by Johannes Berg that expands the main option
2826 "spamd_address" if it starts with a dollar sign.
2828 TK/02 Write list of recipients to X-Envelope-Sender header when building
2829 the mbox-format spool file for content scanning (suggested by Jakob
2832 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2833 (http://www.dcc-servers.net/) support via dccifd. Activated by
2834 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2836 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2837 by Mark Daniel Reidel <mr@df.eu>.
2839 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2840 When building exim an external PCRE library is now needed -
2841 PCRE is a system library on the majority of modern systems.
2842 See entry on PCRE_LIBS in EDITME file.
2844 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2845 conversation. Added nologin parameter to request.
2846 Patch contributed by Kirill Miazine.
2848 TF/01 Do not log submission mode rewrites if they do not change the address.
2850 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2852 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2853 log files in place. Contributed by Roberto Lima.
2855 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2857 TF/03 Bugzilla 615: When checking the local_parts router precondition
2858 after a local_part_suffix or local_part_prefix option, Exim now
2859 does not use the address's named list lookup cache, since this
2860 contains cached lookups for the whole local part.
2862 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2863 Robert Millan. Documentation is in experimental-spec.txt.
2865 TF/04 Bugzilla 668: Fix parallel build (make -j).
2867 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2869 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2870 Patch provided by Jan Srzednicki.
2872 TF/05 Leading white space used to be stripped from $spam_report which
2873 wrecked the formatting. Now it is preserved.
2875 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2876 that they are available at delivery time.
2878 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2880 TF/08 TLS error reporting now respects the incoming_interface and
2881 incoming_port log selectors.
2883 TF/09 Produce a more useful error message if an SMTP transport's hosts
2884 setting expands to an empty string.
2886 NM/06 Bugzilla 744: EXPN did not work under TLS.
2887 Patch provided by Phil Pennock.
2889 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2890 Patch provided by Richard Godbee.
2892 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2893 acl_smtp_notquit, added index entry.
2895 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2896 Patch provided by Eugene Bujak.
2898 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2899 accept(). Patch provided by Maxim Dounin.
2901 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2902 Patch provided by Phil Pennock.
2904 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2906 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2907 Patch provided by Brad "anomie" Jorsch.
2909 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2910 Patch provided by Dean Brooks.
2912 TK/05 Add native DKIM support (does not depend on external libraries).
2914 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2915 Patch provided by Graeme Fowler.
2917 NM/16 Bugzilla 851: Documentation example syntax fix.
2919 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2921 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2924 NM/19 Bugzilla 745: TLS version reporting.
2925 Patch provided by Phil Pennock.
2927 NM/20 Bugzilla 167: bool: condition support.
2928 Patch provided by Phil Pennock.
2930 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2931 clients. Patch provided by Phil Pennock.
2933 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2934 Patch provided by Brad "anomie" Jorsch.
2936 NM/23 Bugzilla 687: Fix misparses in eximstats.
2937 Patch provided by Heiko Schlittermann.
2939 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2940 Patch provided by Heiko Schlittermann.
2942 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2943 plus update to original patch.
2945 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2947 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2948 Patch provided by David Brownlee.
2950 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2952 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2954 NM/30 Bugzilla 888: TLS documentation bugfixes.
2956 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2958 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2959 Unlike the original bugzilla I have changed all shell scripts in src tree.
2961 NM/33 Bugzilla 898: Transport filter timeout fix.
2962 Patch by Todd Rinaldo.
2964 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2965 Patch by Serge Demonchaux.
2967 NM/35 Bugzilla 39: Base64 decode bug fixes.
2968 Patch by Jakob Hirsch.
2970 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2972 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2974 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2976 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2982 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2983 ALT-N's libdkim that I have put here:
2984 http://duncanthrax.net/exim-experimental/
2986 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2987 were called 'true' and it seems that C99 defines that as a reserved
2988 keyword to be used with 'bool' variable types. That means you could
2989 not include C99-style headers which use bools without triggering
2990 build errors in sieve.c.
2992 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2993 as mailq or other aliases. Changed the --help handling significantly
2994 to do whats expected. exim_usage() emits usage/help information.
2996 SC/01 Added the -bylocaldomain option to eximstats.
2998 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3000 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3002 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3008 PH/01 Another patch from the Sieve maintainer.
3010 PH/02 When an IPv6 address is converted to a string for single-key lookup
3011 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3012 dots are used instead of colons so that keys in lsearch files need not
3013 contain colons. This was done some time before quoting was made available
3014 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3015 (notated using the quote facility) so as to distinguish them from IPv4
3016 keys. This meant that lookups for IP addresses in host lists did not work
3017 for iplsearch lookups.
3019 This has been fixed by arranging for IPv6 addresses to be expressed with
3020 colons if the lookup type is iplsearch. This is not incompatible, because
3021 previously such lookups could never work.
3023 The situation is now rather anomalous, since one *can* have colons in
3024 ordinary lsearch keys. However, making the change in all cases is
3025 incompatible and would probably break a number of configurations.
3027 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3030 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3031 conversion specification without a maximum field width, thereby enabling
3032 a rogue spamd server to cause a buffer overflow. While nobody in their
3033 right mind would setup Exim to query an untrusted spamd server, an
3034 attacker that gains access to a server running spamd could potentially
3035 exploit this vulnerability to run arbitrary code as the Exim user.
3037 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3038 $primary_hostname instead of what libspf2 thinks the hosts name is.
3040 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3041 a directory entry by the name of the lookup key. Previously, if a
3042 symlink pointed to a non-existing file or a file in a directory that
3043 Exim lacked permissions to read, a lookup for a key matching that
3044 symlink would fail. Now it is enough that a matching directory entry
3045 exists, symlink or not. (Bugzilla 503.)
3047 PH/03 The body_linecount and body_zerocount variables are now exported in the
3050 PH/04 Added the $dnslist_matched variable.
3052 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3053 This means they are set thereafter only if the connection becomes
3056 PH/06 Added the client_condition to authenticators so that some can be skipped
3057 by clients under certain conditions.
3059 PH/07 The error message for a badly-placed control=no_multiline_responses left
3060 "_responses" off the end of the name.
3062 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3064 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3065 (without spaces) instead of just copying the configuration text.
3067 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3069 PH/11 Added $max_received_linelength.
3071 PH/12 Added +ignore_defer and +include_defer to host lists.
3073 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3074 way in which PCRE > 7.0 is built.
3076 PH/14 Implemented queue_only_load_latch.
3078 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3079 MAIL command. The effect was to mangle the value on 64-bit systems.
3081 PH/16 Another patch from the Sieve maintainer.
3083 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3085 PH/18 If a system quota error occurred while trying to create the file for
3086 a maildir delivery, the message "Mailbox is full" was not appended to the
3087 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3088 only to a quota excession during the actual writing of the file.
3090 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3091 characters?) which causes problems in log lines. The DN values are now
3092 passed through string_printing() before being added to log lines.
3094 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3095 and InterBase are left for another time.)
3097 PH/21 Added message_body_newlines option.
3099 PH/22 Guard against possible overflow in moan_check_errorcopy().
3101 PH/23 POSIX allows open() to be a macro; guard against that.
3103 PH/24 If the recipient of an error message contained an @ in the local part
3104 (suitably quoted, of course), incorrect values were put in $domain and
3105 $local_part during the evaluation of errors_copy.
3111 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3112 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3115 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3116 issue a MAIL command.
3118 PH/02 In an ACL statement such as
3120 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3122 if a client was not listed at all, or was listed with a value other than
3123 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3124 the condition was not true (as it should be), so access was not denied.
3125 The bug was that the ! inversion was incorrectly passed on to the second
3126 item. This has been fixed.
3128 PH/03 Added additional dnslists conditions == and =& which are different from
3129 = and & when the dns lookup returns more than one IP address.
3131 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3132 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3134 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3135 FSYNC, which compiles an option called disable_fsync that allows for
3136 bypassing fsync(). The documentation is heavily laced with warnings.
3138 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3140 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3141 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3142 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3143 including adding "make clean"; (3) Added -fPIC when compiling the test
3144 dynamically loaded module, to get rid of a warning.
3146 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3147 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3148 = 0s. The bug is otherwise harmless.
3150 PH/07 There was a bug in the dovecot authenticator such that the value of
3151 $auth1 could be overwritten, and so not correctly preserved, after a
3152 successful authentication. This usually meant that the value preserved by
3153 the server_setid option was incorrect.
3155 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3157 PH/09 Installed PCRE release 7.0.
3159 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3160 run for batched SMTP input. It is now run at the start of every message
3161 in the batch. While fixing this I discovered that the process information
3162 (output by running exiwhat) was not always getting set for -bs and -bS
3163 input. This is fixed, and it now also says "batched" for BSMTP.
3165 PH/11 Added control=no_pipelining.
3167 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3168 patch, slightly modified), and move the expansion of helo_data till after
3169 the connection is made in the smtp transport (so it can use these
3172 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3174 PH/14 Added log_selector = +pid.
3176 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3178 PH/16 Add ${if forany and ${if forall.
3180 PH/17 Added dsn_from option to vary the From: line in DSNs.
3182 PH/18 Flush SMTP output before performing a callout, unless control =
3183 no_callout_flush is set.
3185 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3186 was true (the default) a successful delivery failed to delete the retry
3187 item, thus causing premature timeout of the address. The bug is now
3190 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3192 PH/21 Long custom messages for fakedefer and fakereject are now split up
3193 into multiline responses in the same way that messages for "deny" and
3194 other ACL rejections are.
3196 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3197 with slight modification.
3199 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3200 draft, changing the syntax and factoring some duplicate code".
3202 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3203 for deliveries of the second and subsequent messages over the same SMTP
3206 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3207 ${reduce, with only minor "tidies".
3209 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3211 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3212 expansion side effects.
3214 PH/27 When a message times out after an over-quota error from an Exim-imposed
3215 quota, the bounce message says "mailbox is full". This message was not
3216 being given when it was a system quota that was exceeded. It now should
3219 MH/03 Made $recipients available in local_scan(). local_scan() already has
3220 better access to the recipient list through recipients_list[], but
3221 $recipients can be useful in postmaster-provided expansion strings.
3223 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3224 in the case of a MAIL command with additional options following the
3225 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3226 were accidentally chopped off.
3228 PH/29 SMTP synchronization checks are implemented when a command is read -
3229 there is a check that no more input is waiting when there shouldn't be
3230 any. However, for some commands, a delay in an ACL can mean that it is
3231 some time before the response is written. In this time, more input might
3232 arrive, invalidly. So now there are extra checks after an ACL has run for
3233 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3234 pipelining has not been advertised.
3236 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3238 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3239 returned in the SMTP response when smtp_return_error_details was false.
3240 This has been fixed.
3242 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3243 instead of the C I/O that was originally supplied, because problems were
3244 reported on Solaris.
3246 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3247 Exim which did not show up earlier: it was assuming that a call to
3248 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3249 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3250 was a macro that became an assignment, so it seemed to work. This has
3251 changed to a proper function call with a void return, hence the compile
3252 error. Exim's code has been fixed.
3254 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3257 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3259 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3261 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3262 because it turns out to be too big or there is a timeout) there is no
3263 "Completed" line in the log. When some messages of this type were
3264 selected by exigrep, they were listed as "not completed". Others were
3265 picked up by some special patterns. I have improved the selection
3266 criteria to be more general.
3268 PH/38 The host_find_failed option in the manualroute router can now be set
3269 to "ignore", to completely ignore a host whose IP address cannot be
3270 found. If all hosts are ignored, the behaviour is controlled by the new
3271 host_all_ignored option.
3273 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3274 homing or because of multiple MX records with /mx) generated more than
3275 one IP address, and the following item turned out to be the local host,
3276 all the secondary addresses of the first item were incorrectly removed
3277 from the list, along with the local host and any following hosts (which
3278 is what is supposed to happen).
3280 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3281 whoever called Exim into the -H file. In the case of the daemon it was
3282 behaving confusingly. When first started, it used values for whoever
3283 started the daemon, but after a SIGHUP it used the Exim user (because it
3284 calls itself on a restart). I have changed the code so that it now always
3287 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3288 message are rejected with the same error (e.g. no authentication or bad
3289 sender address), and a DATA command is nevertheless sent (as can happen
3290 with PIPELINING or a stupid MUA), the error message that was given to the
3291 RCPT commands is included in the rejection of the DATA command. This is
3292 intended to be helpful for MUAs that show only the final error to their
3295 PH/42 Another patch from the Sieve maintainer.
3297 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3298 Eximstats - Fixed some broken HTML links and added missing column headers
3300 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3301 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3303 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3305 PH/43 Yet another patch from the Sieve maintainer.
3307 PH/44 I found a way to check for a TCP/IP connection going away before sending
3308 the response to the final '.' that terminates a message, but only in the
3309 case where the client has not sent further data following the '.'
3310 (unfortunately, this is allowed). However, in many cases there won't be
3311 any further data because there won't be any more messages to send. A call
3312 to select() can be used: if it shows that the input is "ready", there is
3313 either input waiting, or the socket has been closed. An attempt to read
3314 the next input character can distinguish the two cases. Previously, Exim
3315 would have sent an OK response which the client would never have see.
3316 This could lead to message repetition. This fix should cure that, at
3317 least in a lot of common cases.
3319 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3320 advertised in response to EHLO.
3326 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3327 fixed by 4.65/MH/01 (is this a record?) are fixed:
3329 (i) An empty string was always treated as zero by the numeric comparison
3330 operators. This behaviour has been restored.
3332 (ii) It is documented that the numeric comparison operators always treat
3333 their arguments as decimal numbers. This was broken in that numbers
3334 starting with 0 were being interpreted as octal.
3336 While fixing these problems I realized that there was another issue that
3337 hadn't been noticed. Values of message_size_limit (both the global option
3338 and the transport option) were treated as octal if they started with 0.
3339 The documentation was vague. These values are now always treated as
3340 decimal, and I will make that clear in the documentation.
3346 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3347 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3350 MH/01 Don't check that the operands of numeric comparison operators are
3351 integers when their expansion is in "skipping" mode (fixes bug
3352 introduced by 4.64-PH/07).
3354 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3355 child addresses, Exim now panics and dies. Previously, because the count
3356 is held in a short int, deliveries were likely to be lost. As such a
3357 large number of recipients for a single message is ridiculous
3358 (performance will be very, very poor), I have chosen to impose a limit
3359 rather than extend the field.
3365 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3366 leftover -K file (the existence of which was triggered by #402).
3367 While we were at it, introduced process PID as part of the -K
3368 filename. This should rule out race conditions when creating
3371 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3372 processing considerably. Previous code took too long for large mails,
3373 triggering a timeout which in turn triggers #401.
3375 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3376 in the DK code in transports.c. sendfile() is not really portable,
3377 hence the _LINUX specificness.
3379 TF/01 In the add_headers option to the mail command in an Exim filter,
3380 there was a bug that Exim would claim a syntax error in any
3381 header after the first one which had an odd number of characters
3384 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3385 callout verification, Exim cached a "reject" for the entire domain. This
3386 is correct for most verifications, but it is not correct for a recipient
3387 verification with use_sender or use_postmaster set, because in that case
3388 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3389 case of MAIL FROM:<> rejection from other early rejections (e.g.
3390 rejection of HELO). When verifying a recipient using a non-null MAIL
3391 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3392 Whatever the result of the callout, the value of the domain cache is
3393 left unchanged (for any other kind of callout, getting as far as trying
3394 RCPT means that the domain itself is ok).
3396 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3399 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3400 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3401 session, the daemon ignores ECONNECT errors and logs others; it now
3402 ignores EPIPE as well.
3404 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3405 (quoted-printable decoding).
3407 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3408 later the small subsequent patch to fix an introduced bug.
3410 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3412 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3414 PH/08 An error is now given if message_size_limit is specified negative.
3416 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3417 to be given (somewhat) arbitrary names.
3419 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3422 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3423 miscellaneous code fixes
3425 PH/10 Added the log_reject_target ACL modifier to specify where to log
3428 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3429 hostname. This is wrong, because it relates to the incoming message (and
3430 probably the interface on which it is arriving) and not to the outgoing
3431 callout (which could be using a different interface). This has been
3432 changed to use the value of the helo_data option from the smtp transport
3433 instead - this is what is used when a message is actually being sent. If
3434 there is no remote transport (possible with a router that sets up host
3435 addresses), $smtp_active_hostname is used.
3437 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3438 tweaks were necessary in order to get it to work (see also 21 below):
3439 (a) The code assumed that strncpy() returns a negative number on buffer
3440 overflow, which isn't the case. Replaced with Exim's string_format()
3442 (b) There were several signed/unsigned issues. I just did the minimum
3443 hacking in of casts. There is scope for a larger refactoring.
3444 (c) The code used strcasecmp() which is not a standard C function.
3445 Replaced with Exim's strcmpic() function.
3446 (d) The code set only $1; it now sets $auth1 as well.
3447 (e) A simple test gave the error "authentication client didn't specify
3448 service in request". It would seem that Dovecot has changed its
3449 interface. Fortunately there's a specification; I followed it and
3450 changed what the client sends and it appears to be working now.
3452 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3455 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3456 address A is aliased to B and C, where B exists and C does not. Without
3457 -v the output is "A verified" because verification stops after a
3458 successful redirection if more than one address is generated. However,
3459 with -v the child addresses are also verified. Exim was outputting "A
3460 failed to verify" and then showing the successful verification for C,
3461 with its parentage. It now outputs "B failed to verify", showing B's
3462 parentage before showing the successful verification of C.
3464 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3465 look up a TXT record in a specific list after matching in a combined
3468 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3469 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3470 they consult the DNS. I had assumed they would set it the way they
3471 wanted; and indeed my experiments on Linux seem to show that in some
3472 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3473 To be on the safe side, however, I have now made the interface to
3474 host_find_byname() similar to host_find_bydns(), with an argument
3475 containing the DNS resolver options. The host_find_byname() function now
3476 sets these options at its start, just as host_find_bydns() does. The smtp
3477 transport options dns_qualify_single and dns_search_parents are passed to
3478 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3479 of host_find_byname() use the default settings of RES_DEFNAMES
3480 (qualify_single) but not RES_DNSRCH (search_parents).
3482 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3483 spool_read_header() do less string testing, by means of a preliminary
3484 switch on the second character of optional "-foo" lines. (This is
3485 overdue, caused by the large number of possibilities that now exist.
3486 Originally there were few.) While I was there, I also converted the
3487 str(n)cmp tests so they don't re-test the leading "-" and the first
3488 character, in the hope this might squeeze out yet more improvement.
3490 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3491 flag allowing group syntax was set by the header_syntax check but not
3492 turned off, possible causing trouble later; (2) The flag was not being
3493 set at all for the header_verify test, causing "group"-style headers to
3494 be rejected. I have now set it in this case, and also caused header_
3495 verify to ignore an empty address taken from a group. While doing this, I
3496 came across some other cases where the code for allowing group syntax
3497 while scanning a header line wasn't quite right (mostly, not resetting
3498 the flag correctly in the right place). These bugs could have caused
3499 trouble for malformed header lines. I hope it is now all correct.
3501 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3502 with the "reply" argument non-NULL. The code, however (which originally
3503 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3504 but it didn't always do it. This confused somebody who was copying the
3505 code for some other use. I have removed all the tests.
3507 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3508 feature that was used to support insecure browsers during the U.S. crypto
3509 embargo. It requires special client support, and Exim is probably the
3510 only MTA that supported it -- and would never use it because real RSA is
3511 always available. This code has been removed, because it had the bad
3512 effect of slowing Exim down by computing (never used) parameters for the
3513 RSA_EXPORT functionality.
3515 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3516 authenticator to fail if there's a tab character in the incoming data
3517 (there should never be unless someone is messing about, as it's supposed
3518 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3519 if the connection is using TLS or if the remote IP is the same as the
3520 local IP, and the "valid-client-cert option" if a client certificate has
3523 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3524 authenticators. This can be used for authorization after authentication
3525 succeeds. (In the case of plaintext, it servers for both authentication
3528 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3529 if any retry times were supplied.
3531 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3532 connection, where there is no client IP address to check. In this
3533 situation, the verify now always succeeds.
3535 PH/25 Applied John Jetmore's -Mset patch.
3537 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3539 PH/27 In a string expansion for a processed (not raw) header when multiple
3540 headers of the same name were present, leading whitespace was being
3541 removed from all of them, but trailing whitespace was being removed only
3542 from the last one. Now trailing whitespace is removed from each header
3543 before concatenation. Completely empty headers in a concatenation (as
3544 before) are ignored.
3546 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3547 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3549 PH/29 [Removed. This was a change that I later backed out, and forgot to
3550 correct the ChangeLog entry (that I had efficiently created) before
3551 committing the later change.]
3553 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3554 address errors (4xx response to RCPT) over the same connection as other
3555 messages routed to the same hosts. Such deliveries are always "forced",
3556 so retry times are not inspected. This resulted in far too many retries
3557 for the affected addresses. The effect occurred only when there were more
3558 hosts than the hosts_max_try setting in the smtp transport when it had
3559 the 4xx errors. Those hosts that it had tried were not added to the list
3560 of hosts for which the message was waiting, so if all were tried, there
3561 was no problem. Two fixes have been applied:
3563 (i) If there are any address or message errors in an SMTP delivery, none
3564 of the hosts (tried or untried) are now added to the list of hosts
3565 for which the message is waiting, so the message should not be a
3566 candidate for sending over the same connection that was used for a
3567 successful delivery of some other message. This seems entirely
3568 reasonable: after all the message is NOT "waiting for some host".
3569 This is so "obvious" that I'm not sure why it wasn't done
3570 previously. Hope I haven't missed anything, but it can't do any
3571 harm, as the worst effect is to miss an optimization.
3573 (ii) If, despite (i), such a delivery is accidentally attempted, the
3574 routing retry time is respected, so at least it doesn't keep
3575 hammering the server.
3577 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3578 in ${readsocket because some servers need this prod.
3580 PH/32 Added some extra debug output when updating a wait-xxx database.
3582 PH/33 The hint "could be header name not terminated by colon", which has been
3583 given for certain expansion errors for a long time, was not being given
3584 for the ${if def:h_colon_omitted{... case.
3586 PH/34 The spec says: "With one important exception, whenever a domain list is
3587 being scanned, $domain contains the subject domain." There was at least
3588 one case where this was not true.
3590 PH/35 The error "getsockname() failed: connection reset by peer" was being
3591 written to the panic log as well as the main log, but it isn't really
3592 panic-worthy as it just means the connection died rather early on. I have
3593 removed the panic log writing for the ECONNRESET error when getsockname()
3596 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3597 runs only) independently of the message's sender address. This meant
3598 that, if the 4xx error was in fact related to the sender, a different
3599 message to the same recipient with a different sender could confuse
3600 things. In particular, this can happen when sending to a greylisting
3601 server, but other circumstances could also provoke similar problems.
3602 I have changed the default so that the retry time for these errors is now
3603 based a combination of the sender and recipient addresses. This change
3604 can be overridden by setting address_retry_include_sender=false in the
3607 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3608 remote server are returned as part of bounce messages. This was not
3609 happening for LMTP over a pipe (the lmtp transport), but now it is the
3610 same for both kinds of LMTP.
3612 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3613 in header lines that were in fact CNAMEs. This is no longer the case.
3615 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3616 and queue runs started by the daemon processed all messages. This has
3617 been fixed so that -R and -S can now usefully be given with -q<time>.
3619 PH/40 Import PCRE release 6.7 (fixes some bugs).
3621 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3623 PH/42 Give an error if -q is specified more than once.
3625 PH/43 Renamed the variables $interface_address and $interface_port as
3626 $received_ip_address and $received_port, to make it clear that these
3627 values apply to message reception, and not to the outgoing interface when
3628 a message is delivered. (The old names remain recognized, of course.)
3630 PH/44 There was no timeout on the connect() call when using a Unix domain
3631 socket in the ${readsocket expansion. There now is.
3633 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3634 be meaningful with "accept".
3636 SC/01 Eximstats V1.43
3637 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3639 SC/02 Eximstats V1.44
3640 Use a glob alias rather than an array ref in the generated
3641 parser. This improves both readability and performance.
3643 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3644 Collect SpamAssassin and rejection statistics.
3645 Don't display local sender or destination tables unless
3646 there is data to show.
3647 Added average volumes into the top table text output.
3649 SC/04 Eximstats V1.46
3650 Collect data on the number of addresses (recipients)
3651 as well as the number of messages.
3653 SC/05 Eximstats V1.47
3654 Added 'Message too big' to the list of mail rejection
3655 reasons (thanks to Marco Gaiarin).
3657 SC/06 Eximstats V1.48
3658 Mainlog lines which have GMT offsets and are too short to
3659 have a flag are now skipped.
3661 SC/07 Eximstats V1.49 (Alain Williams)
3662 Added the -emptyok flag.
3664 SC/08 Eximstats V1.50
3665 Fixes for obtaining the IP address from reject messages.
3667 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3668 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3669 whitespace changes from 4.64-PH/27
3671 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3674 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3675 are found, allow negative numbers in numeric criteria)
3677 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3679 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3680 to match changes made in 4.64-PH/43
3682 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3684 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3685 "domain" is actually an IP address, and force a failure. This locks out
3686 those revolvers/nameservers that support "A-for-A" lookups, in
3687 contravention of the specifications.
3689 PH/48 When a host name was looked up from an IP address, and the subsequent
3690 forward lookup of the name timed out, the host name was left in
3691 $sender_host_name, contrary to the specification.
3693 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3694 restricted to single-key lookups, Exim was not diagnosing an error if
3695 * or *@ was used with a query-style lookup.
3697 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3699 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3700 long ago, but noone interested enough thought of it. Let's just say that
3701 the "1.1" means that there are some new functions that weren't there at
3702 some point in the past.
3704 PH/51 Error processing for expansion failure of helo_data from an smtp
3705 transport during callout processing was broken.
3707 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3708 tested/used via the -bh/-bhc/-bs options.
3710 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3711 bug, fixed in subsequent PCRE releases).
3713 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3714 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3716 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3722 SC/01 Use a glob alias rather than an array ref in eximstats generated
3723 parser. This improves both readability and performance.
3725 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3726 Don't display local sender or destination tables in eximstats unless
3727 there is data to show.
3728 Added average volumes into the eximstats top table text output.
3730 SC/03 Collect data on the number of addresses (recipients) as well
3731 as the number of messages in eximstats.
3733 TF/01 Correct an error in the documentation for the redirect router. Exim
3734 does (usually) call initgroups() when daemonizing.
3736 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3737 with consistent privilege compared to when running as a daemon.
3739 TF/03 Note in the spec that $authenticated_id is not set for local
3740 submissions from trusted users.
3742 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3743 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3745 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3746 by adding some example configuration directives to the default
3747 configuration file. A little bit of work is required to uncomment the
3748 directives and define how usernames and passwords are checked, but
3749 there is now a framework to start from.
3751 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3752 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3753 without this. I don't know how relevant this is to other LDAP libraries.
3755 PH/02 Add the verb name to the "unknown ACL verb" error.
3757 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3759 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3761 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3762 directory not expanded when it should be if an expanded home directory
3763 was set for the address (which is overridden by the transport).
3765 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3768 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3769 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3770 because it is too late at that time, and has no effect.
3772 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3773 security issue with \' (bugzilla #107). I could not use the
3774 PQescapeStringConn() function, because it needs a PGconn value as one of
3777 PH/08 When testing addresses using -bt, indicate those final addresses that
3778 are duplicates that would not cause an additional delivery. At least one
3779 person was confused, thinking that -bt output corresponded to deliveries.
3780 (Suppressing duplicates isn't a good idea as you lose the information
3781 about possibly different redirections that led to the duplicates.)
3783 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3784 systems where poll() doesn't work, in particular OS X.
3786 PH/10 Added more information to debugging output for retry time not reached.
3788 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3789 operations in malware.c.
3791 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3794 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3795 syslog_duplication set false, log lines that would normally be written
3796 both the the main log and to the reject log were not written to syslog at
3799 PH/14 In the default configuration, change the use of "message" in ACL warn
3800 statements to "add_header".
3802 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3803 not followed by a command (e.g. "seen endif").
3805 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3806 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3809 PH/17 Added extra conditions to the default value of delay_warning_condition
3813 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3814 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3815 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3818 The Auto-Submitted: and various List- headers are standardised, whereas I
3819 don't think Precedence: ever was.
3821 PH/18 Refactored debugging code in route_finduser() to show more information,
3822 in particular, the error code if getpwnam() issues one.
3824 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3825 This is apparently needed in addition to the PH/07 change above to avoid
3826 any possible encoding problems.
3828 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3829 but not after initializing Perl.
3831 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3832 output them only if debugging. By default they are written stderr,
3833 apparently, which is not desirable.
3835 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3838 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3841 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3843 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3844 authenticated or an ident call has been made. Suppress the default
3845 values for $authenticated_id and $authenticated_sender (but permit -oMai
3846 and -oMas) when testing with -bh.
3848 PH/24 Re-jigged the order of the tests in the default configuration so that the
3849 tests for valid domains and recipients precede the DNS black list and CSA
3850 tests, on the grounds that those ones are more expensive.
3852 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3853 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3854 This bug exists in every version of Exim that I still have, right back to
3857 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3858 However, an attempt to turn on case-sensitivity in a regex key by
3859 including (?-i) didn't work because the subject string was already
3860 lowercased, and the effects were non-intuitive. It turns out that a
3861 one-line patch can be used to allow (?-i) to work as expected.
3867 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3868 other effects) broke the use of negated acl sub-conditions.
3870 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3873 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3874 "Deny" causes Exim to reject the incoming connection with a 554 error.
3875 Unfortunately, if there is a major crisis, such as a disk failure,
3876 tcp-wrappers gives "deny", whereas what one would like would be some
3877 kind of temporary error. A kludge has been added to help with this.
3878 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3879 554 error is used if errno is still zero or contains ENOENT (which occurs
3880 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3883 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3885 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3886 errors. Otherwise a message that provokes a temporary error (when other
3887 messages do not) can cause a whole host to time out.
3889 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3890 addresses were routed directly to files or pipes from a redirect router.
3891 File deliveries just didn't batch; pipe deliveries might have suffered
3894 PH/06 A failure to get a lock for a hints database would erroneously always say
3895 "Failed to get write lock", even when it was really a read lock.
3897 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3898 of 0600. This has been changed to use the value of the lockfile_mode
3899 option (which defaults to 0600).
3901 PH/08 Applied small patch from the Sieve maintainer.
3903 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3904 folder from quota calculations, a direct delivery into this folder messed
3905 up the contents of the maildirsize file. This was because the regex was
3906 used only to exclude .Trash (or whatever) when the size of the mailbox
3907 was calculated. There was no check that a delivery was happening into an
3908 excluded directory. This bug has been fixed by ignoring all quota
3909 processing for deliveries into excluded directories.
3911 PH/10 Added the maildirfolder_create_regex option to appendfile.
3917 PH/01 The code for finding all the local interface addresses on a FreeBSD
3918 system running IPv6 was broken. This may well have applied to all BSD
3919 systems, as well as to others that have similar system calls. The broken
3920 code found IPv4 interfaces correctly, but gave incorrect values for the
3921 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3922 that it would not match correctly against @[] and not recognize the IPv6
3925 PH/02 The ipliteral router was not recognizing addresses of the form user@
3926 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3928 PH/03 Added disable_ipv6.
3930 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3931 decoded form, because it is most often used to construct To: headers
3932 lines in autoreplies, and the decoded form may well be syntactically
3933 invalid. However, $reply_address has leading white space removed, and all
3934 newlines turned into spaces so that the autoreply transport does not
3937 PH/05 If group was specified without a user on a router, and no group or user
3938 was specified on a transport, the group from the router was ignored.
3940 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3941 for visible compile-time settings that can be used to change these
3942 numbers, for those that want even more. Backwards compatibility with old
3943 spool files has been maintained. However, going back to a previous Exim
3944 release will lost any variables that are in spool files.
3946 PH/07 Two small changes when running in the test harness: increase delay when
3947 passing a TCP/IP connection to a new process, in case the original
3948 process has to generate a bounce, and remove special handling of
3949 127.0.0.2 (sic), which is no longer necessary.
3951 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3952 be the same on different OS.
3954 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3957 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3958 whether --show-vars was specified or not
3960 JJ/02 exipick: Added support for new ACL variable spool format introduced
3963 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3964 syntactically invalid From: or Reply-to: line, and a filter used this to
3965 generate an autoreply, and therefore failed to obtain an address for the
3966 autoreply, Exim could try to deliver to a non-existent relative file
3967 name, causing unrelated and misleading errors. What now happens is that
3968 it logs this as a hard delivery error, but does not attempt to create a
3971 PH/11 The exinext utility has a -C option for testing purposes, but although
3972 the given file was scanned by exinext itself; it wasn't being passed on
3973 when Exim was called.
3975 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3976 an end-of-file indication when reading a command response.
3978 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3979 compiled. In many other places in Exim, IPv6 addresses are always
3980 recognized, so I have changed this. It also means that IPv4 domain
3981 literals of the form [IPV4:n.n.n.n] are now always recognized.
3983 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3984 used if the router is not running as root, for example, when verifying at
3985 ACL time, or when using -bh. The debugging output from this situation was
3986 non-existent - all you got was a failure to exec. I have made two
3989 (a) Failures to set uid/gid, the current directory, or a process leader
3990 in a subprocess such as that created by queryprogram now generate
3991 suitable debugging output when -d is set.
3993 (b) The queryprogram router detects when it is not running as root,
3994 outputs suitable debugging information if -d is set, and then runs
3995 the subprocess without attempting to change uid/gid.
3997 PH/15 Minor change to Makefile for building test_host (undocumented testing
4000 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4001 additional section of a DNS packet that returns MX or SRV records.
4002 Instead, it always explicitly searches for A/AAAA records. This avoids
4003 major problems that occur when a DNS server includes only records of one
4004 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4005 fixed another bug: if SRV records were looked up and the corresponding
4006 address records were *not* found in the additional section, the port
4007 values from the SRV records were lost.
4009 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4010 using the correct key (the original address) when searching the retry
4011 rules in order to find which one to use for generating the retry hint.
4013 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4014 adding the default one. Similarly, if it contains a Reply-To: header, the
4015 errors_reply_to option, if set, is not used.
4017 PH/19 When calculating a retry time, Exim used to measure the "time since
4018 failure" by looking at the "first failed" field in the retry record. Now
4019 it does not use this if it is later than than the arrival time of the
4020 message. Instead it uses the arrival time. This makes for better
4021 behaviour in cases where some deliveries succeed, thus re-setting the
4022 "first failed" field. An example is a quota failure for a huge message
4023 when small messages continue to be delivered. Without this change, the
4024 "time since failure" will always be short, possible causing more frequent
4025 delivery attempts for the huge message than are intended.
4026 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4028 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4029 $1, $2, $3) because the numerical variables can be reset during some
4030 expansion items (e.g. "match"), thereby losing the authentication data.
4032 PH/21 Make -bV show the size of off_t variables so that the test suite can
4033 decide whether to run tests for quotas > 2G.
4035 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4036 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4037 filecount value is greater than 2G or if a quota value is greater than 2G
4038 on a system where the size of off_t is not greater than 4, a panic error
4041 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4042 never match. The debug and -bh output now contains an explicit error
4043 message indicating a malformed IPv4 address or mask.
4045 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4046 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4047 PH/23 above applies.
4049 PH/25 Do not write to syslog when running in the test harness. The only
4050 occasion when this arises is a failure to open the main or panic logs
4051 (for which there is an explicit test).
4053 PH/26 Added the /no_tell option to "control=freeze".
4055 PH/27 If a host name lookup failed very early in a connection, for example, if
4056 the IP address matched host_lookup and the reverse lookup yielded a name
4057 that did not have a forward lookup, an error message of the form "no IP
4058 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4059 could be logged. Now it outputs the IP address instead of "NULL".
4061 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4062 allows the sender and the authenticated sender to be set when
4063 submitting a message from within Exim. Since child_open_exim() is
4064 documented for local_scan(), the new function should be too.
4066 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4067 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4068 results in an empty string is now treated as unset.
4070 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4072 PH/31 Added sender_verify_fail logging option.
4074 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4075 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4076 tidied the source and removed it altogether.
4078 PH/33 When a queue run was abandoned because the load average was too high, a
4079 log line was always written; now it is written only if the queue_run log
4080 selector is set. In addition, the log line for abandonment now contains
4081 information about the queue run such as the pid. This is always present
4082 in "start" and "stop" lines but was omitted from the "abandon" line.
4084 PH/34 Omit spaces between a header name and the colon in the error message that
4085 is given when verify = headers_syntax fails (if there are lots of them,
4086 the message gets confusing).
4088 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4089 names, as there are now some PTR records that contain slashes. This check
4090 is only to protect against broken name servers that fall over on strange
4091 characters, so the fact that it applies to all lookups doesn't matter.
4093 PH/36 Now that the new test suite is complete, we can remove some of the
4094 special code in Exim that was needed for the old test suite. For example,
4095 sorting DNS records because real resolvers return them in an arbitrary
4096 order. The new test suite's fake resolver always returns records in the
4099 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4100 bounces) except when queue_only is set, to avoid logging races between
4101 the different processes.
4103 PH/38 Panic-die if .include specifies a non-absolute path.
4105 PH/39 A tweak to the "H" retry rule from its user.
4107 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4108 a label. They prevented compilation on older perls.
4110 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4111 a warning to be raised on newish perls.
4113 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4114 on queue. Changes to match documented behaviour of showing count of
4115 messages matching specified criteria.
4117 PH/40 Changed the default ident timeout from 30s to 5s.
4119 PH/41 Added support for the use of login_cap features, on those BSD systems
4120 that have them, for controlling the resources used by pipe deliveries.
4122 PH/42 The content-scanning code uses fopen() to create files in which to put
4123 message data. Previously it was not paying any attention to the mode of
4124 the files. Exim runs with umask(0) because the rest of the code creates
4125 files with open(), and sets the required mode explicitly. Thus, these
4126 files were ending up world-writeable. This was not a big issue, because,
4127 being within the spool directory, they were not world-accessible. I have
4128 created a function called modefopen, which takes an additional mode
4129 argument. It sets umask(777), creates the file, chmods it to the required
4130 mode, then resets the umask. All the relevant calls to fopen() in the
4131 content scanning code have been changed to use this function.
4133 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4134 to 24 hours. This avoids potential overflow problems when processing G
4135 and H retry rules. I suspect nobody ever tinkers with this value.
4137 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4139 PH/45 When the plaintext authenticator is running as a client, the server's
4140 challenges are checked to ensure they are valid base64 strings. By
4141 default, the authentication attempt is cancelled if an invalid string is
4142 received. Setting client_ignore_invalid_base64 true ignores these errors.
4143 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4144 they are received. Thus, the responses can be made to depend on the
4145 challenges. If an invalid string is ignored, an empty string is placed in
4148 PH/46 Messages that are created by the autoreply transport now contains a
4149 References: header, in accordance with RFCs 2822 and 3834.
4151 PH/47 Added authenticated_sender_force to the smtp transport.
4153 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4155 PH/49 Installed latest patch from the Sieve maintainer.
4157 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4158 was also set, the appendfile transport was unnecessarily scanning a
4159 directory of message files (e.g. for maildir delivery) to find the count
4160 of files (along with the size), even though it did not need this
4161 information. It now does the scan only if it needs to find either the
4162 size of the count of files.
4164 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4166 PH/52 Two bugs concerned with error handling when the smtp transport is
4169 (i) Exim was not creating retry information for temporary errors given
4170 for individual recipients after the DATA command when the smtp transport
4171 was used in LMTP mode. This meant that they could be retried too
4172 frequently, and not timed out correctly.
4174 (ii) Exim was setting the flag that allows error details to be returned
4175 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4176 recipients that were returned after the DATA command.
4178 PH/53 This is related to PH/52, but is more general: for any failing address,
4179 when detailed error information was permitted to be returned to the
4180 sender, but the error was temporary, then after the final timeout, only
4181 "retry timeout exceeded" was returned. Now it returns the full error as
4182 well as "retry timeout exceeded".
4184 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4185 do this, and (what is worse) MTAs that accept it.
4187 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4188 will now be deprecated.
4190 PH/56 New os.c-cygwin from the Cygwin maintainer.
4192 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4193 formats (previously only available in exim formats via -bpr, -bpru,
4194 and -bpra. Now also available in native and exiqgrep formats)
4196 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4197 with very large, slow to parse queues
4199 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4201 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4203 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4204 responses to authentication challenges, though it was showing the
4205 challenges; (ii) I've removed the CR characters from the debug output for
4208 PH/58 Allow for the insertion of a newline as well as a space when a string
4209 is turned into more than one encoded-word during RFC 2047 encoding. The
4210 Sieve code now uses this.
4212 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4213 data_4xx, lost_connection, tls_required.
4215 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4216 message was being sent as an SMTP response.
4218 PH/61 Add -l and -k options to exicyclog.
4220 PH/62 When verifying, if an address was redirected to one new address, so that
4221 verification continued, and the new address failed or deferred after
4222 having set something in $address_data, the value of $address_data was not
4223 passed back to the ACL. This was different to the case when no
4224 redirection occurred. The value is now passed back in both cases.
4226 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4227 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4228 use login_cap.h, so on its own it isn't the distinguishing feature. The
4229 new name refers directly to the setclassresources() function.
4231 PH/65 Added configuration files for NetBSD3.
4233 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4235 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4236 is preferred over IPv4.
4238 PH/68 The bounce_return_message and bounce_return_body options were not being
4239 honoured for bounces generated during the reception of non-SMTP messages.
4240 In particular, this applied to messages rejected by the ACL. This bug has
4241 been fixed. However, if bounce_return_message is true and bounce_return_
4242 body is false, the headers that are returned for a non-SMTP message
4243 include only those that have been read before the error was detected.
4244 (In the case of an ACL rejection, they have all been read.)
4246 PH/69 The HTML version of the specification is now built in a directory called
4247 spec_html instead of spec.html, because the latter looks like a path with
4248 a MIME-type, and this confuses some software.
4250 PH/70 Catch two compiler warnings in sieve.c.
4252 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4253 function verify_get_ident() calls ip_connect() to connect a socket, but
4254 if the "connect()" function timed out, ip_connect() used to close the
4255 socket. However, verify_get_ident() also closes the socket later, and in
4256 between Exim writes to the log, which may get opened at this point. When
4257 the socket was closed in ip_connect(), the log could get the same file
4258 descriptor number as the socket. This naturally causes chaos. The fix is
4259 not to close the socket in ip_connect(); the socket should be closed by
4260 the function that creates it. There was only one place in the code where
4261 this was missing, in the iplookup router, which I don't think anybody now
4262 uses, but I've fixed it anyway.
4264 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4265 well as to direct DNS lookups. Otherwise the handling of names in host
4266 lists is inconsistent and therefore confusing.
4272 PH/01 Two changes to the default runtime configuration:
4274 (1) Move the checks for relay_from_hosts and authenticated clients from
4275 after to before the (commented out) DNS black list checks.
4277 (2) Add control=submission to the relay_from_hosts and authenticated
4278 clients checks, on the grounds that messages accepted by these
4279 statements are most likely to be submissions.
4281 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4283 (1) Generate an error if the third argument for the ${prvs expansion is
4286 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4289 (3) Reset the variables that are obtained from the first argument of
4290 ${prvscheck and used in the second argument before leaving the code,
4291 because their memory is reclaimed, so using them afterwards may do
4294 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4295 one (it's much easier than Tom thought :-).
4297 (5) Because of (4), we can now allow for the use of $prvscheck_result
4298 inside the third argument.
4300 PH/03 For some reason, the default setting of PATH when running a command from
4301 a pipe transport was just "/usr/bin". I have changed it to
4304 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4305 anything to be listed in the output from -bV.
4307 PH/05 When a filter generated an autoreply, the entire To: header line was
4308 quoted in the delivery log line, like this:
4310 => >A.N.Other <ano@some.domain> <original@ddress> ...
4312 This has been changed so that it extracts the operative address. There
4313 may be more than one such address. If so, they are comma-separated, like
4316 => >ano@some.domain,ona@other.domain <original@ddress> ...
4318 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4319 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4320 not being looked up in the rDNS to get a host name, Exim was showing the
4321 IP address twice in Received: lines, even though the IP addresses were
4322 identical. For example:
4324 Received: from [1.2.3.4] (helo=[1.2.3.4])
4326 However, if the real host name was known, it was omitting the HELO data
4327 if it matched the actual IP address. This has been tidied up so that it
4328 doesn't show the same IP address twice.
4330 PH/07 When both +timestamp and +memory debugging was on, the value given by
4331 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4332 called by the debug printing, thereby overwriting the timestamp buffer.
4333 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4335 PH/08 When the original message was included in an autoreply transport, it
4336 always said "this is a copy of the message, including all the headers",
4337 even if body_only or headers_only was set. It now gives an appropriate
4340 PH/09 Applied a patch from the Sieve maintainer which:
4342 o fixes some comments
4343 o adds the (disabled) notify extension core
4344 o adds some debug output for the result of if/elsif tests
4345 o points to the current vacation draft in the documentation
4346 and documents the missing references header update
4350 o fixes a bug in processing the envelope test (when testing
4351 multiple envelope elements, the last element determined the
4354 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4355 Electronic Mail") by including:
4357 Auto-submitted: auto-generated
4359 in the messages that it generates (bounce messages and others, such as
4360 warnings). In the case of bounce messages for non-SMTP messages, there was
4361 also a typo: it was using "Auto_submitted" (underscore instead of
4362 hyphen). Since every message generated by Exim is necessarily in response
4363 to another message, thes have all been changed to:
4365 Auto-Submitted: auto-replied
4367 in accordance with these statements in the RFC:
4369 The auto-replied keyword:
4371 - SHOULD be used on messages sent in direct response to another
4372 message by an automatic process,
4374 - MUST NOT be used on manually-generated messages,
4376 - MAY be used on Delivery Status Notifications (DSNs) and Message
4377 Disposition Notifications (MDNs),
4379 - MUST NOT be used on messages generated by automatic or periodic
4380 processes, except for messages which are automatic responses to
4383 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4384 to the default Received: header definition.
4386 PH/12 Added log selector acl_warn_skipped (default on).
4388 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4389 variables because (a) they are in the wrong storage pool and (b) even if
4390 they were copied, it wouldn't work properly because of the caching.
4392 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4393 checking when decoding. Apparently there are clients that generate
4394 overlong encoded strings. Why am I not surprised?
4396 PH/15 If the first argument of "${if match_address" was not empty, but did not
4397 contain an "@" character, Exim crashed. Now it writes a panic log message
4398 and treats the condition as false.
4400 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4402 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4403 extension "envelope-auth". The code is finished and in agreement with
4404 other implementations, but there is no documentation so far and in fact,
4405 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4406 not changing the active code.
4408 Print executed "if" and "elsif" statements when debugging is used. This
4409 helps a great deal to understand what a filter does.
4411 Document more things not specified clearly in RFC3028. I had all this
4412 sorted out, when out of a sudden new issues came to my mind. Oops."
4414 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4417 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4418 canonical form (as documented). However, after a host name lookup from
4419 the IP address, check_host() was doing a simple string comparison with
4420 addresses acquired from the DNS when checking that the found name did
4421 have the original IP as one of its addresses. Since any found IPv6
4422 addresses are likely to be in abbreviated form, the comparison could
4423 fail. Luckily, there already exists a function for doing the comparison
4424 by converting both addresses to binary, so now that is used instead of
4425 the text comparison.
4427 PH/20 There was another similar case to PH/19, when a complete host name was
4428 given in a host list; looking up its IP address could give an abbreviated
4429 form, whereas the current host's name might or might not be abbreviated.
4430 The same fix has been applied.
4436 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4437 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4440 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4441 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4443 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4445 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4446 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4447 Also, for "Auto-Submitted", treat anything other than "no" as
4448 non-personal, in accordance with RFC 3834. (Previously it treated
4449 anything starting "auto-" as non-personal.)
4451 TF/01 The control=submission/name=... option had a problem with syntax
4452 errors if the name included a slash character. The /name= option
4453 now slurps the rest of the string, so it can include any characters
4454 but it must come last in the list of options (after /sender_retain
4457 PH/05 Some modifications to the interface to the fake nameserver for the new
4465 TK/01 Added the "success_on_redirect" address verification option. See
4466 NewStuff for rationale and an example.
4468 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4470 PH/02 Patch to exigrep to allow it to work on syslog lines.
4472 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4473 fread() to skip over the body file's header line, because in Cygwin the
4474 header line is locked and is inaccessible.
4476 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4477 co-exist for some time) to make it clear that it is the Exim ID that is
4478 referenced, not the Message-ID: header line.
4480 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4481 string_format() function, because snprintf() does not exist on all
4484 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4485 ${stat: expansion item.
4487 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4488 protocol synchronization error", to keep the pedants happy.
4490 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4491 well as for IRIX systems, when gcc is being used. See the host.c source
4494 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4496 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4499 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4500 transport and to the smtp transport in LMTP mode.
4502 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4504 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4505 run a filter in a subprocess. This could lead to confusion in subsequent
4506 lookups in the parent process. There should also be a search_tidyup() at
4507 the end of the subprocess.
4509 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4510 only if the host matched helo_try_verify_hosts, which caused the
4511 verification to occur when the EHLO/HELO command was issued. The ACL just
4512 tested the remembered result. Now, if a previous verification attempt has
4513 not happened, "verify = helo" does it there and then.
4515 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4517 TK/03 Fix log output including CR from clamd.
4519 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4520 exist provoked a memory error which could cause a segfault.
4522 PH/15 Installed PCRE 6.2
4524 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4526 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4527 of the problem. Specifically, suggested +O2 rather than +O1 for the
4530 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4532 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4533 record, the port was not correctly logged when the outgoing_port log
4534 selector was set (it logged the transort's default port).
4536 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4537 fallback_hosts, and "hosts" in the smtp transport.
4539 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4540 host errors such as "Connection refused".
4542 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4543 authentication with radiusclient 0.4.9:
4545 - Error returned from rc_read_config was caught wrongly
4546 - Username/password not passed on to radius server due to wrong length.
4548 The presumption is that some radiusclient API changes for 4.51/PH/17
4549 were not taken care of correctly. The code is still untested by me (my
4550 Linux distribution still has 0.3.2 of radiusclient), but it was
4551 contributed by a Radius user.
4553 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4554 expanding the "port" option of the smtp transport.
4556 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4557 while reading a MIME header. Thanks to Tom Hughes for a patch.
4559 PH/24 Include config.h inside local_scan.h so that configuration settings are
4562 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4563 that in an ACL for RCPT (for example), you can examine exactly what was
4566 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4567 commands, but it was not correctly comparing the address with the actual
4568 client host address. Thus, it would show the EHLO address in Received:
4569 header lines when this was not necessary.
4571 PH/27 Added the % operator to ${eval:}.
4573 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4574 it should be ignoring failures (because with -C, for example, it has lost
4575 privilege). It wasn't ignoring creation failures other than "already
4578 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4581 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4582 because an input line was too long, either on its own, or by virtue of
4583 too many continuations, the temporary file was not being removed, and the
4584 return code was incorrect.
4586 PH/31 Missing "BOOL" in function definition in filtertest.c.
4588 PH/32 Applied Sieve patches from the maintainer.
4590 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4592 PH/33 Added "verify = not_blind".
4594 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4595 Local/Makefile (with some defaults set). These are used in built scripts
4596 such as exicyclog, but they have never been used in the exim_install
4597 script (though there are many overriding facilities there). I have
4598 arranged that the exim_install script now takes note of these two
4601 PH/35 Installed configuration files for Dragonfly.
4603 PH/36 When a locally submitted message by a trusted user did not contain a
4604 From: header, and the sender address was obtained from -f or from an SMTP
4605 MAIL command, and the trusted user did not use -F to supply a sender
4606 name, $originator_name was incorrectly used when constructing a From:
4607 header. Furthermore, $originator_name was used for submission mode
4608 messages from external hosts without From: headers in a similar way,
4609 which is clearly wrong.
4611 PH/37 Added control=suppress_local_fixups.
4613 PH/38 When log_selector = +received_sender was set, and the addition of the
4614 sender made the log line's construction buffer exactly full, or one byte
4615 less than full, an overflow happened when the terminating "\n" was
4618 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4619 when the result of a list match is failure because a DNS lookup failed.
4621 PH/40 RM_COMMAND is now used in the building process.
4623 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4624 the "build-* directories that it finds.
4626 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4627 address in a domain literal was a prefix of an interface address.
4629 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4630 when verifying a sender address, unless rewrite_headers is false.
4632 PH/44 Wrote a long comment about why errors_to addresses are verified as
4633 recipients, not senders.
4635 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4636 the ratelimit ACL was added.
4638 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4640 PH/46 Added extra information about PostgreSQL errors to the error string.
4642 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4643 avoiding the need to install special zones in a real server. This is
4644 backwards compatible; if it can't find the fake resolver, it drops back.
4645 Thus, both old and new test suites can be run.
4647 TF/02 Added util/ratelimit.pl
4649 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4650 clock is set back in time.
4652 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4653 Candler <B.Candler@pobox.com>.
4655 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4656 OK for addresses that are the result of redirections.
4658 PH/48 A number of further additions for the benefit of the new test suite,
4659 including a fake gethostbyname() that interfaces to the fake DNS resolver
4662 TF/06 The fix for widen_domains has also been applied to qualify_single and
4663 search_parents which are the other dnslookup options that can cause
4666 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4669 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4671 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4672 service" if it encountered an MX record with an empty target hostname.
4673 The message is now "an MX or SRV record indicated no SMTP service".
4675 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4676 if the DNS of the sending site is misconfigured. This is quite a
4677 common situation. This change restores the behaviour of treating a
4678 helo verification defer as a failure.
4680 PH/51 If self=fail was set on a router, the bounce message did not include the
4681 actual error message.
4687 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4689 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4690 command itself ended in error, the underlying message about the transport
4691 filter timeout was being overwritten with the pipe command error. Now the
4692 underlying error message should be appended to the second error message.
4694 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4696 PH/02 Reduce the amount of output that "make" produces by default. Full output
4697 can still be requested.
4699 PH/03 The warning log line about a condition test deferring for a "warn" verb
4700 was being output only once per connection, rather than after each
4701 occurrence (because it was using the same function as for successful
4702 "warn" verbs). This seems wrong, so I have changed it.
4704 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4705 it should not have, which might have caused a crash in the right
4706 circumstances, but probably never did.
4708 PH/04 Installed a modified version of Tony Finch's patch to make submission
4709 mode fix the return path as well as the Sender: header line, and to
4710 add a /name= option so that you can make the user's friendly name appear
4713 TF/03 Added the control = fakedefer ACL modifier.
4715 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4716 Mark Lowes for thorough testing.
4718 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4720 TK/03 Merged latest SRS patch from Miles Wilton.
4722 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4723 with the definition in sysexits.h (which is #included earlier).
4724 Fortunately, Exim does not actually use EX_OK. The code used to try to
4725 preserve the sysexits.h value, by assuming that macro definitions were
4726 scanned for macro replacements. I have been disabused of this notion,
4727 so now the code just undefines EX_OK before #including unistd.h.
4729 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4730 in the smtp transport. When a block could not be written in a single
4731 write() function, the timeout was being re-applied to each part-write.
4732 This seems wrong - if the receiver was accepting one byte at a time it
4733 would take for ever. The timeout is now adjusted when this happens. It
4734 doesn't have to be particularly precise.
4736 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4737 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4739 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4740 without a domain if the check to <postmaster@domain> fails.
4742 SC/01 Eximstats: added -xls and the ability to specify output files
4743 (patch written by Frank Heydlauf).
4745 SC/02 Eximstats: use FileHandles for outputting results.
4747 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4749 SC/04 Eximstats: fixed display of large numbers with -nvr option
4751 SC/05 Eximstats: fixed merging of reports with empty tables.
4753 SC/06 Eximstats: added the -include_original_destination flag
4755 SC/07 Eximstats: removed tabs and trailing whitespace.
4757 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4759 TK/06 MBOX spool code: Add real "From " MBOX separator line
4760 so the .eml file is really in mbox format (even though
4761 most programs do not really care). Patch from Alex Miller.
4763 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4764 The latter is generated from $received_to and is only set if the
4765 message has one envelope recipient. SA can use these headers,
4766 obviously out-of-the-box. Patch from Alex Miller.
4768 PH/08 The ${def test on a variable was returning false if the variable's
4769 value was "0", contrary to what the specification has always said!
4770 The result should be true unless the variable is empty.
4772 PH/09 The syntax error of a character other than { following "${if
4773 def:variable_name" (after optional whitespace) was not being diagnosed.
4774 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4775 accidental colon was present, for example, could give incorrect results.
4777 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4778 result is used (not including appendfile, where other changes are about
4781 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4782 This involved changing a lot of size variables from int to off_t. It
4783 should work with maildirs and everything.
4785 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4786 spamd dying while we are connected to it.
4788 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4791 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4792 function for BDB 4.3.
4794 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4796 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4797 the sources for PCRE have been re-arranged and more files are now
4800 PH/15 The code I had for printing potentially long long variables in PH/11
4801 above was not the best (it lost precision). The length of off_t variables
4802 is now inspected at build time, and an appropriate printing format (%ld
4803 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4804 to be "long long int" or "long int". This is needed for the internal
4805 formatting function string_vformat().
4807 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4808 the configuration file to be ":syslog", then the script "guesses" where
4809 the logs files are, rather than using the compiled in default. In our
4810 case the guess is not the same as the compiled default, so the script
4811 suddenly stopped working when I started to use syslog. The patch checks
4812 to see if log_file_path is "". If so, it attempts to read it from exim
4813 with no configuration file to get the compiled in version, before it
4814 falls back to the previous guessing code."
4816 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4817 implementing BATV in an Exim configuration. See NewStuff for the gory
4820 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4821 Makefile that are specific to HP-UX.
4823 PH/18 If the "use_postmaster" option was set for a recipient callout together
4824 with the "random" option, the postmaster address was used as the MAIL
4825 FROM address for the random test, but not for the subsequent recipient
4826 test. It is now used for both.
4828 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4829 patch removes a few documentation additions to RFC 3028, because the
4830 latest draft now contains them. It adds the new en;ascii-case comparator
4831 and a new error check for 8bit text in MIME parts. Comparator and
4832 require names are now matched exactly. I enabled the subaddress
4833 extension, but it is not well tested yet (read: it works for me)."
4835 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4836 rework some of the code of TK/09 above to avoid the hardwired use of
4837 "%lld" and "long long". Replaced the call to snprintf() with a call to
4840 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4841 records point to non-existent hosts", "retry timeout exceeded", and
4842 "retry time not reached for any host after a long failure period".
4844 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4845 experimental DomainKeys support:
4847 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4848 (2) On an error such as an illegally used "control", the wrong name for
4849 the control was given.
4851 These problems did NOT occur unless DomainKeys support was compiled.
4853 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4855 PH/24 Added ${if match_ip condition.
4857 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4858 databases so that it will be absolutely obvious if a crash occurs in the
4859 DB library. This is a regular occurrence (often caused by mis-matched
4862 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4863 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4864 picked up on a user's system that detects such things. There doesn't seem
4865 to be a gcc warning option for this - only an attribute that has to be
4866 put on the function's prototype. It seems that in Fedora Core 4 they have
4867 set this on a number of new functions. No doubt there will be more in due
4870 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4871 specify a transport. However, if an address that was verified by such a
4872 router was the subject of a callout, Exim crashed because it tried to
4873 read the rcpt_include_affixes from the non-existent transport. Now it
4874 just assumes that the setting of that option is false. This bug was
4875 introduced by 4.51/PH/31.
4877 PH/28 Changed -d+all to exclude +memory, because that information is very
4878 rarely of interest, but it makes the output a lot bigger. People tend to
4879 do -d+all out of habit.
4881 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4882 code in os-type was giving problems when libc.so lives in lib64, like on
4885 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4886 aren't the modern standard, and it seems that some systems' include files
4887 don't always have them. Exim was already checking for some of the newer
4888 ones like T_AAAA, and defining it itself. I've added checks for all the
4889 record types that Exim uses.
4891 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4892 not automatically generating a new one, as it is supposed to. This
4893 prevented TLS from working. If the file did exist, but contained invalid
4894 data, a new version was generated, as expected. It was only the case of a
4895 non-existent file that was broken.
4897 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4898 with a change in libdomainkeys > 0.64.
4900 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4901 from DNS. If the selector record carries the flag, it now has
4902 precedence over the domain-wide flag.
4904 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4906 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4907 the use of a port name that isn't defined in /etc/services) occurred, the
4908 message was deferred as in a normal delivery, and thus remained on the
4909 spool, instead of being failed because of the mua_wrapper setting. This
4910 is now fixed, and I tidied up some of the mua_wrapper messages at the
4913 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4914 the messages in a hash of arrays rather than using individual hashes.
4915 This is a bit cleaner and results in dramatic memory savings, albeit
4916 at a slight CPU cost.
4918 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4919 as requested by Marc Sherman.
4921 SC/10 Eximstats: added histograms for user specified patterns as requested
4924 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4926 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4927 fopen() in the content-scanning modules that did not already have it.
4933 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4934 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4936 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4938 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4940 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4941 to test Sieve filters that use "vacation".
4943 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4944 that changes the way the GnuTLS parameters are stored in the cache file.
4945 The new format can be generated externally. For backward compatibility,
4946 if the data in the cache doesn't make sense, Exim assumes it has read an
4947 old-format file, and it generates new data and writes a new file. This
4948 means that you can't go back to an older release without removing the
4951 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4952 work if there are any delivery delays because "one_time" forces the
4953 parent to be marked "delivered", so its unseen clone is never tried
4954 again. For this reason, Exim now forbids the simultaneous setting of
4957 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4958 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4959 entry if you want to know the details. The fix, however, neglected to
4960 consider the case when local delivery batching is involved. The test for
4961 "previously delivered" was not happening when checking to see if an
4962 address could be batched with a previous (undelivered) one; under
4963 certain circumstances this could lead to multiple deliveries to the same
4966 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4967 in its include files, and this causes problems building Exim.
4969 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4970 header_syntax) but Exim was just ignoring anything given after a slash.
4971 In particular, this caused confusion with an attempt to use "verify =
4972 reverse_host_lookup/defer_ok". An error is now given when options are
4973 supplied for verify items that do not have them. (Maybe reverse_host_
4974 lookup should have a defer_ok option, but that's a different point.)
4976 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4977 defined by RFC 821) to 2048, because there were problems with some AUTH
4978 commands, and RFC 1869 says the size should be increased for extended
4979 SMTP commands that take arguments.
4981 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4984 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4985 "unknown" error; now it says that the functionality isn't in the binary.
4987 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4988 an address' error message when a string expansion fails (syntax or
4989 whatever). Otherwise the password may appear in the log. Following change
4990 PH/42 below, there is no longer a chance of it appearing in a bounce
4993 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4995 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4996 of its addresses was ever tried. (Bugzilla bug #2.)
4998 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4999 the result incorrectly in the debug output. (It correctly added a newline
5000 to what was transported.)
5002 TF/01 Added $received_time.
5004 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5005 commented out examples of how to interface to a virus scanner and to
5006 SpamAssassin. Also added commented examples of av_scanner and
5007 spamd_address settings.
5009 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5010 and controls are allowed in which ACLs. There were a couple of minor
5011 errors. Some of the entries in the conditions table (which is a table of
5012 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5013 negation of where the condition IS allowed.
5015 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5017 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5018 header file does not have a version number, so I've had to invent a new
5019 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5020 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5021 radiusclient), but it was contributed by a Radius user.
5023 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5024 files or directories, for OpenSSL.
5026 PH/19 When an Exim process that is running as root has to create an Exim log
5027 file, it does so in a subprocess that runs as exim:exim so as to get the
5028 ownership right at creation (otherwise, other Exim processes might see
5029 the file with the wrong ownership). There was no test for failure of this
5030 fork() call, which would lead to the process getting stuck as it waited
5031 for a non-existent subprocess. Forks do occasionally fail when resources
5032 run out. I reviewed all the other calls to fork(); they all seem to check
5035 PH/20 When checking for unexpected SMTP input at connect time (before writing
5036 the banner), Exim was not dealing correctly with a non-positive return
5037 from the read() function. If the client had disconnected by this time,
5038 the result was a log entry for a synchronization error with an empty
5039 string after "input=" when read() returned zero. If read() returned -1
5040 (an event I could not check), uninitialized data bytes were printed.
5041 There were reports of junk text (parts of files, etc) appearing after
5044 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5046 PH/22 Added support for macro redefinition, and (re)definition in between
5047 driver and ACL definitions.
5049 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5050 forgetting to use the resulting value; it was using the unexpanded value.
5052 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5053 hadn't been configured. The fix is from Juergen Kreileder, who
5054 understands it better than I do:
5056 "Here's what I see happening with three configured cyrus_sasl
5057 authenticators configured (plain, login, cram-md5):
5059 On startup auth_cyrus_sasl_init() gets called for each of these.
5060 This means three calls to sasl_listmech() without a specified mech_list.
5061 => SASL tests which mechs of all available mechs actually work
5062 => three warnings about OTP not working
5063 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5065 With the patch, sasl_listmech() also gets called three times. But now
5066 SASL's mech_list option is set to the server_mech specified in the the
5067 authenticator. Or in other words, the answer from sasl_listmech()
5068 gets limited to just the mech you're testing for (which is different
5070 => the return list contains just 'plain' or 'login', 'cram-md5' or
5071 nothing depending on the value of ob->server_mech.
5073 I've just tested the patch: Authentication still works fine,
5074 unavailable mechs specified in the exim configuration are still
5075 caught, and the auth.log warnings about OTP are gone."
5077 PH/25 When debugging is enabled, the contents of the command line are added
5078 to the debugging output, even when log_selector=+arguments is not
5081 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5082 answer is "GNU", and only if the return is "GNU/something" is the answer
5085 PH/27 $acl_verify_message is now set immediately after the failure of a
5086 verification in an ACL, and so is available in subsequent modifiers. In
5087 particular, the message can be preserved by coding like this:
5089 warn !verify = sender
5090 set acl_m0 = $acl_verify_message
5092 Previously, $acl_verify_message was set only while expanding "message"
5093 and "log_message" when a very denied access.
5095 PH/28 Modified OS/os.c-Linux with
5097 -#ifndef OS_LOAD_AVERAGE
5098 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5100 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5101 nomenclature these days.)
5103 PH/29 Installed patch from the Sieve maintainer that adds the options
5104 sieve_useraddress and sieve_subaddress to the redirect router.
5106 PH/30 In these circumstances:
5107 . Two addresses routed to the same list of hosts;
5108 . First host does not offer TLS;
5109 . First host accepts first address;
5110 . First host gives temporary error to second address;
5111 . Second host offers TLS and a TLS session is established;
5112 . Second host accepts second address.
5113 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5114 and peerdn, if requested) that were in fact used only for the second
5117 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5118 attention to any local part prefix or suffix that was matched by the
5119 router that accepted the address. It now behaves in the same way as it
5120 does for delivery: the affixes are removed from the local part unless
5121 rcpt_include_affixes is set on the transport.
5123 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5124 timeout during the DATA phase of an incoming message.
5126 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5127 applied a patch sanctioned by the Sieve maintainer.
5129 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5130 the uid or gid is negative. A case of a negative gid caused this to be
5131 noticed. The fix allows for either to be negative.
5133 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5134 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5137 PH/36 The change PH/12 above was broken. Fixed it.
5139 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5140 the grounds that routing the same address twice would always produce the
5141 same answer. This might have been true once, but it is certainly no
5142 longer true now. Routing a child address may depend on the previous
5143 routing that produced that child. Some complicated redirection strategies
5144 went wrong when messages had multiple recipients, and made Exim's
5145 behaviour dependent on the order in which the addresses were given.
5147 I have moved the duplicate checking until after the routing is complete.
5148 Exim scans the addresses that are assigned to local and remote
5149 transports, and removes any duplicates. This means that more work will be
5150 done, as duplicates will always all be routed, but duplicates are
5151 presumably rare, so I don't expect this is of any significance.
5153 For deliveries to pipes, files, and autoreplies, the duplicate checking
5154 still happens during the routing process, since they are not going to be
5157 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5158 It corrects a timeout issue with spamd. This is Ian's comment: "The
5159 background is that sometimes spamd either never reads data from a
5160 connection it has accepted, or it never writes response data. The exiscan
5161 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5162 blindly assumes that writes won't block so it may never time out."
5164 PH/39 Allow G after quota size as well as K and M.
5166 PH/40 The value set for $authenticated_id in an authenticator may not contain
5167 binary zeroes or newlines because the value is written to log lines and
5168 to spool files. There was no check on this. Now the value is run through
5169 the string_printing() function so that such characters are converted to
5170 printable escape sequences.
5172 PH/41 $message_linecount is a new variable that contains the total number of
5173 lines in the message. Compare $body_linecount, which is the count for the
5176 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5177 bounce and delay warning messages, except in certain special cases, which
5180 (a) An SMTP error message from a remote host;
5181 (b) A message specified in a :fail: redirection;
5182 (c) A message specified in a "fail" command in a system filter;
5183 (d) A message specified in a FAIL return from the queryprogram router;
5184 (e) A message specified by the cannot_route_message router option.
5186 In these cases only, Exim does include the error details in bounce and
5187 warning messages. There are also a few cases where bland messages such
5188 as "unrouteable address" or "local delivery error" are given.
5190 PH/43 $value is now also set for the "else" part of a ${run expansion.
5192 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5193 being worked on, but at least Exim now implements the latest version to
5196 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5197 process to complete was treated as a delivery failure, a timeout while
5198 writing the message to the pipe was logged, but erroneously treated as a
5199 successful delivery. Such timeouts include transport filter timeouts. For
5200 consistency with the overall process timeout, these timeouts are now
5201 treated as errors, giving rise to delivery failures by default. However,
5202 there is now a new Boolean option for the pipe transport called
5203 timeout_defer, which, if set TRUE, converts the failures into defers for
5204 both kinds of timeout. A transport filter timeout is now identified in
5207 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5208 systems where "make" and "gmake" are different, calling "gmake" at top
5209 level broke things. I've arranged for the value of $(MAKE) to be passed
5210 from the Makefile to this script so that it can call the same version of
5214 A note about Exim versions 4.44 and 4.50
5215 ----------------------------------------
5217 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5218 changes of various kinds. As a consequence, a big documentation update was
5219 needed. This delayed the release for rather longer than seemed good, especially
5220 in the light of a couple of (minor) security issues. Therefore, the changes
5221 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5222 release. So 4.44 and 4.50 are in effect two different branches that both start
5225 I have left the 4.50 change log unchanged; it contains all the changes since
5226 4.43. The change log for 4.44 is below; many of its items are identical to
5227 those for 4.50. This seems to be the most sensible way to preserve the
5228 historical information.
5234 1. Minor wording change to the doc/README.SIEVE file.
5236 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5237 computation of the current number of files was incorrect.
5239 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5240 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5243 4. Give more explanation in the error message when the command for a transport
5244 filter fails to execute.
5246 5. There are several places where Exim runs a non-Exim command in a
5247 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5248 was being done only for the command run by the queryprogram router. It is
5249 now done for all such subprocesses. The other cases are: ${run, transport
5250 filters, and the commands run by the lmtp and pipe transports.
5252 6. Added CONFIGURE_GROUP build-time option.
5254 7. Some older OS have a limit of 256 on the maximum number of file
5255 descriptors. Exim was using setrlimit() to set 1000 as a large value
5256 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5257 systems. I've change it so that if it can't get 1000, it tries for 256.
5259 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5260 was an oversight, and furthermore, ever since the addition of extra
5261 controls (e.g. 4.43/32), the checks on when to allow different forms of
5262 "control" were broken. There should now be diagnostics for all cases when a
5263 control that does not make sense is encountered.
5265 9. Added the /retain_sender option to "control=submission".
5267 10. $recipients is now available in the predata ACL (oversight).
5269 11. Tidy the search cache before the fork to do a delivery from a message
5270 received from the command line. Otherwise the child will trigger a lookup
5271 failure and thereby defer the delivery if it tries to use (for example) a
5272 cached ldap connection that the parent has called unbind on.
5274 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5275 of $address_data from the recipient verification was clobbered by the
5276 sender verification.
5278 13. The value of address_data from a sender verification is now available in
5279 $sender_address_data in subsequent conditions in the ACL statement.
5281 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5283 15. Added a new option "connect=<time>" to callout options, to set a different
5286 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5287 was its contents. (It was OK if the option was not defined at all.)
5289 17. A "Completed" log line is now written for messages that are removed from
5290 the spool by the -Mrm option.
5292 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5293 information about exactly what failed.
5295 19. Added -dd to debug only the daemon process.
5297 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5298 handles timeouts, both on the server side and network timeouts. Renamed the
5299 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5301 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5302 It is now set to "smtps".
5304 22. $host_address is now set to the target address during the checking of
5305 ignore_target_hosts.
5307 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5308 being passed; this would have caused $sender_host_name to have been used if
5309 matching the list had actually called for a host name (not very likely,
5310 since this list is usually IP addresses). A host name is now passed as
5313 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5314 code that specifies a non-restarting handler (typically sigaction() in
5315 modern systems) in an attempt to fix a rare and obscure crash bug.
5317 25. Narrowed the window for a race in the daemon that could cause it to ignore
5318 SIGCHLD signals. This is not a major problem, because they are used only to
5319 wake it up if nothing else does.
5321 26. A malformed maildirsize file could cause Exim to calculate negative values
5322 for the mailbox size or file count. Odd effects could occur as a result.
5323 The maildirsize information is now recalculated if the size or filecount
5326 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5327 support for a long time. Removed HAVE_SYS_VFS_H.
5329 28. Installed the latest version of exipick from John Jetmore.
5331 29. In an address list, if the pattern was not a regular expression, an empty
5332 subject address (from a bounce message) matched only if the pattern was an
5333 empty string. Non-empty patterns were not even tested. This was the wrong
5334 because it is perfectly reasonable to use an empty address as part of a
5335 database query. An empty address is now tested by patterns that are
5336 lookups. However, all the other forms of pattern expect the subject to
5337 contain a local part and a domain, and therefore, for them, an empty
5338 address still always fails if the pattern is not itself empty.
5340 30. Exim went into a mad DNS loop when attempting to do a callout where the
5341 host was specified on an smtp transport, and looking it up yielded more
5342 than one IP address.
5344 31. Re-factored the code for checking spool and log partition space into a
5345 function that finds that data and another that does the check. The former
5346 is then used to implement four new variables: $spool_space, $log_space,
5347 $spool_inodes, and $log_inodes.
5349 32. The RFC2047 encoding function was originally intended for short strings
5350 such as real names; it was not keeping to the 75-character limit for
5351 encoded words that the RFC imposes. It now respects the limit, and
5352 generates multiple encoded words if necessary. To be on the safe side, I
5353 have increased the buffer size for the ${rfc2047: expansion operator from
5356 33. It is now permitted to omit both strings after an "if" condition; if the
5357 condition is true, the result is "true". As before, when the second string
5358 is omitted, a false condition yields an empty string. This makes it less
5359 cumbersome to write custom ACL and router conditions.
5361 34. Failure to deliver a bounce message always caused it to be frozen, even if
5362 there was an errors_to setting on the router. The errors_to setting is now
5365 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5366 canonical form (fully expanded) before being placed in
5367 $sender_host_address.
5369 36. The table in the code that translates DNS record types into text (T_A to
5370 "A" for instance) was missing entries for NS and CNAME. It is just possible
5371 that this could have caused confusion if both these types were looked up
5372 for the same domain, because the text type is used as part of Exim's
5373 per-process caching. But the chance of anyone hitting this buglet seems
5376 37. The dnsdb lookup has been extended in a number of ways.
5378 (1) There is a new type, "zns", which walks up the domain tree until it
5379 finds some nameserver records. It should be used with care.
5381 (2) There is a new type, "mxh", which is like "mx" except that it returns
5382 just the host names, not the priorities.
5384 (3) It is now possible to give a list of domains (or IP addresses) to be
5385 looked up. The behaviour when one of the lookups defers can be
5386 controlled by a keyword.
5388 (4) It is now possible to specify the separator character for use when
5389 multiple records are returned.
5391 38. The dnslists ACL condition has been extended: it is now possible to supply
5392 a list of IP addresses and/or domains to be looked up in a particular DNS
5395 39. Added log_selector=+queue_time_overall.
5397 40. When running the queue in the test harness, wait just a tad after forking a
5398 delivery process, to get repeatability of debugging output.
5400 41. Include certificate and key file names in error message when GnuTLS fails
5401 to set them up, because the GnuTLS error message doesn't include the name
5402 of the failing file when there is a problem reading it.
5404 42. Allow both -bf and -bF in the same test run.
5406 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5408 44. The "Exiscan patch" is now merged into the mainline Exim source.
5410 45. Sometimes the final signoff response after QUIT could fail to get
5411 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5412 < 0 before doing a fflush(). This bug looks as though it goes back to the
5413 introduction of TLS in release 3.20, but "sometimes" must have been rare
5414 because the tests only now provoked it.
5416 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5417 (this can affect the format of dates).
5419 47. exim_tidydb, when checking for the continued existence of a message for
5420 which it has found a message-specific retry record, was not finding
5421 messages that were in split spool directories. Consequently, it was
5422 deleting retry records that should have stayed in existence.
5424 48. Steve fixed some bugs in eximstats.
5426 49. The SPA authentication driver was not abandoning authentication and moving
5427 on to the next authenticator when an expansion was forced to fail,
5428 contradicting the general specification for all authenticators. Instead it
5429 was generating a temporary error. It now behaves as specified.
5431 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5432 (the order specifies the preference for clients). The order is now AES256,
5433 AES128, 3DES, ARCFOUR128.
5435 51. Small patch to Sieve code - explicitly set From: when generating an
5438 52. Exim crashed if a remote delivery caused a very long error message to be
5439 recorded - for instance if somebody sent an entire SpamAssassin report back
5440 as a large number of 550 error lines. This bug was coincidentally fixed by
5441 increasing the size of one of Exim's internal buffers (big_buffer) that
5442 happened as part of the Exiscan merge. However, to be on the safe side, I
5443 have made the code more robust (and fixed the comments that describe what
5446 53. Now that there can be additional text after "Completed" in log lines (if
5447 the queue_time_overall log selector is set), a one-byte patch to exigrep
5448 was needed to allow it to recognize "Completed" as not the last thing in
5451 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5452 patch that reportedly fixes this has been added. I am not expert enough to
5453 create a test for it. This is what the patch creator wrote:
5455 "I found a little strange behaviour of ldap code when working with
5456 Windows 2003 AD Domain, where users was placed in more than one
5457 Organization Units. When I tried to give exim partial DN, the exit code
5458 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5459 But simultaneously result of request was absolutely normal ldap result,
5460 so I produce this patch..."
5462 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5463 REFERENCE, so I have modified the code to exclude the patch when that macro
5466 55. Some experimental protocols are using DNS PTR records for new purposes. The
5467 keys for these records are domain names, not reversed IP addresses. The
5468 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5469 leaves it alone. Component reversal etc. now happens only for IP addresses.
5472 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5474 57. Double the size of the debug message buffer (to 2048) so that more of very
5475 long debug lines gets shown.
5477 58. The exicyclog utility now does better if the number of log files to keep
5478 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5480 59. Two changes related to the smtp_active_hostname option:
5482 (1) $smtp_active_hostname is now available as a variable.
5483 (2) The default for smtp_banner uses $smtp_active_hostname instead
5484 of $primary_hostname.
5486 60. The host_aton() function is supposed to be passed a string that is known
5487 to be a valid IP address. However, in the case of IPv6 addresses, it was
5488 not checking this. This is a hostage to fortune. Exim now panics and dies
5489 if the condition is not met. A case was found where this could be provoked
5490 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5491 components; fortuitously, this particular loophole had already been fixed
5492 by change 4.50/55 above.
5494 If there are any other similar loopholes, the new check in host_aton()
5495 itself should stop them being exploited. The report I received stated that
5496 data on the command line could provoke the exploit when Exim was running as
5497 exim, but did not say which command line option was involved. All I could
5498 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5499 running as the user.
5502 61. There was a buffer overflow vulnerability in the SPA authentication code
5503 (which came originally from the Samba project). I have added a test to the
5504 spa_base64_to_bits() function which I hope fixes it.
5507 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5508 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5510 63. The daemon start-up calls getloadavg() while still root for those OS that
5511 need the first call to be done as root, but it missed one case: when
5512 deliver_queue_load_max is set with deliver_drop_privilege. This is
5513 necessary for the benefit of the queue runner, because there is no re-exec
5514 when deliver_drop_privilege is set.
5516 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5517 This has been fixed.
5519 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5520 was in use, was not putting the data itself into the right store pool;
5521 consequently, it could be overwritten for a subsequent message in the same
5522 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5525 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5527 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5528 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5529 the function were treating the return as a boolean value, which happened to
5530 work because 0=false and not-0=true, but is not correct code.
5532 68. The host_aton() function was not handling scoped IPv6 addresses (those
5533 with, for example, "%eth0" on the end) correctly.
5535 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5536 negated items (that is, ~something) in unsigned ints. Some compilers
5537 apparently mutter when there is no cast.
5539 70. If an address verification called from an ACL failed, and did not produce a
5540 user-specific message (i.e. there was only a "system" message), nothing was
5541 put in $acl_verify_message. In this situation, it now puts the system
5544 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5545 session; change 4.31/43 added the unwanted input to the log line - except
5546 that it did not do this in the start of session case. It now does.
5548 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5549 This is wrong and can cause the other end to generate a synchronization
5550 error if it is another Exim or anything else that does the synchronization
5551 check. A QUIT command is no longer sent after a timeout.
5553 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5554 during host lookups.
5556 74. The defer_ok option of callout verification was not working if it was used
5557 when verifying addresses in header lines, that is, for this case:
5559 verify = header_sender/callout=defer_ok
5561 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5562 those file descriptors could be used for SMTP connections. If anything
5563 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5564 could be sent to the SMTP client, causing chaos. The daemon now opens
5565 stdin, stdout, and stderr to /dev/null when it puts itself into the
5568 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5569 log by default. The user can override this with suitable Perl magic.
5571 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5572 the log message when discard triggers, was not working for the DATA ACL or
5573 for the non-SMTP ACL.
5575 78. Error message wording change in sieve.c.
5577 79. If smtp_accept_max_per_host was set, the number of connections could be
5578 restricted to fewer than expected, because the daemon was trying to set up
5579 a new connection before checking whether the processes handling previous
5580 connections had finished. The check for completed processes is now done
5581 earlier. On busy systems, this bug wouldn't be noticed because something
5582 else would have woken the daemon, and it would have reaped the completed
5585 80. If a message was submitted locally by a user whose login name contained one
5586 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5587 It caused a spool format error. I have fixed the spool reading code. A
5588 related problem was that the "from" clause in the Received: line became
5589 illegal because of the space(s). It is now covered by ${quote_local_part.
5591 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5594 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5596 83. Patch from Sieve maintainer to fix unterminated string problem in
5597 "vacation" handling.
5599 84. Some minor changes to the Linux configuration files to help with other
5600 OS variants using glibc.
5602 85. One more patch for Sieve to update vacation handling to latest spec.
5605 ----------------------------------------------------
5606 See the note above about the 4.44 and 4.50 releases.
5607 ----------------------------------------------------
5613 1. Change 4.43/35 introduced a bug that caused file counts to be
5614 incorrectly computed when quota_filecount was set in an appendfile
5617 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5618 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5621 3. Give more explanation in the error message when the command for a transport
5622 filter fails to execute.
5624 4. There are several places where Exim runs a non-Exim command in a
5625 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5626 was being done only for the command run by the queryprogram router. It is
5627 now done for all such subprocesses. The other cases are: ${run, transport
5628 filters, and the commands run by the lmtp and pipe transports.
5630 5. Some older OS have a limit of 256 on the maximum number of file
5631 descriptors. Exim was using setrlimit() to set 1000 as a large value
5632 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5633 systems. I've change it so that if it can't get 1000, it tries for 256.
5635 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5636 was an oversight, and furthermore, ever since the addition of extra
5637 controls (e.g. 4.43/32), the checks on when to allow different forms of
5638 "control" were broken. There should now be diagnostics for all cases when a
5639 control that does not make sense is encountered.
5641 7. $recipients is now available in the predata ACL (oversight).
5643 8. Tidy the search cache before the fork to do a delivery from a message
5644 received from the command line. Otherwise the child will trigger a lookup
5645 failure and thereby defer the delivery if it tries to use (for example) a
5646 cached ldap connection that the parent has called unbind on.
5648 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5649 of $address_data from the recipient verification was clobbered by the
5650 sender verification.
5652 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5653 was its contents. (It was OK if the option was not defined at all.)
5655 11. A "Completed" log line is now written for messages that are removed from
5656 the spool by the -Mrm option.
5658 12. $host_address is now set to the target address during the checking of
5659 ignore_target_hosts.
5661 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5662 being passed; this would have caused $sender_host_name to have been used if
5663 matching the list had actually called for a host name (not very likely,
5664 since this list is usually IP addresses). A host name is now passed as
5667 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5668 code that specifies a non-restarting handler (typically sigaction() in
5669 modern systems) in an attempt to fix a rare and obscure crash bug.
5671 15. Narrowed the window for a race in the daemon that could cause it to ignore
5672 SIGCHLD signals. This is not a major problem, because they are used only to
5673 wake it up if nothing else does.
5675 16. A malformed maildirsize file could cause Exim to calculate negative values
5676 for the mailbox size or file count. Odd effects could occur as a result.
5677 The maildirsize information is now recalculated if the size or filecount
5680 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5681 support for a long time. Removed HAVE_SYS_VFS_H.
5683 18. Updated exipick to current release from John Jetmore.
5685 19. Allow an empty sender to be matched against a lookup in an address list.
5686 Previously the only cases considered were a regular expression, or an
5689 20. Exim went into a mad DNS lookup loop when doing a callout where the
5690 host was specified on the transport, if the DNS lookup yielded more than
5693 21. The RFC2047 encoding function was originally intended for short strings
5694 such as real names; it was not keeping to the 75-character limit for
5695 encoded words that the RFC imposes. It now respects the limit, and
5696 generates multiple encoded words if necessary. To be on the safe side, I
5697 have increased the buffer size for the ${rfc2047: expansion operator from
5700 22. Failure to deliver a bounce message always caused it to be frozen, even if
5701 there was an errors_to setting on the router. The errors_to setting is now
5704 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5705 canonical form (fully expanded) before being placed in
5706 $sender_host_address.
5708 24. Updated eximstats to version 1.33
5710 25. Include certificate and key file names in error message when GnuTLS fails
5711 to set them up, because the GnuTLS error message doesn't include the name
5712 of the failing file when there is a problem reading it.
5714 26. Expand error message when OpenSSL has problems setting up cert/key files.
5717 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5718 (this can affect the format of dates).
5720 28. exim_tidydb, when checking for the continued existence of a message for
5721 which it has found a message-specific retry record, was not finding
5722 messages that were in split spool directories. Consequently, it was
5723 deleting retry records that should have stayed in existence.
5725 29. eximstats updated to version 1.35
5726 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5727 1.35 - bugfix such that pie charts by volume are generated correctly
5729 30. The SPA authentication driver was not abandoning authentication and moving
5730 on to the next authenticator when an expansion was forced to fail,
5731 contradicting the general specification for all authenticators. Instead it
5732 was generating a temporary error. It now behaves as specified.
5734 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5735 (the order specifies the preference for clients). The order is now AES256,
5736 AES128, 3DES, ARCFOUR128.
5738 31. Small patch to Sieve code - explicitly set From: when generating an
5741 32. Exim crashed if a remote delivery caused a very long error message to be
5742 recorded - for instance if somebody sent an entire SpamAssassin report back
5743 as a large number of 550 error lines. This bug was coincidentally fixed by
5744 increasing the size of one of Exim's internal buffers (big_buffer) that
5745 happened as part of the Exiscan merge. However, to be on the safe side, I
5746 have made the code more robust (and fixed the comments that describe what
5749 33. Some experimental protocols are using DNS PTR records for new purposes. The
5750 keys for these records are domain names, not reversed IP addresses. The
5751 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5752 leaves it alone. Component reversal etc. now happens only for IP addresses.
5755 34. The host_aton() function is supposed to be passed a string that is known
5756 to be a valid IP address. However, in the case of IPv6 addresses, it was
5757 not checking this. This is a hostage to fortune. Exim now panics and dies
5758 if the condition is not met. A case was found where this could be provoked
5759 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5760 components; fortuitously, this particular loophole had already been fixed
5761 by change 4.50/55 or 4.44/33 above.
5763 If there are any other similar loopholes, the new check in host_aton()
5764 itself should stop them being exploited. The report I received stated that
5765 data on the command line could provoke the exploit when Exim was running as
5766 exim, but did not say which command line option was involved. All I could
5767 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5768 running as the user.
5771 35. There was a buffer overflow vulnerability in the SPA authentication code
5772 (which came originally from the Samba project). I have added a test to the
5773 spa_base64_to_bits() function which I hope fixes it.
5776 36. The daemon start-up calls getloadavg() while still root for those OS that
5777 need the first call to be done as root, but it missed one case: when
5778 deliver_queue_load_max is set with deliver_drop_privilege. This is
5779 necessary for the benefit of the queue runner, because there is no re-exec
5780 when deliver_drop_privilege is set.
5782 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5783 was in use, was not putting the data itself into the right store pool;
5784 consequently, it could be overwritten for a subsequent message in the same
5785 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5788 38. Sometimes the final signoff response after QUIT could fail to get
5789 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5790 < 0 before doing a fflush(). This bug looks as though it goes back to the
5791 introduction of TLS in release 3.20, but "sometimes" must have been rare
5792 because the tests only now provoked it.
5798 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5799 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5800 Now it returns an integer. A number of calls were still expecting a T/F
5801 return. Fortuitously, in all cases, the tests worked in OK situations,
5802 which is the norm. However, things would have gone wrong on any write
5803 failures on the smtp file descriptor. This function is used when sending
5804 messages over SMTP and also when doing verify callouts.
5806 2. When Exim is called to do synchronous delivery of a locally submitted
5807 message (the -odf or -odi options), it no longer closes stderr before doing
5810 3. Implemented the mua_wrapper option.
5812 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5814 5. Implemented the functions header_remove(), header_testname(),
5815 header_add_at_position(), and receive_remove_recipient(), and exported them
5818 6. If an ACL "warn" statement specified the addition of headers, Exim already
5819 inserted X-ACL-Warn: at the start if there was no header name. However, it
5820 was not making this test for the second and subsequent header lines if
5821 there were newlines in the string. This meant that an invalid header could
5822 be inserted if Exim was badly configured.
5824 7. Allow an ACL "warn" statement to add header lines at the start or after all
5825 the Received: headers, as well as at the end.
5827 8. Added the rcpt_4xx retry error code.
5829 9. Added postmaster_mailfrom=xxx to callout verification option.
5831 10. Added mailfrom=xxxx to the callout verification option, for verify=
5834 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5835 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5837 12. Inserted some casts to stop certain compilers warning when using pointer
5838 differences as field lengths or precisions in printf-type calls (mostly
5839 affecting debugging statements).
5841 13. Added optional readline() support for -be (dynamically loaded).
5843 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5844 same clock tick as a message's arrival, so that its received time was the
5845 same as the "first fail" time on the retry record, and that message
5846 remained on the queue past the ultimate address timeout, every queue runner
5847 would try a delivery (because it was past the ultimate address timeout) but
5848 after another failure, the ultimate address timeout, which should have then
5849 bounced the address, did not kick in. This was a "< instead of <=" error;
5850 in most cases the first failure would have been in the next clock tick
5851 after the received time, and all would be well.
5853 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5854 being recognized when the domain list was tested by the match_domain
5855 condition in an expansion string.
5857 16. Added the ${str2b64: operator.
5859 17. Exim was always calling setrlimit() to set a large limit for the number of
5860 processes, without checking whether the existing limit was already
5861 adequate. (It did check for the limit on file descriptors.) Furthermore,
5862 errors from getrlimit() and setrlimit() were being ignored. Now they are
5863 logged to the main and panic logs, but Exim does carry on, to try to do its
5864 job under whatever limits there are.
5866 18. Imported PCRE 5.0.
5868 19. Trivial typo in log message " temporarily refused connection" (the leading
5871 20. If the log selector return_path_on_delivery was set and an address was
5872 redirected to /dev/null, the delivery process crashed because it assumed
5873 that a return path would always be set for a "successful" delivery. In this
5874 case, the whole delivery is bypassed as an optimization, and therefore no
5877 21. Internal re-arrangement: the function for sending a challenge and reading
5878 a response while authentication was assuming a zero-terminated challenge
5879 string. It's now changed to take a pointer and a length, to allow for
5880 binary data in such strings.
5882 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5884 23. Exim was not respecting finduser_retries when seeking the login of the
5885 uid under which it was called; it was always trying 10 times. (The default
5886 setting of finduser_retries is zero.) Also, it was sleeping after the final
5887 failure, which is pointless.
5889 24. Implemented tls_on_connect_ports.
5891 25. Implemented acl_smtp_predata.
5893 26. If the domain in control=submission is set empty, Exim assumes that the
5894 authenticated id is a complete email address when it generates From: or
5895 Sender: header lines.
5897 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5898 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5899 chown and chgrp in /bin and hostname in /usr/bin.
5901 28. Exim was keeping the "process log" file open after each use, just as it
5902 does for the main log. This opens the possibility of it remaining open for
5903 long periods when the USR1 signal hits a daemon. Occasional processlog
5904 errors were reported, that could have been caused by this. Anyway, it seems
5905 much more sensible not to leave this file open at all, so that is what now
5908 29. The long-running daemon process does not normally write to the log once it
5909 has entered its main loop, and it closes the log before doing so. This is
5910 so that log files can straightforwardly be renamed and moved. However,
5911 there are a couple of unusual error situations where the daemon does write
5912 log entries, and I had neglected to close the log afterwards.
5914 30. The text of an SMTP error response that was received during a remote
5915 delivery was being truncated at 512 bytes. This is too short for some of
5916 the long messages that one sometimes sees. I've increased the limit to
5919 31. It is now possible to make retry rules that apply only when a message has a
5920 specific sender, in particular, an empty sender.
5922 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5923 it possible to be selective about when SMTP synchronization is enforced.
5925 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5927 32. Implemented hosts_connection_nolog.
5929 33. Added an ACL for QUIT.
5931 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5934 35. Added mailbox_size and mailbox_filecount to appendfile.
5936 36. Added control = no_multiline_responses to ACLs.
5938 37. There was a bug in the logic of the code that waits for the clock to tick
5939 in the case where the clock went backwards by a substantial amount such
5940 that the microsecond fraction of "now" was more than the microsecond
5941 fraction of "then" (but the whole seconds number was less).
5943 38. Added support for the libradius Radius client library this is found on
5944 FreeBSD (previously only the radiusclient library was supported).
5950 1. When certain lookups returned multiple values in the form name=value, the
5951 quoting of the values was not always being done properly. Specifically:
5952 (a) If the value started with a double quote, but contained no whitespace,
5954 (b) If the value contained whitespace other than a space character (i.e.
5955 tabs or newlines or carriage returns) it was not quoted.
5956 This fix has been applied to the mysql and pgsql lookups by writing a
5957 separate quoting function and calling it from the lookup code. The fix
5958 should probably also be applied to nisplus, ibase and oracle lookups, but
5959 since I cannot test any of those, I have not disturbed their existing code.
5961 2. A hit in the callout cache for a specific address caused a log line with no
5962 reason for rejecting RCPT. Now it says "Previous (cached) callout
5963 verification failure".
5965 3. There was an off-by-one bug in the queryprogram router. An over-long
5966 return line was truncated at 256 instead of 255 characters, thereby
5967 overflowing its buffer with the terminating zero. As well as fixing this, I
5968 have increased the buffer size to 1024 (and made a note to document this).
5970 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5971 when Exim is waiting for an SMTP response from a remote server, Exim
5972 restarts its select() call on the socket, thereby resetting its timeout.
5973 This is not a problem when such interrupts are rare. Somebody set up a cron
5974 job to run exiwhat every 2 minutes, which is less than the normal select()
5975 timeout (5 or 10 minutes). This meant that the select() timeout never
5976 kicked in because it was always reset. I have fixed this by comparing the
5977 time when an interrupt arrives with the time at the start of the first call
5978 to select(). If more time than the timeout has elapsed, the interrupt is
5979 treated as a timeout.
5981 5. Some internal re-factoring in preparation for the addition of Sieve
5982 extensions (by MH). In particular, the "personal" test is moved to a
5983 separate function, and given an option for scanning Cc: and Bcc: (which is
5984 not set for Exim filters).
5986 6. When Exim created an email address using the login of the caller as the
5987 local part (e.g. when creating a From: or Sender: header line), it was not
5988 quoting the local part when it contained special characters such as @.
5990 7. Installed new OpenBSD configuration files.
5992 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5993 try to make them clearer.
5995 9. Callout options, other than the timeout value, were being ignored when
5996 verifying sender addresses in header lines. For example, when using
5998 verify = header_sender/callout=no_cache
6000 the cache was (incorrectly) being used.
6002 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6003 only the Cygwin environment.
6005 11. When return_path_on_delivery was set as a log selector, if different remote
6006 addresses in the same message used different return paths and parallel
6007 remote delivery occurred, the wrong values would sometimes be logged.
6008 (Whenever a remote delivery process finished, the return path value from
6009 the most recently started remote delivery process was logged.)
6011 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6012 lines when AUTH and/or TLS are in use. This is the "received protocol"
6013 field. Exim used to use "asmtp" for authenticated SMTP, without any
6014 indication (in the protocol name) for TLS use. Now it follows the RFC and
6015 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6016 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6017 names appear in log lines as well as in Received: header lines.
6019 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6020 extensions, and comparison tests, and to fix some bugs.
6022 14. Changes to the "personal" filter test:
6024 (1) The test was buggy in that it was just doing the equivalent of
6025 "contains" tests on header lines. For example, if a user's address was
6026 anne@some.where, the "personal" test would incorrectly be true for
6028 To: susanne@some.where
6030 This test is now done by extracting each address from the header in turn,
6031 and checking the entire address. Other tests that are part of "personal"
6032 are now done using regular expressions (for example, to check local parts
6033 of addresses in From: header lines).
6035 (2) The list of non-personal local parts in From: addresses has been
6036 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6037 taken from the Sieve specification recommendations.
6039 (3) If the message contains any header line starting with "List-" it is
6040 treated as non-personal.
6042 (4) The test for "circular" in the Subject: header line has been removed
6043 because it now seems ill-conceived.
6045 15. Minor typos in src/EDITME comments corrected.
6047 16. Installed latest exipick from John Jetmore.
6049 17. If headers_add on a router specified a text string that was too long for
6050 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6051 of string_sprintf() is now avoided.
6053 18. $message_body_size was not set (it was always zero) when running the DATA
6054 ACL and the local_scan() function.
6056 19. For the "mail" command in an Exim filter, no default was being set for
6057 the once_repeat time, causing a random time value to be used if "once" was
6058 specified. (If the value happened to be <= 0, no repeat happened.) The
6059 default is now 0s, meaning "never repeat". The "vacation" command was OK
6060 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6061 (I found it when inspecting the code).
6063 20. There is now an overall timeout for performing a callout verification. It
6064 defaults to 4 times the callout timeout, which applies to individual SMTP
6065 commands during the callout. The overall timeout applies when there is more
6066 than one host that can be tried. The timeout is checked before trying the
6067 next host. This prevents very long delays if there are a large number of
6068 hosts and all are timing out (e.g. when the network connections are timing
6069 out). The value of the overall timeout can be changed by specifying an
6070 additional sub-option for "callout", called "maxwait". For example:
6072 verify = sender/callout=5s,maxwait=20s
6074 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6075 to the end before writing, but this should make it even safer).
6077 22. Exim was forgetting that it had advertised PIPELINING for the second and
6078 subsequent messages on an SMTP connection. It was also not resetting its
6079 memory on STARTTLS and an internal HELO.
6081 23. When Exim logs an SMTP synchronization error within a session, it now
6082 records whether PIPELINING has been advertised or not.
6084 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6085 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6086 rather than long int.
6088 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6090 26. Added the never_mail option to autoreply.
6096 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6097 crash if the getsockname() call failed; this can happen if a connection is
6098 closed very soon after it is established. The problem was simply in the
6099 order in which certain operations were done, causing Exim to try to write
6100 to the SMTP stream before it had set up the file descriptor. The bug has
6101 been fixed by making things happen in the correct order.
6107 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6108 before the connection was closed, thus losing the rejection response.
6110 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6111 some early Solaris releases, but causes trouble in current releases where
6112 socklen_t is defined.
6114 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6117 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6120 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6121 that releasing the top of it at the end releases what was used for sub-
6122 expansions (unless the block got too big). However, discard this block if
6123 the first thing is a variable or header, so that we can use its block when
6124 it is dynamic (useful for very large $message_headers, for example).
6126 6. Lookups now cache *every* query, not just the most recent. A new, separate
6127 store pool is used for this. It can be recovered when all lookup caches are
6128 flushed. Lookups now release memory at the end of their result strings.
6129 This has involved some general refactoring of the lookup sources.
6131 7. Some code has been added to the store_xxx() functions to reduce the amount
6132 of flapping under certain conditions.
6134 8. log_incoming_interface used to affect only the <= reception log lines. Now
6135 it causes the local interface and port to be added to several more SMTP log
6136 lines, for example "SMTP connection from", and rejection lines.
6138 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6140 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6142 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6143 could be overwritten at the end of the current message (or the start of a
6144 new message if it was set in a HELO ACL). The value is now preserved for
6145 the duration of the SMTP connection.
6147 12. If a transport had a headers_rewrite setting, and a matching header line
6148 contained an unqualified address, that address was qualified, even if it
6149 did not match any rewriting rules. The underlying bug was that the values
6150 of the flags that permit the existence of unqualified sender and recipient
6151 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6152 non-local messages, and by -bnq for local messages) were not being
6153 preserved with the message after it was received.
6155 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6156 "next input=" as part of the text comprising the host identity instead of
6157 the correct text. The code was using the same buffer for two different
6158 strings. However, depending on which order the printing function evaluated
6159 its arguments, the bug did not always show up. Under Linux, for example, my
6160 test suite worked just fine.
6162 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6163 doesn't work with some older versions of Perl. It has been changed to "my",
6164 which in any case is probably the better facility to use.
6166 15. A really picky compiler found some instances of statements for creating
6167 error messages that either had too many or two few arguments for the format
6170 16. The size of the buffer for calls to the DNS resolver has been increased
6171 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6172 for addresses that have a lot of PTR records. This alleviates a problem; it
6173 does not fully solve it.
6175 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6176 buffer now truncates the list and logs the incident, which is the same
6177 action as happens when Exim is looking up a host name and its aliases.
6178 Previously in this situation something unpredictable would happen;
6179 sometimes it was "internal error: store_reset failed".
6181 18. If a server dropped the connection unexpectedly when an Exim client was
6182 using GnuTLS and trying to read a response, the client delivery process
6183 crashed while trying to generate an error log message.
6185 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6186 string, for example:
6188 warn message = H1: something\nH2: something
6190 the text was added as a single header line from Exim's point of view
6191 though it ended up OK in the delivered message. However, searching for the
6192 second and subsequent header lines using $h_h2: did not work. This has been
6193 fixed. Similarly, if a system filter added multiple headers in this way,
6194 the routers could not see them.
6196 20. Expanded the error message when iplsearch is called with an invalid key to
6197 suggest using net-iplsearch in a host list.
6199 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6200 ACLs are no longer actually imposed (and a message to that effect is
6203 22. If a "gecos" field in a passwd entry contained escaped characters, in
6204 particular, if it contained a \" sequence, Exim got it wrong when building
6205 a From: or a Sender: header from that name. A second bug also caused
6206 incorrect handling when an unquoted " was present following a character
6207 that needed quoting.
6209 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6210 was not being matched caselessly.
6212 24. Arranged for all hyphens in the exim.8 source to be escaped with
6215 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6216 properly. Recipient callout cache records were still being keyed to include
6217 the sender, even when use_sender was set false. This led to far more
6218 callouts that were necessary. The sender is no longer included in the key
6219 when use_sender is false.
6221 26. Added "control = submission" modifier to ACLs.
6223 27. Added the ${base62d: operator to decode base 62 numbers.
6225 28. dnsdb lookups can now access SRV records.
6227 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6228 the configuration file.
6230 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6231 (-v) mode. This makes the output for a verbose queue run more intelligible.
6233 31. Added a use_postmaster feature to recipient callouts.
6235 32. Added the $body_zerocount variable, containing the number of binary zero
6236 bytes in the message body.
6238 33. The time of last modification of the "new" subdirectory is now used as the
6239 "mailbox time last read" when there is a quota error for a maildir
6242 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6244 35. Added +ignore_unknown as a special item in host lists.
6246 36. Code for decoding IPv6 addresses in host lists is now included, even if
6247 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6248 address was recognized as an IP address, but was then not correctly decoded
6249 into binary, causing unexpected and incorrect effects when compared with
6256 1. Very minor rewording of debugging text in manualroute to say "list of
6257 hosts" instead of "hostlist".
6259 2. If verify=header_syntax was set, and a header line with an unqualified
6260 address (no domain) and a large number of spaces between the end of the
6261 name and the colon was received, the reception process suffered a buffer
6262 overflow, and (when I tested it) crashed. This was caused by some obsolete
6263 code that should have been removed. The fix is to remove it!
6265 3. When running in the test harness, delay a bit after writing a bounce
6266 message to get a bit more predictability in the log output.
6268 4. Added a call to search_tidyup() just before forking a reception process. In
6269 theory, someone could use a lookup in the expansion of smtp_accept_max_
6270 per_host which, without the tidyup, could leave open a database connection.
6272 5. Added the variables $recipient_data and $sender_data which get set from a
6273 lookup success in an ACL "recipients" or "senders" condition, or a router
6274 "senders" option, similar to $domain_data and $local_part_data.
6276 6. Moved the writing of debug_print from before to after the "senders" test
6279 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6280 problems for message scanning, either using a data ACL, or using
6281 local_scan() because the Received: header was not generated till after they
6282 were called (in order to set the time as the time of reception completion).
6283 I have revised the way this works. The header is now generated after the
6284 body is received, but before the ACL or local_scan() are called. After they
6285 are run, the timestamp in the header is updated.
6291 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6292 before starting a queue runner without re-exec. This happened only when
6293 deliver_drop_privilege was set or when the Exim user was set to root. The
6294 effect of the bug was that timeouts during subsequent deliveries caused
6295 crashes instead of being properly handled. The handler is now left at its
6296 default (and expected) setting.
6298 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6299 message, again when deliver_drop_privilege is set or Exim is run as root.
6300 The bug described in (1) was not present in this case, but the tidying up
6301 of the other signals was missing. I have made the two cases consistent.
6303 3. The ignore_target_hosts setting on a manualroute router was being ignored
6304 for hosts that were looked up using the /MX notation.
6306 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6309 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6310 operated on the sender address. After changing the $sender_address to <>
6311 for the sender address verify, Exim was re-instated it as the original
6312 (before rewriting) address, but remembering that it had rewritten it, so it
6313 wasn't rewriting it again. This bug also had the effect of breaking the
6314 sender address verification caching when the sender address was rewritten.
6316 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6317 This has been changed so that if the ip literal address matches
6318 ignore_target_hosts, the router declines.
6320 7. Added expansion conditions match_domain, match_address, and match_local_
6321 part (NOT match_host).
6323 8. The placeholder for the Received: header didn't have a length field set.
6325 9. Added code to Exim itself and to exim_lock to test for a specific race
6326 condition that could lead to file corruption when using MBX delivery. The
6327 issue is with the lockfile that is created in /tmp. If this file is removed
6328 after a process has opened it but before that process has acquired a lock,
6329 there is the potential for a second process to recreate the file and also
6330 acquire a lock. This could lead to two Exim processes writing to the file
6331 at the same time. The added code performs the same test as UW imapd; it
6332 checks after acquiring the lock that its file descriptor still refers to
6333 the same named file.
6335 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6336 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6339 11. Added the smtp_active_hostname option. If used, this will typically be made
6340 to depend on the incoming interface address. Because $interface_address is
6341 not set up until the daemon has forked a reception process, error responses
6342 that can happen earlier (such as "too many connections") no longer contain
6345 12. If an expansion in a condition on a "warn" statement fails because a lookup
6346 defers, the "warn" statement is abandoned, and the next ACL statement is
6347 processed. Previously this caused the whole ACL to be aborted.
6349 13. Added the iplsearch lookup type.
6351 14. Added ident_timeout as a log selector.
6353 15. Added tls_certificate_verified as a log selector.
6355 16. Added a global option tls_require_ciphers (compare the smtp transport
6356 option of the same name). This controls incoming TLS connections.
6358 17. I finally figured out how to make tls_require_ciphers do a similar thing
6359 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6360 before starting the TLS session.
6362 18. Tabs are now shown as \t in -bP output.
6364 19. If the log selector return_path_on_delivery was set, Exim crashed when
6365 bouncing a message because it had too many Received: header lines.
6367 20. If two routers both had headers_remove settings, and the first one included
6368 a superfluous trailing colon, the final name in the first list and the
6369 first name in the second list were incorrectly joined into one item (with a
6370 colon in the middle).
6376 1. Added -C and -D options to the exinext utility, mainly to make it easier
6377 to include in the automated testing, but these could be helpful when
6378 multiple configurations are in use.
6380 2. The exinext utility was not formatting the output nicely when there was
6381 an alternate port involved in the retry record key, nor when there was a
6382 message id as well (for retries that were specific to a specific message
6383 and a specific host). It was also confused by IPv6 addresses, because of
6384 the additional colons they contain. I have fixed the IPv4 problem, and
6385 patched it up to do a reasonable job for IPv6.
6387 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6388 delivery, the log line now contains "pipelined" if PIPELINING was used.
6390 4. An SMTP transport process used to panic and die if the bind() call to set
6391 an explicit outgoing interface failed. This has been changed; it is now
6392 treated in the same way as a connect() failure.
6394 5. A reference to $sender_host_name in the part of a conditional expansion
6395 that was being skipped was still causing a DNS lookup. This no longer
6398 6. The def: expansion condition was not recognizing references to header lines
6399 that used bh_ and bheader_.
6401 7. Added the _cache feature to named lists.
6403 8. The code for checking quota_filecount in the appendfile transport was
6404 allowing one more file than it should have been.
6406 9. For compatibility with Sendmail, the command line option
6414 and sets the incoming protocol and host name (for trusted callers). The
6415 host name and its colon can be omitted when only the protocol is to be set.
6416 Note the Exim already has two private options, -pd and -ps, that refer to
6417 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6418 "s", but I don't think that's a major issue.
6420 10. A number of refactoring changes to the code, none of which should affect
6423 (a) The number of logging options was getting close to filling up the
6424 32-bit word that was used as a bit map. I have split them into two classes:
6425 those that are passed in the argument to log_write(), and those that are
6426 only ever tested independently outside of that function. These are now in
6427 separate 32-bit words, so there is plenty of room for expansion again.
6428 There is no change in the user interface or the logging behaviour.
6430 (b) When building, for example, log lines, the code previously used a
6431 macro that called string_cat() twice, in order to add two strings. This is
6432 not really sufficiently general. Furthermore, there was one instance where
6433 it was actually wrong because one of the argument was used twice, and in
6434 one call a function was used. (As it happened, calling the function twice
6435 did not affect the overall behaviour.) The macro has been replaced by a
6436 function that can join an arbitrary number of extra strings onto a growing
6439 (c) The code for expansion conditions now uses a table and a binary chop
6440 instead of a serial search (which was left over from when there were very
6441 few conditions). Also, it now recognizes conditions like "pam" even when
6442 the relevant support is not compiled in: a suitably worded error message is
6443 given if an attempt is made to use such a condition.
6445 11. Added ${time_interval:xxxxx}.
6447 12. A bug was causing one of the ddress fields not to be passed back correctly
6448 from remote delivery subprocesses. The field in question was not being
6449 subsequently used, so this caused to problems in practice.
6451 13. Added new log selectors queue_time and deliver_time.
6453 14. Might have fixed a bug in maildirsizefile handling that threw up
6454 "unexpected character" debug warnings, and recalculated the data
6455 unnecessarily. In any case, I expanded the warning message to give more
6458 15. Added the message "Restricted characters in address" to the statements in
6459 the default ACL that block characters like @ and % in local parts.
6461 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6462 Three changes have been made:
6464 (a) There was a serious bug; a negative response to MAIL caused the whole
6465 recipient domain to be cached as invalid, thereby blocking all messages
6466 to all local parts at the same domain, from all senders. This bug has
6467 been fixed. The domain is no longer cached after a negative response to
6468 MAIL if the sender used is not empty.
6470 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6473 (c) A new callout option, "use_sender" has been added for people who want
6474 the modified behaviour.
6480 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6483 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6484 indeed breaks things for older releases.
6486 3. Added additional logging to the case where there is a problem reading data
6487 from a filter that is running in a subprocess using a pipe, in order to
6488 try to track down a specific problem.
6490 4. Testing facility fudge: when running in the test harness and attempting
6491 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6492 getting "No route to host". Convert this to a timeout.
6494 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6497 6. Some OS don't have socklen_t but use size_t instead. This affects the
6498 fifth argument of getsockopt() amongst other things. This is now
6499 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6500 can be set for individual OS. I have set it for SunOS5, OSF1, and
6501 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6502 some earlier ones do not.
6504 7. Change 4.30/15 was not doing the test caselessly.
6506 8. The standard form for an IPv6 address literal was being rejected by address
6507 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6508 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6509 this, as well as the form without the "IPv6" on the front (but only when
6510 address literals are enabled, of course).
6512 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6514 10. Exim crashed if a message with an empty sender address specified by -f
6515 encountered a router with an errors_to setting. This could be provoked only
6516 by a command such as
6520 where an empty string was supplied; "<>" did not hit this bug.
6522 11. Installed PCRE release 4.5.
6524 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6525 remained set. It is now erased.
6527 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6528 times from message ids (which are base 36 rather than the normal 62).
6530 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6531 were being counted as actual protocol errors, and logged if the log
6532 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6533 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6534 and DATA following a set of rejected RCPTs do not count as protocol errors.
6535 In other words, Exim assumes they were pipelined, though this may not
6536 actually be the case. Of course, in all cases the client gets an
6537 appropriate error code.
6539 15. If a lookup fails in an ACL condition, a message about the failure may
6540 be available; it is used if testing the ACL cannot continue, because most
6541 such messages specify what the cause of the deferral is. However, some
6542 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6543 that caused an old message to be retained and used if a later statement
6544 caused a defer, replacing the real cause of the deferral.
6546 16. If an IP address had so many PTR records that the DNS lookup buffer
6547 was not large enough to hold them, Exim could crash while trying to process
6548 the truncated data. It now detects and logs this case.
6550 17. Further to 4.21/58, another change has been made: if (and only if) the
6551 first line of a message (the first header line) ends with CRLF, a bare LF
6552 in a subsequent header line has a space inserted after it, so as not to
6553 terminate the header.
6555 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6556 unnecessarily, used atoi() instead of strtol(), and didn't check the
6557 termination when getting file sizes from file names by regex.
6559 19. Completely re-implemented the support for maildirsize files, in the light
6560 of a number of problems with the previous contributed implementation
6561 (4.30/29). In particular:
6563 . If the quota is zero, the maildirsize file is maintained, but no quota is
6566 . If the maildir directory does not exist, it is created before any attempt
6567 to write a maildirsize file.
6569 . The quota value in the file is just a cache; if the quota is changed in
6570 the transport, the new value overrides.
6572 . A regular expression is available for excluding directories from the
6575 20. The autoreply transport checks the characters in options that define the
6576 message's headers; it allows continued headers, but it was checking with
6577 isspace() after an embedded newline instead of explicitly looking for a
6580 21. If all the "regular" hosts to which an address was routed had passed their
6581 expiry times, and had not reached their retry times, the address was
6582 bounced, even if fallback hosts were defined. Now Exim should go on to try
6585 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6586 equivalent code in the SMTP transport. Some hosts send humungous responses
6587 to HELO/EHLO, more than 1024 it seems.
6589 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6590 gives compiler warnings in some environments. I've now done it "properly",
6593 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6594 (because of problems with the built-in one) was declared to return uschar *
6595 instead of char *, causing compiler failure.
6597 25. Fixed a file descriptor leak when processing alias/forward files.
6599 26. Fixed a minor format string issue in dbfn.c.
6601 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6603 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6604 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6607 29. When a sender address is verified, it is cached, to save repeating the test
6608 when there is more than one recipient in a message. However, when the
6609 verification involves a callout, it is possible for different callout
6610 options to be set for different recipients. It is too complicated to keep
6611 track of this in the cache, so now Exim always runs a verification when a
6612 callout is required, relying on the callout cache for the optimization.
6613 The overhead is duplication of the address routing, but this should not be
6616 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6617 to be verified with callout=postmaster, and the main callout worked but the
6618 postmaster check failed, the verification correctly failed. However, if a
6619 subsequent RCPT command asked for sender verification *without* the
6620 postmaster check, incorrect caching caused this verification also to fail,
6623 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6624 it was not caching the DNS options (qualify_single, search_parents) that
6625 were used when the lookup failed. A subsequent lookup with different
6626 options therefore always gave the same answer, though there were cases
6627 where it should not have. (Example: a "domains = !$mx_any" option on a
6628 dnslookup router: the "domains" option is always processed without any
6629 widening, but the router might have qualify_single set.) Now Exim uses the
6630 cached value only when the same options are set.
6632 32. Added John Jetmore's "exipick" utility to the distribution.
6634 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6635 than a timeout (e.g. a certificate is required, and is not provided), an
6636 Exim server now closes the connection immediately. Previously it waited for
6637 the client to close - but if the client is SSL, it seems that they each
6638 wait for each other, leading to a delay before one of them times out.
6640 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6641 maintained 0.8.x compatibility because I don't think many are using it, and
6642 it is clearly obsolete.
6644 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6647 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6648 client certificate was expired. A simple patch fixes this, though I don't
6649 understand the full logic of why the verify callback is called multiple
6652 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6653 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6654 which causes problems with some clients (such as the Certicom SSL Plus
6655 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6656 disables the coutermeasure allowing Eudora to connect."
6658 38. Exim was not checking that a write() to a log file succeeded. This could
6659 lead to Bad Things if a log got too big, in particular if it hit a file
6660 size limit. Exim now panics and dies if it cannot write to a log file, just
6661 as it does if it cannot open a log file.
6663 39. Modified OS/Makefile-Linux so that it now contains
6665 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6667 The two -D definitions ensure that Exim is compiled with large file
6668 support, which makes it possible to handle log files that are bigger than
6671 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6672 instance) a domain was checked against a named list that involved a lookup,
6673 causing $domain_data to be set, then another domain was checked against the
6674 same list, then the first domain was re-checked, the value of $domain_data
6675 after the final check could be wrong. In particular, if the second check
6676 failed, it could be set empty. This bug probably also applied to
6679 41. The strip_trailing_dot option was not being applied to the address given
6680 with the -f command-line option.
6682 42. The code for reading a message's header from the spool was incrementing
6683 $received_count, but never initializing it. This meant that the value was
6684 incorrect (doubled) while delivering a message in the same process in which
6685 it was received. In the most common configuration of Exim, this never
6686 happens - a fresh exec is done - but it can happen when
6687 deliver_drop_privilege is set.
6689 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6690 it now includes up to 150 characters of the unexpected data in the log
6693 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6694 and building data strings. The size of both of these buffers was 10 000
6695 bytes - far larger than anybody would *ever* want, thought I. Needless to
6696 say, somebody hit the limit. I have increased the maximum line length to
6697 20 000 and the maximum data length of concatenated lines to 100 000. I have
6698 also fixed two bugs, because there was no checking on these buffers. Tsk,
6699 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6700 buffer is too small.
6702 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6703 lsearch lookups. Now it does.
6705 46. When parsing a route_list item in a manualroute router, a fixed-length
6706 buffer was used for the list of hosts. I made this 1024 bytes long,
6707 thinking that nobody would ever have a list of hosts that long. Wrong.
6708 Somebody had a whole pile of complicated expansion conditions, and the
6709 string was silently truncated, leading to an expansion error. It turns out
6710 that it is easier to change to an unlimited length (owing to other changes
6711 that have happened since this code was originally written) than to build
6712 structure for giving a limitation error. The length of the item that
6713 expands into the list of hosts is now unlimited.
6715 47. The lsearch lookup could not handle data where the length of text line was
6716 more than 4095 characters. Such lines were truncated, leading to shortened
6717 data being returned. It should now handle lines of any length.
6719 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6720 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6723 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6724 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6725 no longer happens in comment lines. A list of replacements is now placed
6726 at the head of all of the source files, except those whose only change is
6727 to replace PERL_COMMAND in the very first #! line.
6729 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6730 messages on the queue, with a bottom-up merge sort, using code contributed
6731 by Michael Haardt. This should make operations like -bp somewhat faster on
6732 large queues. It won't affect queue runners, except when queue_run_in_order
6735 51. Installed eximstats 1.31 in the distribution.
6737 52. Added support for SRV lookups to the dnslookup router.
6739 53. If an ACL referred to $message_body or $message_body_end, the value was not
6740 reset for any messages that followed in the same SMTP session.
6742 54. The store-handling optimization for building very long strings was not
6743 differentiating between the different store pools. I don't think this
6744 actually made any difference in practice, but I've tidied it.
6746 55. While running the routers to verify a sender address, $sender_address
6747 was still set to the sender address. This is wrong, because when routing to
6748 send a bounce to the sender, it would be empty. Therefore, I have changed
6749 it so that, while verifying a sender address, $sender_address is set to <>.
6750 (There is no change to what happens when verifying a recipient address.)
6752 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6753 target A or AAAA records (if not already returned) without resetting the
6754 qualify_single or search_parents options of the DNS resolver. These are
6755 inappropriate in this case because the targets of MX and SRV records must
6756 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6757 target that, when qualified, matched something in the local domain. These
6758 two options are now turned off when doing these lookups.
6760 57. It seems that at least some releases of Reiserfs (which does not have the
6761 concept of a fixed number of inodes) returns zero and not -1 for the
6762 number of available inodes. This interacted badly with check_spool_inodes,
6763 which assumed that -1 was the "no such thing" setting. What I have done is
6764 to check that the total number of inodes is greater than zero before doing
6765 the test of how many are available.
6767 58. When a "warn" ACL statement has a log_message modifier, the message is
6768 remembered, and not repeated. This is to avoid a lot of repetition when a
6769 message has many recipients that cause the same warning to be written.
6770 However, Exim was preserving the list of already written lines for an
6771 entire SMTP session, which doesn't seem right. The memory is now reset if a
6772 new message is started.
6774 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6775 debugging output unless log_rewrite was also set.
6777 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6778 of (int)(handle) when we know that handle contains (void *)(-1).
6780 61. The Exim daemon panic-logs an error return when it closes the incoming
6781 connection. However "connection reset by peer" seems to be common, and
6782 isn't really an error worthy of noting specially, so that particular error
6785 62. When Exim is trying to find all the local interfaces, it used to panic and
6786 die if the ioctl to get the interface flags failed. However, it seems that
6787 on at least one OS (Solaris 9) it is possible to have an interface that is
6788 included in the list of interfaces, but for which you get a failure error
6789 for this call. This happens when the interface is not "plumbed" into a
6790 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6791 failure of the "get flags" call assumes that the interface is down.
6793 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6794 makes life easier for people who are doing arithmetic on fields extracted
6795 from dates, where you often get leading zeros that should not be
6796 interpreted as octal.
6798 64. Added qualify_domain to the redirect router, to override the global
6801 65. If a pathologically long header line contained very many addresses (the
6802 report of this problem mentioned 10 000) and each of them was rewritten,
6803 Exim could use up a very large amount of memory. (It kept on making new
6804 copies of the header line as it rewrote, and never released the old ones.)
6805 At the expense of a bit more processing, the header rewriting function has
6806 been changed so that it no longer eats memory in this way.
6808 66. The generation of the Received: header has been moved from the time that a
6809 message starts to be received, to the time that it finishes. The timestamp
6810 in the Received: header should now be very close to that of the <= log
6811 line. There are two side-effects of this change:
6813 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6814 logged header lines no longer include the local Received: line, because
6815 it has not yet been created. The same applies to a copy of the message
6816 that is returned to a non-SMTP sender when a message is rejected.
6818 (b) When a filter file is tested using -bf, no additional Received: header
6819 is added to the test message. After some thought, I decided that this
6822 This change does not affect the value of $received_for. It is still set
6823 after address rewriting, but before local_scan() is called.
6825 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6827 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6828 gave an unhelpful panic error message, and a defer error. I have managed to
6829 change this behaviour so that it now rejects any supplied certificate,
6830 which seems right, as the list of acceptable certificates is empty.
6832 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6833 gave an unhelpful defer error. I have not managed to make this reject any
6834 supplied certificates, but the error message it gives is "no certificate
6835 supplied", which is not helpful.
6837 70. exigrep's output now also includes lines that are not associated with any
6838 message, but which match the given pattern. Implemented by a patch from
6839 Martin Sluka, which also tidied up the Perl a bit.
6841 71. Recipient callout verification, like sender verification, was using <> in
6842 the MAIL FROM command. This isn't really the right thing, since the actual
6843 sender may affect whether the remote host accepts the recipient or not. I
6844 have changed it to use the actual sender in the callout; this means that
6845 the cache record is now keyed on a recipient/sender pair, not just the
6846 recipient address. There doesn't seem to be a real danger of callout loops,
6847 since a callout by the remote host to check the sender would use <>.
6848 [SEE ABOVE: changed after hitting problems.]
6850 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6851 temporary errors. However, in the case of such a code being given after
6852 the end of a data transmission (i.e. after ".") Exim was failing to write
6853 a retry record for the message. (Yes, there was some broken host that was
6854 actually sending 8xx at this point.)
6856 73. An unknown lookup type in a host list could cause Exim to panic-die when
6857 the list was checked. (An example that provoked this was putting <; in the
6858 middle of a list instead of at the start.) If this happened during a DATA
6859 ACL check, a -D file could be left lying around. This kind of configuration
6860 error no longer causes Exim to die; instead it causes a defer error. The
6861 incident is still logged to the main and panic logs.
6863 74. Buglet left over from Exim 3 conversion. The message "too many messages
6864 in one connection" was written to the rejectlog but not the mainlog, except
6865 when address rewriting (yes!) was being logged.
6867 75. Added write_rejectlog option.
6869 76. When a system filter was run not as root (that is, when system_filter_user
6870 was set), the values of the $n variables were not being returned to the
6871 main process; thus, they were not subsequently available in the $sn
6874 77. Added +return_path_on_delivery log selector.
6876 78. A connection timeout was being treated differently from recipients deferred
6877 when testing hosts_max_try with a message that was older than the host's
6878 retry timeout. (The host should not be counted, thus allowing all hosts to
6879 be tried at least once before bouncing.) This may have been the cause of an
6880 occasionally reported bug whereby a message would remain on the queue
6881 longer than the retry timeout, but would be bounced if a delivery was
6882 forced. I say "may" because I never totally pinned down the problem;
6883 setting up timeout/retry tests is difficult. See also the next item.
6885 79. The ultimate address timeout was not being applied to errors that involved
6886 a combination of host plus message (for example, a timeout on a MAIL
6887 command). When an address resolved to a number of possible hosts, and they
6888 were not all tried for each delivery (e.g. because of hosts_max_try), a
6889 message could remain on the queue longer than the retry timeout.
6891 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6894 81. Fixed an obscure SMTP outgoing bug which required at least the following
6895 conditions: (a) there was another message waiting for the same server;
6896 (b) the server returned 5xx to all RCPT commands in the first message so
6897 that the message was not completed; (c) the server dropped the connection
6898 or gave a negative response to the RSET that Exim sends to abort the
6899 transaction. The observed case was a dropped connection after DATA that had
6900 been sent in pipelining mode. That is, the server had advertised PIPELINING
6901 but was not implementing it correctly. The effect of the bug was incorrect
6902 behaviour, such as trying another host, and this could lead to a crash.
6908 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6909 and daemon.c were passed as pointers to ints; they should have been
6910 pointers to socklen_t variables (which are typically unsigned ints).
6912 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6915 3. Fixed a really odd bug that affected only the testing scheme; patching a
6916 certain fixed string in the binary changed the value of another string that
6917 happened to be identical to the end of the original first string.
6919 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6920 name", it returns that address as the IP address. On some operating
6921 systems (e.g. Solaris), it also passes back the IP address string as the
6922 "host name". However, on others (e.g. Linux), it passes back an empty
6923 string. Exim wasn't checking for this, and was changing the host name to an
6924 empty string, assuming it had been canonicalized.
6926 5. Although rare, it is permitted to have more than one PTR record for a given
6927 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6928 all the names associated with an address, because they do in Solaris.
6929 However, it seems that they do not in Linux for data that comes from the
6930 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6931 I found this out when I moved to a new Linux workstation and tried to run
6932 the Exim test suite.
6934 To get round this problem I have changed the code so that it now does its
6935 own call to the DNS to look up PTR records when searching for a host name.
6936 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6937 addresses that are only in /etc/hosts are still found.
6939 This behaviour is, however, controlled by an option called host_lookup_
6940 order, which defaults to "bydns:byaddr". If people want to use the other
6941 order, or indeed, just use one or the other means of lookup, they can
6942 specify it in this variable.
6944 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6945 some operating systems, this comes back from gethostbyaddr() as an empty
6946 string, and this is what Exim used to test for. However, it seems that in
6947 other systems, "." is yielded. Exim now tests for this case too.
6949 7. The values of check_spool_space and check_log_space are now held internally
6950 as a number of kilobytes instead of an absolute number of bytes. If a
6951 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6952 kilobyte. This means that much larger values can be stored.
6954 8. Exim monitor: an attempt to get the action menu when not actually pointing
6955 at a message produces an empty menu entitled "No message selected". This
6956 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6957 no entries in it ("Shell widget menu has zero width and/or height"). So I
6958 have added a single, blank menu entry in this case.
6960 9. Added ${quote_local_part.
6962 10. MIME decoding is now applied to the contents of Subject: header lines when
6965 11. Now that a reference to $sender_host_address automatically causes a reverse
6966 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6967 host lookup before query-style lookups in lists that might use this
6968 variable. This has therefore been abolished, and the "net-" prefix is no
6969 longer necessary for query-style lookups.
6971 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6972 have been a typo for LFLAGS, so it has been changed.
6974 13. The install script calls Exim with "-C /dev/null" in order to find the
6975 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6976 to be output. However, since Exim outputs its version number before the
6977 error, it didn't break the script. It just looked ugly. I fixed this by
6978 always allowing "-C /dev/null" if the caller is root.
6980 14. Ignore overlarge ACL variable number when reading spool file - insurance
6981 against a later release with more variables having written the file.
6983 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6984 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6985 this, as well as the form without the "IPv6" on the front.
6987 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6988 OS/Makefile-Darwin file.
6990 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6991 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6993 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6994 variable. However, some OS use ints for inodes. I've added cast to long int
6995 to get rid of the compiler warning.
6997 19. I had forgotten to lock out "/../" in configuration file names when
6998 ALT_CONFIG_PREFIX was set.
7000 20. Routers used for verification do not need to specify transports. However,
7001 if such a router generated a host list, and callout was configured, Exim
7002 crashed, because it could not find a port number from the (non-existent)
7003 transport. It now assumes port 25 in this circumstance.
7005 21. Added the -t option to exigrep.
7007 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7008 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7009 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7010 them is set, LOOKUP_LSEARCH is forced.
7012 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7014 24. Added sender and host information to the "rejected by local_scan()" log
7015 line; previously there was no indication of these.
7017 25. Added .include_if_exists.
7019 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7020 Linux. It turns out that not all file systems support this. Apparently some
7021 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7022 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7023 not supported on the file descriptor, is now ignored when Exim is trying to
7024 sync a directory. This applies only to Linux.
7026 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7028 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7029 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7032 29. Added support for maildirsize files from supplied patch (modified a bit).
7034 30. The use of :fail: followed by an empty string could lead Exim to respond to
7035 sender verification failures with (e.g.):
7037 550 Verification failed for <xxx>
7038 550 Sender verify failed
7040 where the first response line was missing the '-' that indicates it is not
7041 the final line of the response.
7043 31. The loop for finding the name of the user that called Exim had a hardwired
7044 limit of 10; it now uses the value of finduser_retries, which is used for
7045 all other user lookups.
7047 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7050 33. Exim was neglecting to zero errno before one call of strtol() when
7051 expanding a string and expecting an integer value. On some systems this
7052 resulted in spurious "integer overflow" errors. Also, it was casting the
7053 result into an int without checking.
7055 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7056 did not work. The code looks as if it has *never* worked, though it appears
7057 to have been documented since at least release 1.62. I have made it work.
7059 35. The "timeout_DNS" error in retry rules, also documented since at least
7060 1.62, also never worked. As it isn't clear exactly what this means, and
7061 clearly it isn't a major issue, I have abolished the feature by treating it
7062 as "timeout", and writing a warning to the main and panic logs.
7064 36. The display of retry rules for -brt wasn't always showing the error code
7067 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7068 timeout_connect_A, timeout_connect_MX.
7070 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7071 to the empty sender.
7073 39. The daemon was not analysing the content of -oX till after it had closed
7074 stderr and disconnected from the controlling terminal. This meant that any
7075 syntax errors were only noted on the panic log, and the return code from
7076 the command was 0. By re-arranging the code a little, I've made the
7077 decoding happen first, so such errors now appear on stderr, and the return
7078 code is 1. However, the actual setting up of the sockets still happens in
7079 the disconnected process, so errors there are still only recorded on the
7082 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7083 connections (as happens on some IP stacks) was logged at start up time as
7084 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7085 it from "IPv6 and IPv4", which means that two separate sockets are being
7088 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7089 says whether AF_INET or AF_INET6 was passed as an argument.
7091 42. Exiwhat output was messed up when time zones were included in log
7094 43. Exiwhat now gives more information about the daemon's listening ports,
7095 and whether -tls-on-connect was used.
7097 44. The "port" option of the smtp transport is now expanded.
7099 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7100 silently ignored. Now an error message is written to the main and panic
7103 46. There's a new ACL modifier called "logwrite" which writes to a log file
7104 as soon as it is encountered.
7106 47. Added $local_user_uid and $local_user_gid at routing time.
7108 48. Exim crashed when trying to verify a sender address that was being
7111 49. Exim was recognizing only a space character after ".include". It now also
7112 recognizes a tab character.
7114 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7115 extracting the relevant information from the specification. The man page no
7116 longer contains scrambled data for the -d option, and I've added a section
7117 at the front about calling Exim under different names.
7119 51. Added "extra_headers" argument to the "mail" command in filter files.
7121 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7124 53. Installed eximstats 1.29.
7126 54. Added transport_filter_timeout as a generic transport option.
7128 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7129 Cc: header lines. This was required by RFC 822, but it not required by RFC
7132 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7133 incoming messages that did not have them. Now it does so only if the
7134 message originates locally, that is, if there is no associated remote host
7135 address. When Resent- header lines are present, this applies to the Resent-
7136 lines rather than the non-Resent- lines.
7138 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7139 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7141 58. Messages for configuration errors now include the name of the main
7142 configuration file - useful now that there may be more than one file in a
7143 list (.included file names were always shown).
7145 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7146 for those rare installations that do not start the daemon as root or run it
7147 setuid root. I've cut out the call to initgroups() if the daemon is not
7150 60. The Exim user and group can now be bound into the binary as text strings
7151 that are looked up at the start of Exim's processing.
7153 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7155 62. Added $mailstore_basename variable.
7157 63. Installed patch to sieve.c from Michael Haardt.
7159 64. When Exim failed to open the panic log after failing to open the main log,
7160 the original message it was trying to log was written to stderr and debug
7161 output, but if they were not available (the usual case in production), it
7162 was lost. Now it is written to syslog before the two lines that record the
7163 failures to open the logs.
7165 65. Users' Exim filters run in subprocesses under the user's uid. It is
7166 possible for a "deliver" command or an alias in a "personal" command to
7167 provoke an address rewrite. If logging of address rewriting is configured,
7168 this fails because the process is not running as root or exim. There may be
7169 a better way of dealing with this, but for the moment (because 4.30 needs
7170 to be released), I have disabled address rewrite logging when running a
7171 filter in a non-root, non-exim process.
7177 1. The buildconfig auxiliary program wasn't quoting the value set for
7178 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7179 not defined. This bug was masked in 4.22 by the effect that was fixed in
7182 2. Some messages that were rejected after a message id was allocated were
7183 shown as "incomplete" by exigrep. It no longer does this for messages that
7184 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7186 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7187 have allow_domain_literals set, the ID did not get logged in the <= line.
7188 Domain literals are now always recognized in Message-ID: header lines.
7190 4. The first argument for a ${extract expansion item is the key name or field
7191 number. Leading and trailing spaces in this item were not being ignored,
7192 causing some misleading effects.
7194 5. When deliver_drop_privilege was set, single queue runner processes started
7195 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7196 same command in the process it spins off) were not dropping privilege.
7198 6. When the daemon running as "exim" started a queue runner, it always
7199 re-executed Exim in the spun-off process. This is a waste of effort when
7200 deliver_drop_privilege is set. The new process now just calls the
7201 queue-runner function directly.
7207 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7210 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7211 ignored. Though the use of -oP was forcing the writing of a pid file, it
7212 was always written to the default place.
7214 3. If the message "no IP address found for host xxxx" is generated during
7215 incoming verification, it is now followed by identification of the incoming
7216 connection (so you can more easily find what provoked it).
7218 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7220 5. Added some features to "harden" Exim a bit more against certain attacks:
7222 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7223 be put in Local/Makefile. This is like the never_users runtime option,
7224 but it cannot be overridden. The default setting is "root".
7226 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7227 prefix string with which any file named in a -C command line option
7230 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7231 is retained for -C and -D only if the caller of Exim is root. Without
7232 it, the exim user may also use -C and -D and retain privilege.
7234 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7235 command line option is disabled.
7237 6. Macro names set by the -D option must start with an upper case letter, just
7238 like macro names defined in the configuration file.
7240 7. Added "dereference=" facility to LDAP.
7242 8. Two instances of the typo "uknown" in the source files are fixed.
7244 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7245 the Configure-Makefile script screwed up while processing it.
7247 10. Incorporated PCRE 4.4.
7249 11. The SMTP synchronization check was not operating right at the start of an
7250 SMTP session. For example, it could not catch a HELO sent before the client
7251 waited for the greeting. There is now a check for outstanding input at the
7252 point when the greeting is written. Because of the duplex, asynchronous
7253 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7254 way, but not yet received, when the check is performed.
7256 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7257 on TCP/IP sockets, because this apparently causes some broken clients to
7260 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7261 unchanged) from the Cygwin maintainer.
7263 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7264 was supported instead of testing for mbx. Effectively a typo.
7266 15. The spa authenticator server code was not checking that the input it
7267 received was valid base64.
7269 16. The debug output line for the "set" modifier in ACLs was not showing the
7270 name of the variable that was being set.
7272 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7274 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7275 lookup on its own. Something else was needed to trigger the lookup. For
7276 example, a match in host_lookup or the need for a host name in a host list.
7277 Now, if $sender_host_name is referenced and the host name has not yet been
7278 looked up, a lookup is performed. If the lookup fails, the variable remains
7279 empty, and $host_lookup_failed is set to "1".
7281 19. Added "eqi" as a case-independent comparison operator.
7283 20. The saslauthd authentication condition could segfault if neither service
7284 nor realm was specified.
7286 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7287 error message that was logged was misleading, and incoming SMTP
7288 connections were dropped. The message is now more accurate, and temporary
7289 errors are given to SMTP connections.
7291 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7292 (except RSET) with a 421 error, until QUIT is received. However, it was
7293 failing to send a response to QUIT.
7295 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7296 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7297 ACL is now run after the helo_try_verify_hosts code.
7299 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7300 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7301 case-independent, so other case variants are also recognized). Apparently
7302 some systems use these upper case variants.
7304 25. If more than two messages were waiting for the same host, and a transport
7305 filter was specified for the transport, Exim sent two messages over the
7306 same TCP/IP connection, and then failed with "socket operation on non-
7307 socket" when it tried to send the third.
7309 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7311 27. The extern definition of crypt16() in expand.c was not being excluded when
7312 the OS had its own crypt16() function.
7314 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7315 as a preferred synonym for return_size_limit, both as an option and as an
7318 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7320 30. Changed the default configuration ACL to relax the local part checking rule
7321 for addresses that are not in any local domains. For these addresses,
7322 slashes and pipe symbols are allowed within local parts, but the sequence
7323 /../ is explicitly forbidden.
7325 31. SPA server authentication was not clearing the challenge buffer before
7328 32. log_message in a "warn" ACL statement was writing to the reject log as
7329 well as to the main log, which contradicts the documentation and doesn't
7330 seem right (because no rejection is happening). So I have stopped it.
7332 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7333 However, I am unable to do any testing of this.
7335 34. Fixed an infelicity in the appendfile transport. When checking directories
7336 for a mailbox, to see if any needed to be created, it was accidentally
7337 using path names with one or more superfluous leading slashes; tracing
7338 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7340 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7341 contents are added to the log line that is written for every discarded
7342 recipient. (Previously a log_message setting was ignored.)
7344 36. The ${quote: operator now quotes the string if it is empty.
7346 37. The install script runs exim in order to find its version number. If for
7347 some reason other than non-existence or emptiness, which it checks, it
7348 could not run './exim', it was installing it with an empty version number,
7349 i.e. as "exim-". This error state is now caught, and the installation is
7352 38. An argument was missing from the function that creates an error message
7353 when Exim fails to connect to the socket for saslauthd authentication.
7354 This could cause Exim to crash, or give a corrupted message.
7356 39. Added isip, isip4, and isip6 to ${if conditions.
7358 40. The ACL variables $acl_xx are now saved with the message, and can be
7359 accessed later in routers, transports, and filters.
7361 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7362 strings in the file are not string-expanded.
7364 42. If a MAIL command specified a SIZE value that was too large to fit into an
7365 int variable, the check against message_size_limit failed. Such values are
7366 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7367 day this will have to be increased, but I don't think I want to be around
7368 when emails are that large.
7375 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7376 iconv() is not standard in FreeBSD.
7378 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7379 IPv6 enabled. The observed symptom was a segmentation fault on return from
7380 the function os_common_find_running_interfaces() in src/os.c.
7382 3. In the check_special_case() function in daemon.c I had used "errno" as an
7383 argument name, which causes warnings on some systems. This was basically a
7384 typo, since it was named "eno" in the comments!
7386 4. The code that waits for the clock to tick (at a resolution of some fraction
7387 of a second) so as to ensure message-id uniqueness was always waiting for
7388 at least one whole tick, when it could have waited for less. [This is
7389 almost certainly not relevant at current processor speeds, where it is
7390 unlikely to ever wait at all. But we try to future-proof.]
7392 5. The function that sleeps for a time interval that includes fractions of a
7393 second contained a race. It did not block SIGALRM between setting the
7394 timer, and suspending (a couple of lines later). If the interval was short
7395 and the sigsuspend() was delayed until after it had expired, the suspension
7396 never ended. On busy systems this could lead to processes getting stuck for
7399 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7400 process, before it forks any delivery processes. The open lookup caching
7401 mechanism meant that the open file or database connection was passed into
7402 the delivery process. The problem was that delivery processes always tidy
7403 up cached lookup data. This could cause a problem for the next delivery
7404 process started by the queue runner, because the external queue runner
7405 process does not know about the closure. So the next delivery process
7406 still has data in the lookup cache. In the case of a file lookup, there was
7407 no problem because closing a file descriptor in a subprocess doesn't affect
7408 the parent. However, if the lookup was caching a connection to a database,
7409 the connection was closed, and the second delivery process was likely to
7410 see errors such as "PGSQL: query failed: server closed the connection
7411 unexpectedly". The problem has been fixed by closing all cached lookups
7412 in a queue runner before running a delivery process.
7414 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7415 seem to have the "const" qualifier which it has on other OS. I've
7418 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7419 *of the same type* (client or server) with the same public name that an
7420 error should be diagnosed.
7422 9. When Exim looked up a host name for an IP address, but failed to find the
7423 original IP address when looking up the host name (a safety check), it
7424 output the message "<ip address> does not match any IP for NULL", which was
7425 confusing, to say the least. The bug was that the host name should have
7426 appeared instead of "NULL".
7428 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7429 user that is built into the binary, and the -C or -D options is used, root
7430 privilege is dropped before the configuration file is read. In addition,
7431 logging is switched to stderr instead of the normal log files. If the
7432 configuration then re-defines the Exim user, the unprivileged environment
7433 is probably not what is expected, so Exim logs a panic warning message (but
7436 However, if deliver_drop_privilege is set, the unprivileged state may well
7437 be exactly what is intended, so the warning has been cut out in that case,
7438 and Exim is allowed to try to write to its normal log files.
7444 1. smtp_return_error_details was not giving details for temporary sender
7445 or receiver verification errors.
7447 2. Diagnose a configuration error if two authenticators have the same public
7450 3. Exim used not to create the message log file for a message until the first
7451 delivery attempt. This could be confusing when incoming messages were held
7452 for policy or load reasons. The message log file is now created at the time
7453 the message is received, and an initial "Received" line is written to it.
7455 4. The automatically generated man page for command line options had a minor
7456 bug that caused no ill effects; however, a more serious problem was that
7457 the procedure for building the man page automatically didn't always
7458 operate. Consequently, release 4.20 contains an out-of-date version. This
7459 shouldn't happen again.
7461 5. When building Exim with embedded Perl support, the script that builds the
7462 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7463 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7465 6. The freeze_tell option was not being used for messages that were frozen on
7466 arrival, either by an ACL or by local_scan().
7468 7. Added the smtp_incomplete_transaction log selector.
7470 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7471 was accepting AUTH without a new EHLO.
7473 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7474 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7477 10. Logging of TCP/IP connections (when configured) now happens in the main
7478 daemon process instead of the child process, so that the TCP/IP connection
7479 count is more accurate (but it can never be perfect).
7481 11. The use of "drop" in a nested ACL was not being handled correctly in the
7482 outer ACL. Now, if condition failure induced by the nested "drop" causes
7483 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7484 or "require"), the connection is dropped.
7486 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7487 that yield "discard" can now be used with an "accept" or a "discard" verb,
7488 but an error is generated for any others (because I can't see a useful way
7489 to define what should happen).
7491 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7492 are now processed in the same way as lines in the Exim configuration file.
7493 In particular, continuation lines are supported.
7495 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7497 15. Added -ti meaning -t -i.
7499 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7500 domains, and warn by logging if others are found.
7502 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7503 returned from GIFCONF when Exim is trying to find the list of interfaces on
7504 a host. The code in os.c has been modified to copy each ifreq to an aligned
7505 structure in all cases.
7507 Also, in some cases, the returned ifreq's were being copied to a 'struct
7508 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7509 means the last couple of bytes of an IPv6 address could be chopped if the
7510 ifreq contained only a normal sockaddr (14 bytes storage).
7512 18. Named domain lists were not supported in the hosts_treat_as_local option.
7513 An entry such as +xxxx was not recognized, and was treated as a literal
7516 19. Ensure that header lines added by a DATA ACL are included in the reject log
7517 if the ACL subsequently rejects the message.
7519 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7520 MD5 (which is deprecated).
7522 21. When testing a filter file using -bf, Exim was writing a message when it
7523 took the sender from a "From " line in the message, but it was not doing so
7524 when it took $return_path from a Return-Path: header line. It now does.
7526 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7527 with a valid header line field name (a series of printing characters
7528 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7530 23. Changed "disc" in the source to "disk" to conform to the documentation and
7531 the book and for uniformity.
7533 24. Ignore Sendmail's -Ooption=value command line item.
7535 25. When execve() failed while trying to run a command in a pipe transport,
7536 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7537 could be confused with a return value of 69 from the command itself. This
7538 has been changed to 127, the value the shell returns if it is asked to run
7539 a non-existent command. The wording for the related log line suggests a
7540 non-existent command as the problem.
7542 26. If received_header_text expands to an empty string, do not add a Received:
7543 header line to the message. (Well, it adds a token one on the spool, but
7544 marks it "old" so that it doesn't get used or transmitted.)
7546 27. Installed eximstats 1.28 (addition of -nt option).
7548 28. There was no check for failure on the call to getsockname() in the daemon
7549 code. This can fail if there is a shortage of resources on the system, with
7550 ENOMEM, for example. A temporary error is now given on failure.
7552 29. Contrary to the C standard, it seems that in some environments, the
7553 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7554 program. Exim now does this explicitly; it affects the formatting of
7555 timestamps using strftime().
7557 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7558 complaints. I've now initialized all the variables, to avoid this.
7560 32. Header lines added by a system filter were not being "seen" during
7561 transport-time rewrites.
7563 33. The info_callback() function passed to OpenSSL is set up with type void
7564 (*)(SSL *, int, int), as described somewhere. However, when calling the
7565 function (actually a macro) that sets it up, the type void(*)() is
7566 expected. I've put in a cast to prevent warnings from picky compilers.
7568 34. If a DNS black list lookup found a CNAME record, but there were no A
7569 records associated with the domain it pointed at, Exim crashed.
7571 35. If a DNS black list lookup returned more than one A record, Exim ignored
7572 all but the first. It now scans all returned addresses if a particular IP
7573 value is being sought. In this situation, the contents of the
7574 $dnslist_value variable are a list of all the addresses, separated by a
7577 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7578 to accept a host name and all its aliases if the forward lookup for any of
7579 them yielded the IP address of the incoming connection. Now it accepts only
7580 those names whose forward lookup yields the correct IP address. Any other
7581 names are discarded. This closes a loophole whereby a rogue DNS
7582 administrator could create reverse DNS records to break through a
7583 wildcarded host restriction in an ACL.
7585 37. If a user filter or a system filter that ran in a subprocess used any of
7586 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7587 the wrong values were passed to the pipe command ($thisaddress had the
7588 value of $0, $0 had the value of $1, etc). This bug was introduced by
7589 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7591 38. Improved the line breaking for long SMTP error messages from ACLs.
7592 Previously, if there was no break point between 40 and 75 characters, Exim
7593 left the rest of the message alone. Two changes have been made: (a) I've
7594 reduced the minimum length to 35 characters; (b) if it can't find a break
7595 point between 35 and 75 characters, it looks ahead and uses the first one
7596 that it finds. This may give the occasional overlong line, but at least the
7597 remaining text gets split now.
7599 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7600 file descriptors might be low, and that setting 1000 would always raise it.
7601 It turns out that in some environments, the limit is already over 1000 and
7602 that lowering it causes trouble. So now Exim takes care not to decrease it.
7604 40. When delivering a message, the value of $return_path is set to $sender_
7605 address at the start of routing (routers may change the value). By an
7606 oversight, this default was not being set up when an address was tested by
7607 -bt or -bv, which affected the outcome if any router or filter referred to
7610 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7611 log only when "message" or "log_message" are set. However, if one of the
7612 conditions was an address verification, or a call to a nested ACL, the
7613 messages generated by the underlying test were being passed through. This
7614 no longer happens. The underlying message is available in $acl_verify_
7615 message for both "message" and "log_message" expansions, so it can be
7616 passed through if needed.
7618 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7619 new expansion $bh_ to give the encoded byte string without charset
7620 translation. Translation happens only if iconv() is available; HAVE_ICONV
7621 indicates this at build time. HEADERS_CHARSET gives the charset to
7622 translate to; headers_charset can change it in the configuration, and
7623 "headers charset" can change it in an individual filter file.
7625 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7626 that creates RFC 2047 encoded "words" labels them as that charset instead
7627 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7628 expansion operator; (ii) when Exim creates a From: line for a local
7629 message; (iii) when a header line is rewritten to include a "phrase" part.
7631 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7632 buggy, causing it to skip the first lines of messages whose message ID
7633 ended in 'D'. This would not have bitten before Exim release 4.14, because
7634 message IDs were unlikely to end in 'D' before then. The effect was to have
7635 incorrect size information for certain domains.
7637 45. #include "config.h" was missing at the start of the crypt16.c module. This
7638 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7641 46. If there was a timeout during a "random" callout check, Exim treated it as
7642 a failure of the random address, and carried on sending RSET and the real
7643 address. If the delay was just some slowness somewhere, the response to the
7644 original RCPT would be taken as a response to RSET and so on, causing
7645 mayhem of various kinds.
7647 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7648 when I implemented it. It didn't allow for the fact that some option values
7649 may legitimately be negative (e.g. size_addition), and it didn't even do
7650 the right test for positive values.
7652 48. Domain names in DNS records are case-independent. Exim always looks them up
7653 in lower case. Some resolvers return domain names in exactly the case they
7654 appear in the zone file, that is, they may contain uppercase letters. Not
7655 all resolvers do this - some return always lower case. Exim was treating a
7656 change of case by a resolver as a change of domain, similar to a widening
7657 of a domain abbreviation. This triggered its re-routing code and so it was
7658 trying to route what was effectively the same domain again. This normally
7659 caused routing to fail (because the router wouldn't handle the domain
7660 twice). Now Exim checks for this case specially, and just changes the
7661 casing of the domain that it ultimately uses when it transmits the message
7664 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7667 50. If a filter generated a file delivery with a non-absolute name (possible if
7668 no home directory exists for the router), the forbid_file option was not
7671 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7672 the existing equality matching.
7674 52. Exim was using ints instead of ino_t variables in some places where it was
7675 dealing with inode numbers.
7677 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7678 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7679 called TMPDIR, and if it finds it is different, it changes its value.
7681 54. The smtp_printf() function is now made available to local_scan() so
7682 additional output lines can be written before returning. There is also an
7683 smtp_fflush() function to enable the detection of a dropped connection.
7684 The variables smtp_input and smtp_batched_input are exported to
7687 55. Changed the default runtime configuration: the message "Unknown user"
7688 has been removed from the ACL, and instead placed on the localuser router,
7689 using the cannot_route_message feature. This means that any verification
7690 failures that generate their own messages won't get overridden. Similarly,
7691 the "Unrouteable address" message that was in the ACL for unverifiable
7692 relay addresses has also been removed.
7694 56. Added hosts_avoid_esmtp to the smtp transport.
7696 57. The exicyclog script was not checking for the esoteric option
7697 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7698 will work only if exicyclog is run under the appropriate euid.
7700 58. Following a discussion on the list, the rules by which Exim recognises line
7701 endings on incoming messages have been changed. The -dropcr and drop_cr
7702 options are now no-ops, retained only for backwards compatibility. The
7703 following line terminators are recognized: LF CRLF CR. However, special
7704 processing applies to CR:
7706 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7707 nor a local message in the state where . is a terminator.
7709 (ii) If a bare CR is encountered in a header line, an extra space is added
7710 after the line terminator so as not to end the header. The reasoning
7711 behind this is that bare CRs in header lines are most likely either
7712 to be mistakes, or people trying to play silly games.
7714 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7715 was being incorrectly given as 18 bytes larger than it should have been.
7716 This is a VOB (very old bug).
7718 60. This may never have affected anything current, but just in case it has:
7719 When the local host is found other than at the start of a list of hosts,
7720 the local host, those with the same MX, and any that follow, are discarded.
7721 When the list in question was part of a longer list of hosts, the following
7722 hosts (not currently being processed) were also being discarded. This no
7723 longer happens. I'm not sure if this situation could ever has previously
7726 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7729 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7730 header. This is something that is recommended in a new Internet Draft, and
7731 is something that is documented as being done by Sendmail. There are two
7732 possible values. For messages generated by the autoreply transport, Exim
7735 Auto-Submitted: auto-replied
7737 whereas for all other generated messages (e.g. bounces) it adds
7739 Auto-Submitted: auto-generated
7741 63. The "personal" condition in filters now includes a test for the
7742 Auto-Submitted: header. If it contains the string "auto-" the message it
7743 not considered personal.
7745 64. Added rcpt_include_affixes as a generic transport option.
7747 65. Added queue_only_override (default true).
7749 66. Added the syslog_duplication option.
7751 67. If what should have been the first header line of a message consisted of
7752 a space followed by a colon, Exim was mis-interpreting it as a header line.
7753 It isn't of course - it is syntactically invalid and should therefore be
7754 treated as the start of the message body. The misbehaviour could have
7755 caused a number of strange effects, including loss of data in subsequent
7756 header lines, and spool format errors.
7758 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7759 client host had authenticated. This control can now be exercised by an ACL
7760 for more flexibility.
7762 69. By default, callouts do not happen when testing with -bh. There is now a
7763 variant, -bhc, which does actually run the callout code, including
7764 consulting and updating the callout cache.
7766 70. Added support for saslauthd authentication, courtesy of Alexander
7769 71. If statvfs() failed on the spool or log directories while checking their
7770 size for availability, Exim confusingly gave the error "space shortage".
7771 Furthermore, in debugging mode it crashed with a floating point exception.
7772 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7773 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7774 serious problem, Exim now writes to the main and panic logs when this
7775 happens, with details of the failure. It then refuses to accept the
7776 incoming message, giving the message "spool directory problem" or "log
7777 directory problem" with a 421 code for SMTP messages.
7779 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7780 0, 1, and 2 exist, because some OS complain for execs without them (see
7781 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7782 descriptors. However, the code omitted to check that the open succeeded,
7783 causing mysterious errors if for some reason the permissions on /dev/null
7784 got screwed. Now Exim writes a message to the main and panic logs, and
7785 bombs out if it can't open /dev/null.
7787 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7788 interact so that it is all more flexible. It is supposed to remain
7789 backwards compatible. Also added extra_local_interfaces.
7791 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7792 to bomb out with an assertion failure - to the client this appears as a
7793 connection drop. This problem occurs in the part of the code that was taken
7794 from the Samba project. Fortunately, the assertion is in a very simple
7795 function, so I have fixed this by reproducing the function inline in the
7796 one place where it is called, and arranging for authentication to fail
7797 instead of killing the process with assert().
7799 75. The SPA client code was not working when the server requested OEM rather
7800 than Unicode encoding.
7802 76. Added code to make require_files with a specific uid setting more usable in
7803 the case where statting the file as root fails - usually a non-root-mounted
7804 NFS file system. When this happens and the failure is EACCES, Exim now
7805 forks a subprocess and does the per-uid checking as the relevant uid.
7807 77. Added process_log_path.
7809 78. If log_file_path was not explicitly set, a setting of check_log_space or
7810 check_log_inodes was ignored.
7812 79. If a space check for the spool or log partitions fails, the incident is now
7813 logged. Of course, in the latter case the data may get lost...
7815 80. Added the %p formatting code to string_format() so that it can be used to
7816 print addresses in debug_print(). Adjusted all the address printing in the
7817 debugging in store.c to use %p rather than %d.
7819 81. There was a concern that a line of code in smtp_in.c could overflow a
7820 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7821 initially expressed, the concern was not well-founded, because trailing
7822 spaces are removed early. However, if the trailing spaces were followed by
7823 a NULL, they did not get removed, so the overflow was possible. Two fixes
7826 (a) I re-wrote the offending code in a cleaner fashion.
7827 (b) If an incoming SMTP command contains a NULL character, it is rejected
7830 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7831 runs initgroups(), so that if the Exim user is in any additional groups,
7832 they will be used during message reception.
7838 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob