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.
55 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
56 was given for the sqlite_dbfile a trap resulted.
58 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
59 "name" argument resulted in a trap. There is no reason to disallow such;
60 this was a coding error.
62 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
63 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
64 spent suspended, ignoring the Posix definition. Previously we assumed
65 it did and a constant offset from real time could be used as a correction.
66 Change to using the same clock source for the start-of-message and the
67 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
68 exists, just to get a clock slightly more aligned to reality.
70 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
71 RFC says it is optional some validators care. The missing char was not
72 intended but triggered by a line-wrap alignement. Discovery and fix by
73 Guillaume Outters, hacked on by JH.
75 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
76 name being quoted was tainted a trap would be taken. Fix by using
77 dynamicaly created buffers. The routine could have been called by a
78 rewrite with the "h" flag, by using the "-F" command-line option, or
79 by using a "name=" option on a control=submission ACL modifier.
81 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
82 Previously when a whitespace character was specified it was not inserted
83 after removing the newline.
85 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
86 is_tainted() had an off-by-one error in the overenthusiastic direction.
87 Find and fix by Gavan. Although NetBSD is not a supported platform for
88 4.94 this bug could affect other platforms.
90 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
91 is_tainted() had an off-by-one error in the overenthusiastic direction.
92 Find and fix by Gavan. Although NetBSD is not a supported platform for
93 4.94 this bug could affect other platforms.
94 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
95 Previously when a whitespace character was specified it was not inserted
96 after removing the newline.
98 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
99 the domain part of the recipient address. This overrides any tls_sni
100 option set, which was previously used.
102 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
105 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
106 than one server was defined and depending on the platform memory layout
107 details, an internal consistency trap could be hit while walking the list
110 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
111 files. The value is used by the authresults expansion item. Previously
112 if this was used in a router or transport, a crash could result.
114 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
115 applied. This resulted, if any header-line rewrite rules were configured,
116 in a panic-log trigerrable by sending a message with a long address in
117 a header. Fix by increaing the arbitrary limit to larger than a single
118 (dewrapped) 5322 header line maximum size.
120 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
121 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
122 RFC 6648 which deprecates X- options in protocols as a general practice.
123 Changeover between the implementations is handled by the mechanisms
126 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
127 a remote which uses callout/hold. Previously the local was queued.
129 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
132 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
134 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
135 authenticator client_send option. Previously the next char, after a pair
136 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
137 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
138 documentation. There is still no way to get a leading ^ immediately
139 after a NUL (ie. for the password of a PLAIN method authenticator.
141 JH/39 Bug 2691: fix $local_part_data. When the matching list element
142 referred to a file, bad data was returned. This likely also affected
145 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
146 left undeleted; the attempt to re-create it then failed - resulting in
147 the usual "SIGHUP tp have daemon reload configuration" to not work.
148 This affected any platform not supporting "abstract" Unix-domain
149 sockets (i.e. not Linux).
151 JH/42 Bug 2692: Harden against a peer which reneges on a 452 "too many
152 recipients" response to RCPT in a later response, with a 250. The
153 previous coding assumed this would not happen, and under PIPELINING
154 would result in both lost and duplicate recipients for a message.
156 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
157 Previously the weighting was incorrectly applied. Similar fix for socks
158 proxies. Found and fixed by Heiko Schlichting.
160 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
161 not handle sub-lists included using the +namedlist syntax. While
162 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
163 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
166 HS/01 Enforce absolute PID file path name.
168 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
170 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
172 PP/02 Bug 2643: Correct TLS DH constants.
173 A missing NUL termination in our code-generation tool had led to some
174 incorrect Diffie-Hellman constants in the Exim source.
175 Reported by kylon94, code-gen tool fix by Simon Arlott.
177 PP/03 Impose security length checks on various command-line options.
178 Fixes CVE-2020-SPRSS reported by Qualys.
180 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
181 better. Reported by Qualys.
183 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
184 providing a particularly obnoxious sender full name.
187 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
189 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
192 PP/08 Change default for recipients_max from unlimited to 50,000.
194 PP/09 Fix security issue with too many recipients on a message (to remove a
195 known security problem if someone does set recipients_max to unlimited,
196 or if local additions add to the recipient list).
197 Fixes CVE-2020-RCPTL reported by Qualys.
199 PP/10 Fix security issue in SMTP verb option parsing
200 Fixes CVE-2020-EXOPT reported by Qualys.
202 PP/11 Fix security issue in BDAT state confusion.
203 Ensure we reset known-good where we know we need to not be reading BDAT
204 data, as a general case fix, and move the places where we switch to BDAT
205 mode until after various protocol state checks.
206 Fixes CVE-2020-BDATA reported by Qualys.
208 HS/03 Die on "/../" in msglog file names
210 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the euid of
211 the Exim runtime user are allowed to create files.
217 JH/01 Avoid costly startup code when not strictly needed. This reduces time
218 for some exim process initialisations. It does mean that the logging
219 of TLS configuration problems is only done for the daemon startup.
221 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
223 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
224 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
227 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
228 DKIM signing is being done. Previously a transport_filter would always
229 disable CHUNKING, falling back to traditional DATA.
231 JH/05 Regard command-line receipients as tainted.
233 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
235 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
236 PAM library frees one of the arguments given to it, despite the
237 documentation. Therefore a plain malloc must be used.
239 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
240 on-stack buffers were used, resulting in a taint trap when DSN information
241 copied from a received message was written into the buffer.
243 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
244 the ordering of its ARC headers. This caused a crash.
246 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
247 a new record was being constructed with information from the peer, a trap
250 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
251 installation would get error messages from DMARC verify, when it hit the
252 nonexistent file indicated by the default. Distros wanting DMARC enabled
253 should both provide the file and set the option.
254 Also enforce no DMARC verification for command-line sourced messages.
256 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
257 could, depending on the platform, hang at the STARTTLS response.
259 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
260 message on a connection. Previously if one message had ARC headers and
261 the following one did not, a crash could result when adding an
262 Authentication-Results: header.
264 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
265 Exim main code and Exim-related utities. The introduction of taint
266 tracking also did many adjustments to string handling. Since then, eximon
267 frequently terminated with an assert failure.
269 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
270 check for 452 responses. This slightly helps the inefficieny of doing
271 a large alias-expansion into a recipient-limited target. The max_rcpt
272 transport option still applies (and at the current default, will override
273 the new feature). The check is done for either cause of synch, and forces
274 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
275 connection. The new facility is not tunable at this time.
277 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
278 library live data was being used, so the results became garbage. Make
279 copies while it is still usable.
281 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
282 on delivery deferred (==) and failed (**) lines (if a delivery was
283 attemtped). Previously it was only on completion (=>) lines.
285 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
286 for the expansion of the server_scram_iter and server_scram_salt options.
288 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
289 are now specifically given a NO_DATA response without hitting the system
290 resolver. The library goes on to do the now-standard TXT lookup.
291 Use of dnsdb lookups is not affected.
293 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
294 only retrieve the errormessage once. Previously two calls to dlerror()
295 were used, and the second one (for mainlog/paniclog) retrieved null
298 JH/20 Taint checking: disallow use of tainted data for
299 - the appendfile transport file and directory options
300 - the pipe transport command
301 - the autoreply transport file, log and once options
302 - file names used by the redirect router (including filter files)
304 - paths used by single-key lookups
305 Previously this was permitted.
307 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
308 adjusted the size of a major service buffer; this failed because the
309 buffer was in use at the time. Change to a compile-time increase in the
310 buffer size, when this authenticator is compiled into exim.
312 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
313 previous fast-mode was untenable in the face of glibs using mmap to
314 support larger malloc requests.
316 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
317 New values supported, if defined on system where compiled:
318 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
319 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
321 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
322 running a limited number of proceses in parallel, a benefit is gained. The
323 amount varies with the platform hardware and load. The use of the option
324 queue_run_in_order means we cannot do this, as ordering becomes
327 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
328 had introduced a string-copy (for ensuring NUL-termination) which was not
329 appropriate for that case, which can include embedded NUL bytes in the
330 block of data. Investigation showed the copy to actually be needless, the
331 data being length-specified.
333 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
334 done during a receiving connection, and both used TLS, global info was
335 used rather than per-connection info for tracking the state of data
336 queued for transmission. This could result in a connection hang.
338 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
339 Previously, when delivering serveral messages down a single connection
340 only the first would provide a SIZE. This was due to the size information
341 not being properly tracked.
343 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
344 TAI (at 37 seconds currently), pretend to be in UTC for time-related
345 expansion and logging. Previously, spurious values such as a future
346 minute could be seen.
348 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
349 it could crash from a null-deref. This could also affect the
350 ${addresses: } operator and ${readsock } item.
352 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
353 message following a mime one, the variable was not reset.
355 JH/30 When an pipelined-connect fails at the first response, assume incorrect
356 cached capability (perhaps the peer reneged?) and immediately retry in
359 JH/31 Fix spurious detection of timeout while writing to transport filter.
361 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
362 an attempt to copy the string was made before checking it.
364 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
365 taint of the lookup key was maintained; we now regard the presence in the
366 filesystem as sufficient validation.
368 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
369 argument is supplied.
371 JH/35 The dsearch lookup now requires that the directory is an absolute path.
372 Previously this was not checked, and nonempty relative paths made an
373 access under Exim's current working directory.
375 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
376 Previously no event was raised.
378 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
379 parameter supplied by the sender MAIL FROM command. Previously it was
380 ignored, and only the check_spool_space option value for the required
383 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
384 the size of the signing public-key. Previously it was instead giving
385 the size of the signature hash.
387 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
388 the default. See the (new) dkim_verify_min_keysizes option.
390 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
391 and an ACL use a lookup for checking either the local_part or domain,
392 stale data could be accessed. Ensure that variable references are
393 dropped between messages.
395 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
396 by the client was not checked as pointing within response data before
397 being used. A malicious client could thus cause an out-of-bounds read and
398 possibly gain authentication. Fix by adding the check.
400 JH/42 Internationalisation: change the default for downconversion in the smtp
401 transport to be "if needed". Previously it was "as previously set" for
402 the message, which usually meant "if needed" for message-submission but
403 "no" for everything else. However, MTAs have been seen using SMTPUTF8
404 even when the envelope addresses did not need it, resulting in forwarding
405 failures to non-supporting MTAs. A downconvert in such cases will be
406 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
407 transport. The change does mean that addresses needing conversion will
408 be converted when previously a delivery failure would occur.
410 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
411 response was received it would be used unchecked in a fail-DSN, violating
412 standards on line-length limits. Truncate if needed.
414 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
415 doesn't work. (Additionally add a new main config option to configure the
422 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
423 side, to decode a TLS 1.3 packet capture.
425 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
426 Previously the default library behaviour applied, sending two, each in
429 JH/03 Debug output for ACL now gives the config file name and line number for
432 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
434 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
436 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
437 buffer overrun for (non-chunking) other transports.
439 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
440 TLS1.3, means that a server rejecting a client certificate is not visible
441 to the client until the first read of encrypted data (typically the
442 response to EHLO). Add detection for that case and treat it as a failed
443 TLS connection attempt, so that the normal retry-in-clear can work (if
444 suitably configured).
446 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
447 and/or domain. Found and fixed by Jason Betts.
449 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
450 configuration). If a CNAME target was not a wellformed name pattern, a
453 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
454 the OS reports them interleaved with other addresses.
456 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
457 used both for input and for a verify callout, both encrypted, SMTP
458 responses being sent by the server could be lost. This resulted in
459 dropped connections and sometimes bounces generated by a peer sending
462 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
463 string. Previously it was possible to cause undefined behaviour in a
464 library routine (usually a crash). Found by "zerons".
466 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
469 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
470 API was removed, so update to use the newer ones.
472 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
473 any timeout set, is taking a long time. Previously we would hang on to a
474 rotated logfile "forever" if the input was arriving with long gaps
475 (a previous attempt to fix addressed lack, for a long time, of initial
478 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
479 shared (NFS) environment. The length of the tempfile name is now
480 4 + 16 ("hdr.$message_exim_id") which might break on file
481 systems which restrict the file name length to lower values.
484 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
485 shared (NFS) environment.
487 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
488 did for all versions <4.90). Notably -M, -m, --invert, -I may be
491 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
492 on some platforms for bit 31.
494 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
495 to changes apparently associated with TLS1.3 handling some of the APIs
496 previously used were either nonfunctional or inappropriate. Strings
497 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
498 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
499 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
500 This affects log line X= elements, the $tls_{in,out}_cipher variables,
501 and the use of specific cipher names in the encrypted= ACL condition.
503 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
505 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
506 verification result was not updated unless hosts_require_ocsp applied.
508 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
509 queue_list_requires_admin set to false, non-admin users were denied the
512 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
513 directory-of-certs mode. Previously they were advertised despite the
516 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
517 A single TCP connection by a client will now hold a TLS connection open
518 for multiple message deliveries, by default. Previously the default was to
521 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
522 default. If built with the facility, DANE will be used. The facility
523 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
525 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
526 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
527 must be defined and you must still, unless you define DISABLE_TLS, manage
528 the the include-dir and library-file requirements that go with that
529 choice. Non-TLS builds are still supported.
531 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
534 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
535 default. If the platform supports and has the facility enabled, it will
536 be requested on all coneections.
538 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
539 controlled by the build-time option SUPPORT_PIPE_CONNECT.
541 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
543 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
544 success-DSN messages. Previously the From: header was always the default
545 one for these; the option was ignored.
547 JH/28 Fix the timeout on smtp response to apply to the whole response.
548 Previously it was reset for every read, so a teergrubing peer sending
549 single bytes within the time limit could extend the connection for a
550 long time. Credit to Qualsys Security Advisory Team for the discovery.
552 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
553 delivery address, which leaked information of the results of local
554 forwarding. Change to the original envelope recipient address, per
557 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
558 requested. Previously not bounce was generated and a log entry of
559 error ignored was made.
561 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
563 JH/32 Introduce a general tainting mechanism for values read from the input
564 channel, and values derived from them. Refuse to expand any tainted
565 values, to catch one form of exploit.
567 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
568 was unused and the unexpanded text used for the test. Found and
569 fixed by Ruben Jenster.
571 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
572 an attempt to use a TLS library read routine dereffed a nul pointer,
575 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
576 them in our smtp response.
578 JH/36 Have the general router option retry_use_local_part default to true when
579 any of the restrictive preconditions are set (to anything). Previously it
580 was only for check_local user. The change removes one item of manual
581 configuration which is required for proper retries when a remote router
582 handles a subset of addresses for a domain.
584 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
585 link count into consideration.
587 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
588 caused the extension of big_buffer, the following lines were ignored.
590 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
591 accordance with RFC 2308. Previously there was no expiry, so a longlived
592 receive process (eg. due to ACL delays) versus a short SOA value could
595 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
597 JH/39 Promote DMARC support to mainline.
599 JH/40 Bug 2452: Add a References: header to DSNs.
601 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
602 parameters. The relevant library call is documented as "Deprecated: This
603 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
604 3.6.0, DH parameters are negotiated following RFC7919."
606 HS/06 Change the default of dnssec_request_domains to "*"
608 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
609 carried on and emitted a BDAT command, even when PIPELINING was not
612 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
613 buffer was used for the filename, resulting in a trap when tainted
614 arguments (eg. $domain) were used.
616 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
617 recommended to avoid a possible server-load attack. The feature can be
618 re-enabled via the openssl_options main cofiguration option.
620 JH/45 local_scan API: documented the current smtp_printf() call. This changed
621 for version 4.90 - adding a "more data" boolean to the arguments.
622 Bumped the ABI version number also, this having been missed previously;
623 release versions 4.90 to 4.92.3 inclusive were effectively broken in
624 respect of usage of smtp_printf() by either local_scan code or libraries
625 accessed via the ${dlfunc } expansion item. Both will need coding
626 adjustment for any calls to smtp_printf() to match the new function
627 signature; a FALSE value for the new argument is always safe.
629 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
630 the file-offset (which the Linux syscall does, and exim expects); this
631 resulted in an indefinite loop.
633 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
634 to do ARC verification. The Authentication-Results: header line added
635 by the configuration then had no ARC item.
641 JH/01 Remove code calling the customisable local_scan function, unless a new
642 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
644 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
645 non-signal-safe functions being used.
647 JH/03 Bug 2269: When presented with a received message having a stupidly large
648 number of DKIM-Signature headers, disable DKIM verification to avoid
649 a resource-consumption attack. The limit is set at twenty.
651 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
652 report of oldest_pass in ${authres } in consequence, and separate out
653 some descriptions of reasons for verification fail.
655 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
656 files in the spool were present and unlocked. A queue-runner could spot
657 them, resulting in a duplicate delivery. Fix that by doing the unlock
658 after the unlink. Investigation by Tim Stewart. Take the opportunity to
659 add more error-checking on spoolfile handling while that code is being
662 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
663 No known attacks, no CVE, this is defensive hardening.
665 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
666 a queue-runner could start a delivery while other operations were ongoing.
667 Cutthrough delivery was a common victim, resulting in duplicate delivery.
668 Found and investigated by Tim Stewart. Fix by using the open message data
669 file handle rather than opening another, and not locally closing it (which
670 releases a lock) for that case, while creating the temporary .eml format
671 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
673 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
674 $sender_verify_failure/$recipient_verify_failure to "random".
676 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
679 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
680 Previously this would segfault.
682 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
685 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
686 like zero, since the resolver should be doing this for us, But we need one
687 as a CNAME but no MX presence gets the CNAME returned; we need to check
688 that doesn't point to an MX to declare it "no MX returned" rather than
689 "error, loop". A new main option is added so the older capability of
690 following some limited number of chain links is maintained.
692 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
694 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
695 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
696 now-deprecated earlier definitions used only specified the range up to TLS
697 1.2 (in the older-version library docs).
699 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
701 JH/15 Rework TLS client-side context management. Stop using a global, and
702 explicitly pass a context around. This enables future use of TLS for
703 connections to service-daemons (eg. malware scanning) while a client smtp
704 connection is using TLS; with cutthrough connections this is quite likely.
706 JH/16 Fix ARC verification to do AS checks in reverse order.
708 JH/17 Support a "tls" option on the ${readsocket } expansion item.
710 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
711 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
712 Previously the "utf8" would be re-prepended for every additional message.
714 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
715 Previously thery were accepted, resulting in issues when attempting to
716 forward messages to a non-supporting MTA.
718 PP/02 Let -n work with printing macros too, not just options.
720 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
721 one parent address was copied, and bogus data was used at delivery-logging
722 time. Either a crash (after delivery) or bogus log data could result.
723 Discovery and analysis by Tim Stewart.
725 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
726 Previously if the string ended mid-character, we did not insert the
727 promised '?' replacement.
729 PP/04 Documentation: current string operators work on bytes, not codepoints.
731 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
732 should pack well giving a smaller memory footprint so better caching and
733 therefore performance. Group the declarations where this can't be done so
734 that the byte-sized flag variables are not interspersed among pointer
735 variables, giving a better chance of good packing by the compiler.
737 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
738 non-null, to avoid issues with sites running BATV. Previously reports were
739 sent with an empty envelope sender so looked like bounces.
741 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
742 The ignore_error flag wasn't being returned from the filter subprocess so
743 was not set for later routers. Investigation and fix by Matthias Kurz.
745 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
746 and a msg:complete for the whole, when a message is manually removed using
747 -Mrm. Developement by Matthias Kurz, hacked on by JH.
749 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
750 a "Gnu special" function, asprintf() in the DB utility binary builds; I
751 hope that is portable enough.
753 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
754 requiring a known-CA anchor certificate; make it now rely entirely on the
755 TLSA as an anchor. Checking the name on the leaf cert against the name
756 on the A-record for the host is still done for TA (but not for EE mode).
758 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
759 included in delivery lines for non-proxied connections, when compiled with
760 SUPPORT_SOCKS and running with proxy logging enabled.
762 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
763 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
764 move the existing event to fire before the normal logging of message
765 failure so that custom logging is bracketed by normal logging.
767 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
768 msg:fail:internal event. Developement by Matthias Kurz.
770 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
771 far too small for todays use of crypto signatures stored there. Go all
772 the way to the max DNS message size of 64kB, even though this might be
773 overmuch for IOT constrained device use.
775 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
776 copy a string over itself. The library routine is documented as not
777 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
779 JH/32 For main options check_spool_space and check_inode_space, where the
780 platform supports 64b integers, support more than the previous 2^31 kB
781 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
782 the previous G, M, k.
784 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
785 $authenticated_fail_id variable on authentication failure. Previously
788 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
789 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
790 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
791 GNUTLS_SEC_PARAM_MEDIUM.
793 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
794 side. Previously we would continue as if no SNI had been received.
796 JH/36 Harden the handling of string-lists. When a list consisted of a sole
797 "<" character, which should be a list-separator specification, we walked
798 off past the nul-terimation.
800 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
801 causes) even when the retry time is not yet met. Previously they were
802 not, meaning that when (say) an account was over-quota and temp-rejecting,
803 and multiple senders' messages were queued, only one sender would get
804 notified on each configured delay_warning cycle.
806 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
808 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
809 cipher-suites, an error can be left on the stack even for a succeeding
810 accept; this results in impossible error messages when a later operation
813 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
814 return error codes indicating retry. Under TLS1.3 this becomes required.
816 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
817 it only wrote the new authenticators, resulting in a lack of tracking of
818 peer changes of ESMTP extensions until the next cache flush.
820 JH/41 Fix the loop reading a message header line to check for integer overflow,
821 and more-often against header_maxsize. Previously a crafted message could
822 induce a crash of the recive process; now the message is cleanly rejected.
824 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
825 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
831 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
832 When redis_servers is set to a list of > 1 element, and the Redis servers
833 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
834 case of MOVED into a DEFER case instead, thus moving the query onto the
835 next server in the list. For a cluster of N elements, all N servers must
836 be defined in redis_servers.
838 GF/02 Catch and remove uninitialized value warning in exiqsumm
839 Check for existence of @ARGV before looking at $ARGV[0]
841 JH/01 Replace the store_release() internal interface with store_newblock(),
842 which internalises the check required to safely use the old one, plus
843 the allocate and data copy operations duplicated in both (!) of the
844 extant use locations.
846 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
847 modifier. This matches the restriction on the commandline.
849 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
850 Previously only the last row was returned.
852 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
853 we assumed that tags in the header were well-formed, and parsed the
854 element content after inspecting only the first char of the tag.
855 Assumptions at that stage could crash the receive process on malformed
858 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
859 While running the DKIM ACL we operate on the Permanent memory pool so that
860 variables created with "set" persist to the DATA ACL. Also (at any time)
861 DNS lookups that fail create cache records using the Permanent pool. But
862 expansions release any allocations made on the current pool - so a dnsdb
863 lookup expansion done in the DKIM ACL releases the memory used for the
864 DNS negative-cache, and bad things result. Solution is to switch to the
865 Main pool for expansions.
866 While we're in that code, add checks on the DNS cache during store_reset,
867 active in the testsuite.
868 Problem spotted, and debugging aided, by Wolfgang Breyha.
870 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
871 When none of the hosts presented to a transport match an already-open
872 connection, close it and proceed with the list. Previously we would
873 queue the message. Spotted by Lena with Yahoo, probably involving
876 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
877 Previously a spurious "250 OK id=" response was appended to the proper
880 JH/08 The "support for" informational output now, which built with Content
881 Scanning support, has a line for the malware scanner interfaces compiled
882 in. Interface can be individually included or not at build time.
884 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
885 by the template makefile "src/EDITME". The "STREAM" support for an older
886 ClamAV interface method is removed.
888 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
889 rows affected is given instead).
891 JH/11 The runtime Berkeley DB library version is now additionally output by
892 "exim -d -bV". Previously only the compile-time version was shown.
894 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
895 SMTP connection. Previously, when one had more receipients than the
896 first, an abortive onward connection was made. Move to full support for
897 multiple onward connections in sequence, handling cutthrough connection
898 for all multi-message initiating connections.
900 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
901 routers. Previously, a multi-recipient message would fail to match the
902 onward-connection opened for the first recipient, and cause its closure.
904 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
905 a timeout on read on a GnuTLS initiating connection, resulting in the
906 initiating connection being dropped. This mattered most when the callout
907 was marked defer_ok. Fix to keep the two timeout-detection methods
910 JH/15 Relax results from ACL control request to enable cutthrough, in
911 unsupported situations, from error to silently (except under debug)
912 ignoring. This covers use with PRDR, frozen messages, queue-only and
915 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
917 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
918 metadata, resulting in a crash in free().
920 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
921 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
922 Broken also in d185889f4, with init system revamp.
924 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
925 Previously we abruptly closed the connection after reading a malware-
926 found indication; now we go on to read the "scan ok" response line,
929 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
930 ACL. Previously, a crash would result.
932 JH/19 Speed up macro lookups during configuration file read, by skipping non-
933 macro text after a replacement (previously it was only once per line) and
934 by skipping builtin macros when searching for an uppercase lead character.
936 JH/20 DANE support moved from Experimental to mainline. The Makefile control
937 for the build is renamed.
939 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
940 was allocated for every new TLS startup, meaning one per message. Fix
941 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
943 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
944 reported the original. Fix to report (as far as possible) the ACL
945 result replacing the original.
947 JH/23 Fix memory leak during multi-message connections using STARTTLS under
948 OpenSSL. Certificate information is loaded for every new TLS startup,
949 and the resources needed to be freed.
951 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
953 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
956 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
957 DATA response info to the (existing) per-recipient response info for
958 the "C=" log element. It can have useful tracking info from the
959 destination system. Patch from Simon Arlott.
961 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
962 length value. Previously this would segfault.
964 HS/02 Support Avast multiline protoocol, this allows passing flags to
965 newer versions of the scanner.
967 JH/28 Ensure that variables possibly set during message acceptance are marked
968 dead before release of memory in the daemon loop. This stops complaints
969 about them when the debug_store option is enabled. Discovered specifically
970 for sender_rate_period, but applies to a whole set of variables.
971 Do the same for the queue-runner and queue-list loops, for variables set
972 from spool message files. Do the same for the SMTP per-message loop, for
973 certain variables indirectly set in ACL operations.
975 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
976 as a multi-recipient message from a mailinglist manager). The coding had
977 an arbitrary cutoff number of characters while checking for more input;
978 enforced by writing a NUL into the buffer. This corrupted long / fast
979 input. The problem was exposed more widely when more pipelineing of SMTP
980 responses was introduced, and one Exim system was feeding another.
981 The symptom is log complaints of SMTP syntax error (NUL chars) on the
982 receiving system, and refused recipients seen by the sending system
983 (propating to people being dropped from mailing lists).
984 Discovered and pinpointed by David Carter.
986 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
987 replaced by the ${authresults } expansion.
989 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
991 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
992 allows proper process termination in container environments.
994 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
995 Previously the "final dot" had a newline after it; ensure it is CR,LF.
997 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
998 and "err_perm", deprecated since 4.83 when the RFC-defined words
999 "temperror" and "permerror" were introduced.
1001 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1002 transport-filters or DKIM-signing. The restriction was lost in the
1003 consolidation of verify-callout and delivery SMTP handling.
1004 Extend the restriction to also cover ARC-signing.
1006 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1007 in defer=pass mode supply a 450 to the initiator. Previously the message
1010 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1011 tls_require_ciphers is used as before.
1013 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1014 "pass_unscanned". Only tmpfails from the scanner are written to
1015 the paniclog, as they may require admin intervention (permission
1016 denied, license issues). Other scanner errors (like decompression
1017 bombs) do not cause a paniclog entry.
1019 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1020 Previously it was possible to log spurious information in receive log
1023 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1024 triggered odd behaviour from Outlook Express clients.
1026 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1029 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1030 since the IETF WG has not yet settled on that versus the original
1031 "bare" representation.
1033 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1034 Previously the millisecond value corrupted the output.
1035 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1036 corrupted the output.
1042 JH/01 Rework error string handling in TLS interface so that the caller in
1043 more cases is responsible for logging. This permits library-sourced
1044 string to be attached to addresses during delivery, and collapses
1045 pairs of long lines into single ones.
1047 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1048 during configuration. Wildcards are allowed and expanded.
1050 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1051 This permits better logging.
1053 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1054 do not close it down and have the child transport start it up again on
1055 the passed-on TCP connection. Instead, proxy the child (and any
1056 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1057 affected: the continued delivery log lines do not have any DNSSEC, TLS
1058 Certificate or OCSP information. TLS cipher information is still logged.
1060 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1061 identical IP addresses on different listening ports. Will also affect
1064 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1065 add noisy ifdef guards to special-case this sillyness.
1066 Patch from Bernd Kuhls.
1068 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1069 than 255 are no longer allowed.
1071 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1072 Disable the session-cache too, which might reduce our load. Since we
1073 currrectly use a new context for every connection, both as server and
1074 client, there is no benefit for these.
1075 GnuTLS appears to not support tickets server-side by default (we don't
1076 call gnutls_session_ticket_enable_server()) but client side is enabled
1077 by default on recent versions (3.1.3 +) unless the PFS priority string
1080 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1081 <https://reproducible-builds.org/specs/source-date-epoch/>.
1083 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1084 the check for any unsuccessful recipients did not notice the limit, and
1085 erroneously found still-pending ones.
1087 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1088 MSG_MORE. Only in-clear (not on TLS connections).
1090 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1091 transport-filter is involved we can buffer the headers in memory for
1092 creating the signature, and read the spool data file once for the
1093 signature and again for transmission.
1095 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1096 4.77 as the kernel support then wasn't solid, having issues in 64bit
1097 mode. Now, it's been long enough. Add support for FreeBSD also.
1099 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1100 case where the routing stage had gathered several addresses to send to
1101 a host before calling the transport for the first, we previously failed
1102 to close down TLS in the old transport process before passing the TCP
1103 connection to the new process. The new one sent a STARTTLS command
1104 which naturally failed, giving a failed delivery and bloating the retry
1105 database. Investigation and fix prototype from Wolfgang Breyha.
1107 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1108 false-negatives in the check that the sender had not preempted a response
1109 or prompt from Exim (running as a server), due to that code's lack of
1110 awareness of the SMTP input buffering.
1112 PP/04 Add commandline_checks_require_admin option.
1113 Exim drops privileges sanely, various checks such as -be aren't a
1114 security problem, as long as you trust local users with access to their
1115 own account. When invoked by services which pass untrusted data to
1116 Exim, this might be an issue. Set this option in main configuration
1117 AND make fixes to the calling application, such as using `--` to stop
1120 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1121 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1122 responses to those, into a single TLS record each way (this usually means
1123 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1126 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1127 affects you only if you're dancing at the edge of the param size limits.
1128 If you are, and this message makes sense to you, then: raise the
1129 configured limit or use OpenSSL 1.1. Nothing we can do for older
1132 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1133 cmdline element to get the documented default one. Previously it was
1136 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1137 are unsuable for non-message-specific reasons - eg. connection timeout,
1138 banner-time rejection.
1140 JH/16 Fix logging of delivery remote port, when specified by router, under
1143 PP/06 Repair manualroute's ability to take options in any order, even if one
1144 is the name of a transport.
1147 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1149 JH/17 Change the list-building routines interface to use the expanding-string
1150 triplet model, for better allocation and copying behaviour.
1152 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1153 Previously it was constructed the first time a possibly-matching string
1154 was met in the configuration file input during startup; now it is done
1157 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1158 compatible one, to avoid the (poorly documented) possibility of a config
1159 file in the working directory redirecting the DB files, possibly correpting
1160 some existing file. CVE-2017-10140 assigned for BDB.
1162 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1163 cache-hot. Previously, although the result was properly cached, the
1164 initial verify call returned a defer.
1166 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1167 the main verify for receipient in uncached-mode.
1169 JH/22 Retire historical build files to an "unsupported" subdir. These are
1170 defined as "ones for which we have no current evidence of testing".
1172 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1173 if present. Previously it was ignored.
1175 JH/24 Start using specified-initialisers in C structure init coding. This is
1176 a C99 feature (it's 2017, so now considered safe).
1178 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1179 if was a fixed-sized field and bitmask ops via macros; it is now more
1182 PP/07 GitHub PR 56: Apply MariaDB build fix.
1183 Patch provided by Jaroslav Škarvada.
1185 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1186 during Coverity cleanups [4.87 JH/47]
1187 Diagnosis and fix provided by Michael Fischer v. Mollard.
1189 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1190 the right size to place the terminating semicolon on its own folded
1191 line, the header hash was calculated to an incorrect value thanks to
1192 the (relaxed) space the fold became.
1194 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1195 and confused the parent.
1197 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1198 which could crash as a result. This could lead to undeliverable messages.
1200 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1203 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1204 matters on fast-turnover and PID-randomising systems, which were getting
1205 out-of-order delivery.
1207 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1208 a possibly-overlapping copy. The symptom was that "Remote host closed
1209 connection in response to HELO" was logged instead of the actual 4xx
1212 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1213 Previously only that bufferd was discarded, resulting in SYMTP command
1216 JH/32 DKIM: when a message has multiple signatures matching an identity given
1217 in dkim_verify_signers, run the dkim acl once for each. Previously only
1218 one run was done. Bug 2189.
1220 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1221 "panic the current process" to "deliberately defer". The panic log is
1222 still written with the problem list name; the mail and reject logs now
1223 get a temp-reject line for the message that was being handled, saying
1224 something like "domains check lookup or other defer". The SMTP 451
1225 message is still "Temporary local problem".
1227 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1228 A crafted sequence of BDAT commands could result in in-use memory beeing
1229 freed. CVE-2017-16943.
1231 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1232 from SMTP input. Previously it was always done; now only done for DATA
1233 and not BDAT commands. CVE-2017-16944.
1235 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1236 to the message (such as an overlong header line). Previously this was
1237 not done and we did not exit BDAT mode. Followon from the previous item
1238 though a different problem.
1244 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1245 than -2003 did; needs libidn2 in addition to libidn.
1247 JH/02 The path option on a pipe transport is now expanded before use.
1249 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1250 Patch provided by "Björn", documentation fix added too.
1252 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1253 missing a wire-to-host endian conversion.
1255 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1256 close after a BDAT command line could be taken as a following command,
1257 giving a synch failure. Fix by only checking for synch immediately
1258 before acknowledging the chunk.
1260 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1261 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1262 macro. Patches provided by Josh Soref.
1264 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1265 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1266 by IANA as a service extension. However, John Klensin suggests that we
1269 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1270 the dkim code may be unix-mode line endings rather than smtp wire-format
1271 CRLF, so prepend a CR to any bare LF.
1273 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1274 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1276 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1277 the first were themselves being wrongly included in the feed into dkim
1278 processing; with most chunk sizes in use this resulted in an incorrect
1279 body hash calculated value.
1281 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1282 DKIM signature block, for verification. Although advised against by
1283 standards it is specifically not ruled illegal.
1285 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1287 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1288 missing a body hash (the bh= tag).
1290 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1291 It seems that HAProxy sends the Proxy Protocol information in clear and
1292 only then does a TLS startup, so do the same.
1294 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1295 TCP connections (such as for Spamd) unless the daemon successfully set
1296 Fast Open mode on its listening sockets. This fixes breakage seen on
1297 too-old kernels or those not configured for Fast Open, at the cost of
1298 requiring both directions being enabled for TFO, and TFO never being used
1299 by non-daemon-related Exim processes.
1301 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1302 endings, at least on the first header line. Try to canonify any that get
1303 past that check, despite the cost.
1305 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1306 now limited to an arbitrary five deep, while parsing addresses with the
1307 strip_excess_angle_brackets option enabled.
1309 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1310 instead leave the unprompted TLS handshake in socket buffer for the
1311 TLS library to consume.
1313 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1315 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1317 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1318 region in one operation, for speed, and this leaves assigned pointers
1319 dangling. Add checks run only under the testsuite which checks all
1320 variables at a store-reset and panics on a dangling pointer; add code
1321 explicitly nulling out all the variables discovered. Fixes one known
1322 bug: a transport crash, where a dangling pointer for $sending_ip_address
1323 originally assigned in a verify callout, is re-used.
1325 PP/06 Drop '.' from @INC in various Perl scripts.
1327 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1329 PP/08 Reduce a number of compilation warnings under clang; building with
1330 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1331 should be warning-free.
1333 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1335 HS/01 Fix portability problems introduced by PP/08 for platforms where
1336 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1338 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1339 chunk. This allows us to accept broken chunked messages. We need a more
1340 general solution here.
1342 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1343 already-broken messages in the queue.
1345 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1347 JH/19 Fix reference counting bug in routing-generated-address tracking.
1353 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1354 supports it and a size is available (ie. the sending peer gave us one).
1356 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1357 years of being deprecated). The replacements are the ACLs
1358 acl_smtp_mime and acl_not_smtp_mime.
1360 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1361 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1362 or even an in-clear connection were permitted. Now, if the host lookup
1363 was dnssec and dane was requested then the host is only used if the
1364 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1365 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1366 if one fails this test.
1367 This means that a poorly-configured remote DNS will make it incommunicado;
1368 but it protects against a DNS-interception attack on it.
1370 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1371 non-noisy when a race steals the message being considered.
1373 JH/05 If main configuration option tls_certificate is unset, generate a
1374 self-signed certificate for inbound TLS connections.
1376 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1377 in rewrites and routers.
1379 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1380 and logged a warning sing 4.83; now they are a configuration file error.
1382 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1383 (lacking @domain). Apply the same qualification processing as RCPT.
1385 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1387 JH/10 Support ${sha256:} applied to a string (as well as the previous
1390 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1391 a cutthrough deliver is pending, as we always want to make a connection.
1392 This also avoids re-routing the message when later placing the cutthrough
1393 connection after a verify cache hit.
1394 Do not update it with the verify result either.
1396 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1397 when routing results in more than one destination address.
1399 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1400 signing (which inhibits the cutthrough capability). Previously only
1401 the presence of an option was tested; now an expansion evaluating as
1402 empty is permissible (obviously it should depend only on data available
1403 when the cutthrough connection is made).
1405 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1406 the relevant preceding SMTP command did not note the pipelining mode.
1408 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1409 Previously they were not counted.
1411 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1412 as one having no matching records. Previously we deferred the message
1413 that needed the lookup.
1415 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1416 distinguished as "(=".
1418 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1419 for missing MX records. Previously it only worked for missing A records.
1421 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1423 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1424 after the data-go-ahead and data-ack. Patch from Jason Betts.
1426 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1427 even for a "none" policy. Patch from Tony Meyer.
1429 JH/22 Fix continued use of a connection for further deliveries. If a port was
1430 specified by a router, it must also match for the delivery to be
1433 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1434 When one of the recipients of a message was unsuitable for the connection
1435 (has no matching addresses), we lost track of needing to mark it
1436 deferred. As a result mail would be lost.
1438 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1440 JH/25 Decoding ACL controls is now done using a binary search; the source code
1441 takes up less space and should be simpler to maintain. Merge the ACL
1442 condition decode tables also, with similar effect.
1444 JH/26 Fix problem with one_time used on a redirect router which returned the
1445 parent address unchanged. A retry would see the parent address marked as
1446 delivered, so not attempt the (identical) child. As a result mail would
1449 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1450 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1451 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1454 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1455 Main config options check_{log,spool}_{inodes,space} are now
1456 100 inodes, 10MB unless set otherwise in the configuration.
1458 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1459 Previously it only applied to the main-section connection policy
1462 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1464 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1465 by me. Added RFC7919 DH primes as an alternative.
1467 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1468 are not in the system include path.
1470 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1471 GnuTLS, when a session startup failed (eg because the client disconnected)
1472 Exim did stdio operations after fclose. This was exposed by a recent
1473 change which nulled out the file handle after the fclose.
1475 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1476 signed directly by the cert-signing cert, rather than an intermediate
1477 OCSP-signing cert. This is the model used by LetsEncrypt.
1479 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1481 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1482 an incoming connection.
1484 HS/02 Bug 1802: Do not half-close the connection after sending a request
1487 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1488 fallback to "prime256v1".
1490 JH/34 SECURITY: Use proper copy of DATA command in error message.
1491 Could leak key material. Remotely exploitable. CVE-2016-9963.
1497 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1498 and 3.4.4 - once the server is enabled to respond to an OCSP request
1499 it does even when not requested, resulting in a stapling non-aware
1500 client dropping the TLS connection.
1502 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1503 support variable-length bit vectors. No functional change.
1505 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1506 The I= interface field on outgoing lines is now after the H= remote
1507 host field, same as incoming lines. There is a separate
1508 outgoing_interface log selector which allows you to disable the
1511 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1512 If not running log_selector +smtp_connection the mainlog would be held
1513 open indefinitely after a "too many connections" event, including to a
1514 deleted file after a log rotate. Leave the per net connection logging
1515 leaving it open for efficiency as that will be quickly detected by the
1516 check on the next write.
1518 HS/01 Bug 1671: Fix post transport crash.
1519 Processing the wait-<transport> messages could crash the delivery
1520 process if the message IDs didn't exist for some reason. When
1521 using 'split_spool_directory=yes' the construction of the spool
1522 file name failed already, exposing the same netto behaviour.
1524 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1525 mime_regex ACL conditions.
1527 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1528 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1529 to HELO, local diagnostic string.
1531 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1532 log line to debug. Even when configured with a tls authenticator many
1533 client connections are expected to not authenticate in this way, so
1534 an authenticate fail is not an error.
1536 HS/02 Add the Exim version string to the process info. This way exiwhat
1537 gives some more detail about the running daemon.
1539 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1540 matter for fast-change records such as DNSBLs.
1542 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1543 retry record, even if constant. There may be multiple transports with
1544 different interface settings and the retry behaviour needs to be kept
1547 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1549 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1551 JH/10 Bug 840: fix log_defer_output option of pipe transport
1553 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1554 hosts_randomize. This matters a lot when combined with mua_wrapper.
1556 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1557 ${quote_pgsql:<string>} operator.
1559 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1561 JH/14 Change header file name for freeradius-client. Relevant if compiling
1562 with Radius support; from the Gentoo tree and checked under Fedora.
1564 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1566 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1567 option transport_filter, meaning no filtering.
1569 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1571 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1572 defaults to "*" (all hosts). The variable is now available when not built
1573 with TLS, default unset, mainly to enable keeping the testsuite sane.
1574 If a server certificate is not supplied (via tls_certificate) an error is
1575 logged, and clients will find TLS connections fail on startup. Presumably
1576 they will retry in-clear.
1577 Packagers of Exim are strongly encouraged to create a server certificate
1578 at installation time.
1580 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1581 with the $config_file variable.
1583 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1584 in transport context, after the attempt, and per-recipient. The latter type
1585 is per host attempted. The event data is the error message, and the errno
1586 information encodes the lookup type (A vs. MX) used for the (first) host,
1587 and the trailing two digits of the smtp 4xx response.
1589 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1590 to write to mainlog (or rejectlog, paniclog) in the window between file
1591 creation and permissions/ownership being changed. Particularly affects
1592 installations where exicyclog is run as root, rather than exim user;
1593 result is that the running daemon panics and dies.
1595 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1597 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1598 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1599 "pri" and "weight". Note that the previous implicit priority given by the
1600 list order is no longer honoured.
1602 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1603 for DKIM processing.
1605 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1606 by defining SUPPORT_SOCKS.
1608 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1609 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1610 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1611 variables are renamed to proxy_{local,external}_{address,port}.
1613 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1614 for a build by defining SUPPORT_I18N
1616 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1617 of the query string, and make ${quote_redis:} do that quoting.
1619 JH/29 Move Events support from Experimental to mainline, enabled by default
1620 and removable for a build by defining DISABLE_EVENT.
1622 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1624 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1625 cached by the daemon.
1627 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1628 by defining LOOKUP_REDIS. The libhiredis library is required.
1630 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1631 keys are given for lookup.
1633 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1634 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1635 only supported when built with TLS support. The PolarSSL SHA routines
1636 are still used when the TLS library is too old for convenient support.
1638 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1639 openssl_options), for security. OpenSSL forces this from version 1.1.0
1640 server-side so match that on older versions.
1642 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1643 allocation for $value could be released as the expansion processing
1644 concluded, but leaving the global pointer active for it.
1646 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1647 and to use the domains and local_parts ACL conditions.
1649 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1650 incorrectly not doubled on cutthrough transmission, hence seen as a
1651 body-termination at the receiving system - resulting in truncated mails.
1652 Commonly the sender saw a TCP-level error, and retransmitted the message
1653 via the normal store-and-forward channel. This could result in duplicates
1654 received - but deduplicating mailstores were liable to retain only the
1655 initial truncated version.
1657 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1659 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1661 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1662 we're in there, support oversigning also; bug 1309.
1664 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1666 HS/04 Add support for keep_environment and add_environment options.
1668 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1669 either intentional arithmetic overflow during PRNG, or testing config-
1672 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1673 delivery resulted in actual delivery. Cancel cutthrough before DATA
1676 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1677 rejects a recipient: pass the reject to the originator.
1679 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1680 Many were false-positives and ignorable, but it's worth fixing the
1683 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1684 for the new environment-manipulation done at startup. Move the routines
1685 from being local to tls.c to being global via the os.c file.
1687 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1688 an extract embedded as result-arg for a map, the first arg for extract
1689 is unavailable so we cannot tell if this is a numbered or keyed
1690 extraction. Accept either.
1696 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1699 JH/02 The smtp transport option "multi_domain" is now expanded.
1701 JH/03 The smtp transport now requests PRDR by default, if the server offers
1704 JH/04 Certificate name checking on server certificates, when exim is a client,
1705 is now done by default. The transport option tls_verify_cert_hostnames
1706 can be used to disable this per-host. The build option
1707 EXPERIMENTAL_CERTNAMES is withdrawn.
1709 JH/05 The value of the tls_verify_certificates smtp transport and main options
1710 default to the word "system" to access the system default CA bundle.
1711 For GnuTLS, only version 3.0.20 or later.
1713 JH/06 Verification of the server certificate for a TLS connection is now tried
1714 (but not required) by default. The verification status is now logged by
1715 default, for both outbound TLS and client-certificate supplying inbound
1718 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1721 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1722 Status Notification (bounce) messages are now MIME format per RFC 3464.
1723 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1724 under the control of the dsn_advertise_hosts option, and routers may
1725 have a dsn_lasthop option.
1727 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1728 default, modifiable by a malware= option. The list separator for
1729 the options can now be changed in the usual way. Bug 68.
1731 JH/10 The smtp_receive_timeout main option is now expanded before use.
1733 JH/11 The incoming_interface log option now also enables logging of the
1734 local interface on delivery outgoing connections.
1736 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1737 if the interface and destination host and port all match.
1739 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1742 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1743 Patch from Andrew Lewis.
1745 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1746 now supports optional time-restrictions, weighting, and priority
1747 modifiers per server. Patch originally by <rommer@active.by>.
1749 JH/16 The spamd_address main option now supports a mixed list of local
1750 and remote servers. Remote servers can be IPv6 addresses, and
1751 specify a port-range.
1753 JH/17 Bug 68: The spamd_address main option now supports an optional
1754 timeout value per server.
1756 JH/18 Bug 1581: Router and transport options headers_add/remove can
1757 now have the list separator specified.
1759 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1762 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1765 JH/21 Support for the A6 type of dns record is withdrawn.
1767 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1768 rather than the verbs used.
1770 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1771 from 255 to 1024 chars.
1773 JH/24 Verification callouts now attempt to use TLS by default.
1775 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1776 are generic router options now. The defaults didn't change.
1778 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1779 Original patch from Alexander Shikoff, worked over by JH.
1781 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1782 files if ZCAT_COMMAND is not executable.
1784 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1786 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1788 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1789 Normally benign, it bites when the pair was led to by a CNAME;
1790 modern usage is to not canonicalize the domain to a CNAME target
1791 (and we were inconsistent anyway for A-only vs AAAA+A).
1793 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1795 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1796 when evaluating $sender_host_dnssec.
1798 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1799 $sender_helo_dnssec variable.
1801 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1803 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1805 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1807 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1808 documented as working, but never had. Support all but $spam_report.
1810 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1811 added for tls authenticator.
1813 HS/03 Add perl_taintmode main config option
1819 TL/01 When running the test suite, the README says that variables such as
1820 no_msglog_check are global and can be placed anywhere in a specific
1821 test's script, however it was observed that placement needed to be near
1822 the beginning for it to behave that way. Changed the runtest perl
1823 script to read through the entire script once to detect and set these
1824 variables, reset to the beginning of the script, and then run through
1825 the script parsing/test process like normal.
1827 TL/02 The BSD's have an arc4random API. One of the functions to induce
1828 adding randomness was arc4random_stir(), but it has been removed in
1829 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1830 function when detected.
1832 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1833 cause callback expansion.
1835 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1836 syntax errors in an expansion can be treated as a string instead of
1837 logging or causing an error, due to the internal use of bool_lax
1838 instead of bool when processing it.
1840 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1841 server certificates when making smtp deliveries.
1843 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1845 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1847 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1849 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1850 Merged patch from Sebastian Wiedenroth.
1852 JH/05 Fix results-pipe from transport process. Several recipients, combined
1853 with certificate use, exposed issues where response data items split
1854 over buffer boundaries were not parsed properly. This eventually
1855 resulted in duplicates being sent. This issue only became common enough
1856 to notice due to the introduction of connection certificate information,
1857 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1859 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1860 size buffer was used, resulting in syntax errors when an expansion
1863 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1864 version 3.3.6 or later.
1866 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1867 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1868 both become "event_action", the variables become $event_name, $event_data
1869 and $event_defer_errno. There is a new variable $verify_mode, usable in
1870 routers, transports and related events. The tls:cert event is now also
1871 raised for inbound connections, if the main configuration event_action
1874 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1875 early OCSP support, but no stapling (appears to be less than 1.0.0).
1877 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1878 server certificate names available under the smtp transport option
1879 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1882 JH/10 Time-related extraction expansions from certificates now use the main
1883 option "timezone" setting for output formatting, and are consistent
1884 between OpenSSL and GnuTLS compilations. Bug 1541.
1886 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1887 encoded parameter in the incoming message. Bug 1558.
1889 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1890 include certificate info, eximon was claiming there were spoolfile
1893 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1895 JH/14 Log delivery-related information more consistently, using the sequence
1896 "H=<name> [<ip>]" wherever possible.
1898 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1899 are problematic for Debian distribution, omit them from the release
1902 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1904 JH/16 Fix string representation of time values on 64bit time_t architectures.
1907 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1908 output list separator was used.
1913 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1914 checkers that were complaining about end of non-void function with no
1917 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1918 This was a regression introduced in 4.83 by another bugfix.
1920 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1922 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1923 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1929 TF/01 Correctly close the server side of TLS when forking for delivery.
1931 When a message was received over SMTP with TLS, Exim failed to clear up
1932 the incoming connection properly after forking off the child process to
1933 deliver the message. In some situations the subsequent outgoing
1934 delivery connection happened to have the same fd number as the incoming
1935 connection previously had. Exim would try to use TLS and fail, logging
1936 a "Bad file descriptor" error.
1938 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1939 utilities have not been installed.
1941 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1942 temporary space as the ACL may create new global variables.
1944 TL/01 LDAP support uses per connection or global context settings, depending
1945 upon the detected version of the libraries at build time.
1947 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1948 to extract and use the src ip:port in logging and expansions as if it
1949 were a direct connection from the outside internet. PPv2 support was
1950 updated based on HAProxy spec change in May 2014.
1952 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1954 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1955 Properly escape header and check for NULL return.
1957 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1960 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1962 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1963 characters in header names, implemented as a verify condition.
1964 Contributed by Michael Fischer v. Mollard.
1966 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1967 results permerror and temperror. Previous values are deprecated but
1968 still accepted. In a future release, err_perm and err_temp will be
1969 completely removed, which will be a backward incompatibility if the
1970 ACL tests for either of these two old results. Patch contributed by
1971 user bes-internal on the mailing list.
1973 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1975 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1976 selectors, in both main and reject logs.
1978 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1981 JH/07 Add malware type "sock" for talking to simple daemon.
1983 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1985 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1986 routers/transports under cutthrough routing.
1988 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1989 numbers. Touch up "bool" conditional to keep the same definition.
1991 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1993 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1995 JH/12 Expand items in router/transport headers_add or headers_remove lists
1996 individually rather than the list as a whole. Bug 1452.
1998 Required for reasonable handling of multiple headers_ options when
1999 they may be empty; requires that headers_remove items with embedded
2000 colons must have them doubled (or the list-separator changed).
2002 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2003 view the policy declared in the DMARC record. Currently, $dmarc_status
2004 is a combined value of both the record presence and the result of the
2007 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2009 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2010 dnslookup router and the smtp transport (applying to the forward
2013 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2014 of ldap servers used for a specific lookup. Patch provided by Heiko
2017 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2018 New variable $lookup_dnssec_authenticated for observability.
2020 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2021 Patch submitted by Lars Timman.
2023 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2025 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2026 Requires trusted mode and valid format message id, aborts otherwise.
2027 Patch contributed by Heiko Schlichting.
2029 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2030 certextract with support for various fields. Bug 1358.
2032 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2033 is requested by default, modifiable by smtp transport option
2036 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2037 operate on certificate variables to give certificate fingerprints
2038 Also new ${sha256:cert_variable}.
2040 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2042 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2043 Christian Aistleitner.
2045 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2047 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2048 file. Patch from Wolfgang Breyha.
2050 JH/25 Expand the coverage of the delivery $host and $host_address to
2051 client authenticators run in verify callout. Bug 1476.
2053 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2054 align with daemon_smtp_ports. Bug 72.
2056 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2057 support and error reporting did not work properly.
2059 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2060 and is readable. Patch from Andrew Colin Kissa.
2062 TL/14 Enhance documentation of ${run expansion and how it parses the
2063 commandline after expansion, particularly in the case when an
2064 unquoted variable expansion results in an empty value.
2066 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2068 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2069 by renaming away from it. Addresses GH issue 15, reported by
2072 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2074 TL/15 SECURITY: prevent double expansion in math comparison functions
2075 (can expand unsanitized data). Not remotely exploitable.
2082 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2084 PP/02 Make -n do something, by making it not do something.
2085 When combined with -bP, the name of an option is not output.
2087 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2090 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2091 $sender_host_name and config options to manage this, and basic check
2094 PP/05 DSCP support for outbound connections and control modifier for inbound.
2096 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2097 (Only plugin which currently uses this is kerberos4, which nobody should
2098 be using, but we should make it available and other future plugins might
2099 conceivably use it, even though it would break NAT; stuff *should* be
2100 using channel bindings instead).
2102 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2103 name; added for Sendmail compatibility; requires admin caller.
2104 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2105 just ignore it); requires trusted caller.
2106 Also parse but ignore: -Ac -Am -X<logfile>
2109 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2111 TL/02 Add +smtp_confirmation as a default logging option.
2113 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2114 Patch by Magnus Holmgren from 2007-02-20.
2116 TL/04 Bugzilla 1281 - Spec typo.
2117 Bugzilla 1283 - Spec typo.
2118 Bugzilla 1290 - Spec grammar fixes.
2120 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2122 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2124 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2125 mailing list by by Dmitry Isaikin.
2127 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2129 JH/02 Support "G" suffix to numbers in ${if comparisons.
2131 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2133 NM/01 Bugzilla 1197 - Spec typo
2134 Bugzilla 1196 - Spec examples corrections
2136 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2138 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2139 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2142 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2143 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2145 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2146 "acl {{name}{arg}...}", and optional args on acl condition
2149 JH/05 Permit multiple router/transport headers_add/remove lines.
2151 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2153 JH/07 Avoid using a waiting database for a single-message-only transport.
2154 Performance patch from Paul Fisher. Bugzilla 1262.
2156 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2159 JH/09 Add $headers_added variable, with content from use of ACL modifier
2160 add_header (but not yet added to the message). Bugzilla 199.
2162 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2163 Pulled from Bugzilla 817 by Wolfgang Breyha.
2165 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2167 (nb: this is the same fix as in Exim 4.80.1)
2169 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2172 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2175 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2177 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2178 advertise SMTP AUTH mechanism to us, instead of a generic
2179 protocol violation error. Also, make Exim more robust to bad
2180 data from the Dovecot auth socket.
2182 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2184 When a queue runner is handling a message, Exim first routes the
2185 recipient addresses, during which it prunes them based on the retry
2186 hints database. After that it attempts to deliver the message to
2187 any remaining recipients. It then updates the hints database using
2190 So if a recipient address works intermittently, it can get repeatedly
2191 deferred at routing time. The retry hints record remains fresh so the
2192 address never reaches the final cutoff time.
2194 This is a fairly common occurrence when a user is bumping up against
2195 their storage quota. Exim had some logic in its local delivery code
2196 to deal with this. However it did not apply to per-recipient defers
2197 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2199 This change adds a proper retry rule check during routing so that the
2200 final cutoff time is checked against the message's age. We only do
2201 this check if there is an address retry record and there is not a
2202 domain retry record; this implies that previous attempts to handle
2203 the address had the retry_use_local_parts option turned on. We use
2204 this as an approximation for the destination being like a local
2205 delivery, as in LMTP.
2207 I suspect this new check makes the old local delivery cutoff check
2208 redundant, but I have not verified this so I left the code in place.
2210 TF/02 Correct gecos expansion when From: is a prefix of the username.
2212 Test 0254 submits a message to Exim with the header
2216 When I ran the test suite under the user fanf2, Exim expanded
2217 the header to contain my full name, whereas it should have added
2218 a Resent-Sender: header. It erroneously treats any prefix of the
2219 username as equal to the username.
2221 This change corrects that bug.
2223 GF/01 DCC debug and logging tidyup
2224 Error conditions log to paniclog rather than rejectlog.
2225 Debug lines prefixed by "DCC: " to remove any ambiguity.
2227 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2229 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2230 Bug spotted by Jeremy Harris; was flawed since initial commit.
2231 Would have resulted in OCSP responses post-SNI triggering an Exim
2232 NULL dereference and crash.
2234 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2236 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2237 Bug detection, analysis and fix by Samuel Thibault.
2238 Bugzilla 1331, Debian bug #698092.
2240 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2242 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2243 Server implementation by Todd Lyons, client by JH.
2244 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2245 config variable "prdr_enable" controls whether the server
2246 advertises the facility. If the client requests PRDR a new
2247 acl_data_smtp_prdr ACL is called once for each recipient, after
2248 the body content is received and before the acl_smtp_data ACL.
2249 The client is controlled by both of: a hosts_try_prdr option
2250 on the smtp transport, and the server advertisement.
2251 Default client logging of deliveries and rejections involving
2252 PRDR are flagged with the string "PRDR".
2254 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2255 fclose(). Diagnosis by Todd Lyons.
2257 PP/17 Update configure.default to handle IPv6 localhost better.
2258 Patch by Alain Williams (plus minor tweaks).
2261 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2262 This is now consistent with GnuTLS, and is now documented: the
2263 previous undocumented portable approach to treating the option as
2264 unset was to force an expansion failure. That still works, and
2265 an empty string is now equivalent.
2267 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2268 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2269 not performing validation itself.
2271 PP/20 Added force_command boolean option to pipe transport.
2272 Patch from Nick Koston, of cPanel Inc.
2274 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2277 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2279 PP/21 Fix eximon continuous updating with timestamped log-files.
2280 Broken in a format-string cleanup in 4.80, missed when I repaired the
2281 other false fix of the same issue.
2282 Report and fix from Heiko Schlichting.
2285 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2286 Report from Prashanth Katuri.
2288 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2289 It's SecureTransport, so affects any MacOS clients which use the
2290 system-integrated TLS libraries, including email clients.
2292 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2293 using a MIME ACL for non-SMTP local injection.
2294 Report and assistance in diagnosis by Warren Baker.
2296 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2298 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2300 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2301 last id that failed so it may be referenced in subsequent ACL's.
2303 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2306 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2307 ldap library initialization, allowing self-signed CA's to be
2308 used. Also properly sets require_cert option later in code by
2309 using NULL (global ldap config) instead of ldap handle (per
2310 session). Bug diagnosis and testing by alxgomz.
2312 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2313 the src/util/ subdirectory.
2315 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2316 renamed to Transport Post Delivery Action by Jeremy Harris, as
2319 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2320 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2321 redis_servers = needs to be configured which will be used by the redis
2322 lookup. Patch from Warren Baker, of The Packet Hub.
2324 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2326 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2327 hostname or reverse DNS when processing a host list. Used suggestions
2328 from multiple comments on this bug.
2330 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2332 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2333 Missed a few lines, added it to make the runtest require no keyboard
2336 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2337 contains upper case chars. Make router use caseful_local_part.
2339 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2340 support when GnuTLS has been built with p11-kit.
2346 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2348 This, or similar/improved, will also be change PP/11 of 4.82.
2354 PP/01 Handle short writes when writing local log-files.
2355 In practice, only affects FreeBSD (8 onwards).
2356 Bugzilla 1053, with thanks to Dmitry Isaikin.
2358 NM/01 Bugzilla 949 - Documentation tweak
2360 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2363 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2365 PP/02 Implemented gsasl authenticator.
2367 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2369 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2370 `pkg-config foo` for cflags/libs.
2372 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2373 with rest of GSASL and with heimdal_gssapi.
2375 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2376 `pkg-config foo` for cflags/libs for the TLS implementation.
2378 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2379 properties get this fed in as external SSF. A number of robustness
2380 and debugging improvements to the cyrus_sasl authenticator.
2382 PP/08 cyrus_sasl server now expands the server_realm option.
2384 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2385 Patch by Jeremy Harris.
2387 PP/10 Added dbmjz lookup type.
2389 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2391 PP/12 MAIL args handles TAB as well as SP, for better interop with
2392 non-compliant senders.
2393 Analysis and variant patch by Todd Lyons.
2395 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2396 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2397 Patch from Dirk Mueller <dmueller@suse.com>
2399 PP/13 tls_peerdn now print-escaped for spool files.
2400 Observed some $tls_peerdn in wild which contained \n, which resulted
2401 in spool file corruption.
2403 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2404 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2405 or write after TLS renegotiation, which otherwise led to messages
2408 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2409 as a tracking header (ie: a signed header comes before the signature).
2410 Patch from Wolfgang Breyha.
2412 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2413 comma-sep list; embedded commas doubled.
2415 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2417 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2419 Report and patch from Dmitry Banschikov.
2421 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2422 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2423 failures. We appear to now support TLS1.1+ with Exim.
2425 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2426 lets Exim select keys and certificates based upon TLS SNI from client.
2427 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2428 before an outbound SMTP session. New log_selector, +tls_sni.
2430 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2431 NULL dereference. Report and patch from Alun Jones.
2433 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2434 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2435 Not seeing resolver debug output on NetBSD, but suspect this is a
2436 resolver implementation change.
2438 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2439 Left warnings. Added "eximon gdb" invocation mode.
2441 PP/21 Defaulting "accept_8bitmime" to true, not false.
2443 PP/22 Added -bw for inetd wait mode support.
2445 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2446 locate the relevant includes and libraries. Made this the default.
2448 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2449 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2451 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2452 This may cause build issues on older platforms.
2454 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2455 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2456 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2457 Added SNI support via GnuTLS too.
2458 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2460 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2462 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2463 Applied second patch from Janne, implementing suggestion to default
2464 multiple-strings-in-record handling to match SPF spec.
2466 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2468 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2469 read-only, out of scope).
2470 Patch from Wolfgang Breyha, report from Stuart Northfield.
2472 PP/29 Fix three issues highlighted by clang analyser static analysis.
2473 Only crash-plausible issue would require the Cambridge-specific
2474 iplookup router and a misconfiguration.
2475 Report from Marcin Mirosław.
2477 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2479 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2480 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2481 As part of this, removing so much warning spew let me fix some minor
2482 real issues in debug logging.
2484 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2485 assignment on my part. Fixed.
2487 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2488 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2489 Janne Snabb (who went above and beyond: thank you).
2491 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2492 string otherwise requires a connection and a bunch more work and it's
2493 relatively easy to get wrong. Should also expose TLS library linkage
2496 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2497 64-bit ${eval} (JH/03).
2499 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2500 GNU libc to support some of the 64-bit stuff, should not lead to
2501 conflicts. Defined before os.h is pulled in, so if a given platform
2502 needs to override this, it can.
2504 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2505 protection layer was required, which is not implemented.
2506 Bugzilla 1254, patch from Wolfgang Breyha.
2508 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2509 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2510 tls_dhparam take prime identifiers. Also unbreak combination of
2511 OpenSSL+DH_params+TLSSNI.
2513 PP/39 Disable SSLv2 by default in OpenSSL support.
2519 PP/01 Solaris build fix for Oracle's LDAP libraries.
2520 Bugzilla 1109, patch from Stephen Usher.
2522 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2524 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2527 TF/02 Fix a couple more cases where we did not log the error message
2528 when unlink() failed. See also change 4.74-TF/03.
2530 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2531 lock up or crash if it happened to be inside a call to libc when it
2532 got a SIGUSR1 from exiwhat.
2534 The SIGUSR1 handler appends the current process status to the process
2535 log which is later printed by exiwhat. It used to use the general
2536 purpose logging code to do this, but several functions it calls are
2537 not safe for signals.
2539 The new output code in the SIGUSR1 handler is specific to the process
2540 log, and simple enough that it's easy to inspect for signal safety.
2541 Removing some special cases also simplifies the general logging code.
2542 Removing the spurious timestamps from the process log simplifies
2545 TF/04 Improved ratelimit ACL condition.
2547 The /noupdate option has been deprecated in favour of /readonly which
2548 has clearer semantics. The /leaky, /strict, and /readonly update modes
2549 are mutually exclusive. The update mode is no longer included in the
2550 database key; it just determines when the database is updated. (This
2551 means that when you upgrade Exim will forget old rate measurements.)
2553 Exim now checks that the per_* options are used with an update mode that
2554 makes sense for the current ACL. For example, when Exim is processing a
2555 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2556 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2557 must specify per_mail/readonly. If you omit the update mode it defaults to
2558 /leaky where that makes sense (as before) or /readonly where required.
2560 The /noupdate option is now undocumented but still supported for
2561 backwards compatibility. It is equivalent to /readonly except that in
2562 ACLs where /readonly is required you may specify /leaky/noupdate or
2563 /strict/noupdate which are treated the same as /readonly.
2565 A useful new feature is the /count= option. This is a generalization
2566 of the per_byte option, so that you can measure the throughput of other
2567 aggregate values. For example, the per_byte option is now equivalent
2568 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2570 The per_rcpt option has been generalized using the /count= mechanism
2571 (though it's more complicated than the per_byte equivalence). When it is
2572 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2573 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2574 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2575 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2576 Note that using per_rcpt with a non-readonly update mode in more than
2577 one ACL will cause the recipients to be double-counted. (The per_mail
2578 and per_byte options don't have this problem.)
2580 The handling of very low rates has changed slightly. If the computed rate
2581 is less than the event's count (usually one) then this event is the first
2582 after a long gap. In this case the rate is set to the same as this event's
2583 count, so that the first message of a spam run is counted properly.
2585 The major new feature is a mechanism for counting the rate of unique
2586 events. The new per_addr option counts the number of different
2587 recipients that someone has sent messages to in the last time period. It
2588 behaves like per_rcpt if all the recipient addresses are different, but
2589 duplicate recipient addresses do not increase the measured rate. Like
2590 the /count= option this is a general mechanism, so the per_addr option
2591 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2592 example, measure the rate that a client uses different sender addresses
2593 with the options per_mail/unique=$sender_address. There are further
2594 details in the main documentation.
2596 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2598 TF/06 Removed a few PCRE remnants.
2600 TF/07 Automatically extract Exim's version number from tags in the git
2601 repository when doing development or release builds.
2603 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2604 Bugzilla 879. Patch from Paul Fisher.
2606 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2607 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2610 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2612 PP/05 Variable $av_failed, true if the AV scanner deferred.
2613 Bugzilla 1078. Patch from John Horne.
2615 PP/06 Stop make process more reliably on build failure.
2616 Bugzilla 1087. Patch from Heiko Schlittermann.
2618 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2619 Bugzilla 1089. Patch from Heiko Schlittermann.
2621 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2622 Bugzilla 1131. Patch from Holger Weiß.
2624 PP/09 Handle IPv6 addresses with SPF.
2625 Bugzilla 860. Patch from Wolfgang Breyha.
2627 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2629 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2632 PP/11 match_* no longer expand right-hand-side by default.
2633 New compile-time build option, EXPAND_LISTMATCH_RHS.
2634 New expansion conditions, "inlist", "inlisti".
2636 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2638 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2640 PP/14 fix log_write() format string regression from TF/03.
2641 Bugzilla 1152. Patch from Dmitry Isaikin.
2647 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2649 PP/02 Harmonised TLS library version reporting; only show if debugging.
2650 Layout now matches that introduced for other libraries in 4.74 PP/03.
2652 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2654 PP/04 New "dns_use_edns0" global option.
2656 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2659 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2660 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2662 TK/01 Updated PolarSSL code to 0.14.2.
2663 Bugzilla 1097. Patch from Andreas Metzler.
2665 PP/07 Catch divide-by-zero in ${eval:...}.
2666 Fixes bugzilla 1102.
2668 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2671 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2672 format-string attack -- SECURITY: remote arbitrary code execution.
2674 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2675 time unintentionally subject to list matching rules, letting the header
2676 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2677 arbitrary string expansion). This allowed for information disclosure.
2679 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2680 INT_MIN/-1 -- value coerced to INT_MAX.
2686 NM/01 Workaround for PCRE version dependency in version reporting
2689 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2690 This fixes portability to compilers other than gcc, notably
2691 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2693 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2694 makefiles for portability to HP-UX and POSIX correctness.
2696 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2697 Also via indented variable definition in the Makefile.
2698 (Debugging by Oliver Heesakkers).
2700 PP/02 Restore caching of spamd results with expanded spamd_address.
2701 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2703 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2704 Improves build reliability. Fix from: Frank Elsner
2706 NM/02 Fix wide character breakage in the rfc2047 coding
2707 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2709 NM/03 Allow underscore in dnslist lookups
2710 Fixes bug 1026. Patch from Graeme Fowler
2712 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2713 Code patches from Adam Ciarcinski of NetBSD.
2715 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2718 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2719 is logged, to avoid truncation. Patch from John Horne.
2721 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2722 Patch from Jakob Hirsch.
2724 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2725 SQL string expansion failure details.
2726 Patch from Andrey Oktyabrski.
2728 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2729 Patch from Simon Arlott.
2731 PP/09 New lookups functionality failed to compile on old gcc which rejects
2732 extern declarations in function scope.
2733 Patch from Oliver Fleischmann
2735 PP/10 Use sig_atomic_t for flags set from signal handlers.
2736 Check getgroups() return and improve debugging.
2737 Fixed developed for diagnosis in bug 927 (which turned out to be
2740 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2741 Patch from Mark Zealey.
2743 PP/12 Bugzilla 1056: Improved spamd server selection.
2744 Patch from Mark Zealey.
2746 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2747 Based on patch from Heiko Schlittermann.
2749 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2750 Patch from Uwe Doering, sign-off by Michael Haardt.
2752 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2753 variable declaration deep within a block. Bug and patch from
2756 PP/15 lookups-Makefile IRIX compatibility coercion.
2758 PP/16 Make DISABLE_DKIM build knob functional.
2760 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2761 Patch by Simon Arlott
2763 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2764 variable argument macros. Our copy now differs from upstream.
2770 TF/01 Failure to get a lock on a hints database can have serious
2771 consequences so log it to the panic log.
2773 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2774 controlled using the smtp_confirmation log selector.
2776 TF/03 Include the error message when we fail to unlink a spool file.
2778 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2779 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2780 for maintaining out-of-tree patches for some time.
2782 PP/01 Bugzilla 139: Documentation and portability issues.
2783 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2784 Handle per-OS dynamic-module compilation flags.
2786 PP/02 Let /dev/null have normal permissions.
2787 The 4.73 fixes were a little too stringent and complained about the
2788 permissions on /dev/null. Exempt it from some checks.
2789 Reported by Andreas M. Kirchwitz.
2791 PP/03 Report version information for many libraries, including
2792 Exim version information for dynamically loaded libraries. Created
2793 version.h, now support a version extension string for distributors
2794 who patch heavily. Dynamic module ABI change.
2796 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2797 privilege escalation vulnerability whereby the Exim run-time user
2798 can cause root to append content of the attacker's choosing to
2801 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2804 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2805 If dropping privileges for untrusted macros, we disabled normal logging
2806 on the basis that it would fail; for the Exim run-time user, this is not
2807 the case, and it resulted in successful deliveries going unlogged.
2808 Fixed. Reported by Andreas Metzler.
2814 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2815 only prepend for the Resent-* case. Fixes regression introduced in
2816 Exim 4.70 by NM/22 for Bugzilla 607.
2818 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2819 increasing numbers of administrators be bitten by this.
2821 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2823 PP/03 Bugzilla 994: added openssl_options main configuration option.
2825 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2827 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2829 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2831 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2832 without a peer certificate, leading to a segfault because of an
2833 assumption that peers always have certificates. Be a little more
2834 paranoid. Problem reported by Martin Tscholak.
2836 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2837 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2838 NB: ClamAV planning to remove STREAM in "middle of 2010".
2839 CL also introduces -bmalware, various -d+acl logging additions and
2840 more caution in buffer sizes.
2842 PP/09 Implemented reverse_ip expansion operator.
2844 PP/10 Bugzilla 937: provide a "debug" ACL control.
2846 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2848 PP/12 Bugzilla 973: Implement --version.
2850 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2852 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2854 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2856 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2857 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2858 ignore trailing whitespace.
2860 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2862 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2865 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2866 Notification from Dr Andrew Aitchison.
2868 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2869 ExtendedDetectionInfo response format.
2870 Notification from John Horne.
2872 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2875 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2876 XSL and documented dependency on system catalogs, with examples of how
2879 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2882 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2883 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2884 configuration file which is writeable by the Exim user or group.
2886 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2887 of configuration files to cover files specified with the -C option if
2888 they are going to be used with root privileges, not just the default
2891 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2892 option (effectively making it always true).
2894 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2895 files to be used while preserving root privileges.
2897 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2898 that rogue child processes cannot use them.
2900 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2901 run-time user, instead of root.
2903 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2904 Exim run-time user without dropping privileges.
2906 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2907 result string, instead of calling string_vformat() twice with the same
2910 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2911 for other users. Others should always drop root privileges if they use
2912 -C on the command line, even for a whitelisted configure file.
2914 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2916 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2922 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2923 $data_path, and $header_path variables; fixed documentation bugs and
2926 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2927 exipick to access non-standard spools, including the "frozen" queue
2930 NM/01 Bugzilla 965: Support mysql stored procedures.
2931 Patch from Alain Williams
2933 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2935 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2936 Patch from Andreas Metzler
2938 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2939 Patch from Kirill Miazine
2941 NM/05 Bugzilla 671: Added umask to procmail example.
2943 JJ/03 installed exipick 20100323.0, fixing doc bug
2945 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2946 directory. Notification and patch from Dan Rosenberg.
2948 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2950 TK/02 Improve log output when DKIM signing operation fails.
2952 MH/01 Treat the transport option dkim_domain as a colon separated
2953 list, not as a single string, and sign the message with each element,
2954 omitting multiple occurences of the same signer.
2956 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2957 Bugzilla 985, 986. Patch by Simon Arlott
2959 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2960 Patch by Simon Arlott
2962 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2963 MBX locking. Notification from Dan Rosenberg.
2969 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2971 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2973 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2975 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2977 NM/04 Bugzilla 915: Flag broken perl installation during build.
2983 TK/01 Added patch by Johannes Berg that expands the main option
2984 "spamd_address" if it starts with a dollar sign.
2986 TK/02 Write list of recipients to X-Envelope-Sender header when building
2987 the mbox-format spool file for content scanning (suggested by Jakob
2990 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2991 (http://www.dcc-servers.net/) support via dccifd. Activated by
2992 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2994 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2995 by Mark Daniel Reidel <mr@df.eu>.
2997 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2998 When building exim an external PCRE library is now needed -
2999 PCRE is a system library on the majority of modern systems.
3000 See entry on PCRE_LIBS in EDITME file.
3002 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3003 conversation. Added nologin parameter to request.
3004 Patch contributed by Kirill Miazine.
3006 TF/01 Do not log submission mode rewrites if they do not change the address.
3008 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3010 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3011 log files in place. Contributed by Roberto Lima.
3013 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3015 TF/03 Bugzilla 615: When checking the local_parts router precondition
3016 after a local_part_suffix or local_part_prefix option, Exim now
3017 does not use the address's named list lookup cache, since this
3018 contains cached lookups for the whole local part.
3020 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3021 Robert Millan. Documentation is in experimental-spec.txt.
3023 TF/04 Bugzilla 668: Fix parallel build (make -j).
3025 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3027 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3028 Patch provided by Jan Srzednicki.
3030 TF/05 Leading white space used to be stripped from $spam_report which
3031 wrecked the formatting. Now it is preserved.
3033 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3034 that they are available at delivery time.
3036 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3038 TF/08 TLS error reporting now respects the incoming_interface and
3039 incoming_port log selectors.
3041 TF/09 Produce a more useful error message if an SMTP transport's hosts
3042 setting expands to an empty string.
3044 NM/06 Bugzilla 744: EXPN did not work under TLS.
3045 Patch provided by Phil Pennock.
3047 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3048 Patch provided by Richard Godbee.
3050 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3051 acl_smtp_notquit, added index entry.
3053 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3054 Patch provided by Eugene Bujak.
3056 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3057 accept(). Patch provided by Maxim Dounin.
3059 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3060 Patch provided by Phil Pennock.
3062 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3064 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3065 Patch provided by Brad "anomie" Jorsch.
3067 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3068 Patch provided by Dean Brooks.
3070 TK/05 Add native DKIM support (does not depend on external libraries).
3072 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3073 Patch provided by Graeme Fowler.
3075 NM/16 Bugzilla 851: Documentation example syntax fix.
3077 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3079 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3082 NM/19 Bugzilla 745: TLS version reporting.
3083 Patch provided by Phil Pennock.
3085 NM/20 Bugzilla 167: bool: condition support.
3086 Patch provided by Phil Pennock.
3088 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3089 clients. Patch provided by Phil Pennock.
3091 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3092 Patch provided by Brad "anomie" Jorsch.
3094 NM/23 Bugzilla 687: Fix misparses in eximstats.
3095 Patch provided by Heiko Schlittermann.
3097 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3098 Patch provided by Heiko Schlittermann.
3100 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3101 plus update to original patch.
3103 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3105 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3106 Patch provided by David Brownlee.
3108 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3110 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3112 NM/30 Bugzilla 888: TLS documentation bugfixes.
3114 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3116 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3117 Unlike the original bugzilla I have changed all shell scripts in src tree.
3119 NM/33 Bugzilla 898: Transport filter timeout fix.
3120 Patch by Todd Rinaldo.
3122 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3123 Patch by Serge Demonchaux.
3125 NM/35 Bugzilla 39: Base64 decode bug fixes.
3126 Patch by Jakob Hirsch.
3128 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3130 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3132 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3134 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3140 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3141 ALT-N's libdkim that I have put here:
3142 http://duncanthrax.net/exim-experimental/
3144 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3145 were called 'true' and it seems that C99 defines that as a reserved
3146 keyword to be used with 'bool' variable types. That means you could
3147 not include C99-style headers which use bools without triggering
3148 build errors in sieve.c.
3150 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3151 as mailq or other aliases. Changed the --help handling significantly
3152 to do whats expected. exim_usage() emits usage/help information.
3154 SC/01 Added the -bylocaldomain option to eximstats.
3156 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3158 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3160 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3166 PH/01 Another patch from the Sieve maintainer.
3168 PH/02 When an IPv6 address is converted to a string for single-key lookup
3169 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3170 dots are used instead of colons so that keys in lsearch files need not
3171 contain colons. This was done some time before quoting was made available
3172 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3173 (notated using the quote facility) so as to distinguish them from IPv4
3174 keys. This meant that lookups for IP addresses in host lists did not work
3175 for iplsearch lookups.
3177 This has been fixed by arranging for IPv6 addresses to be expressed with
3178 colons if the lookup type is iplsearch. This is not incompatible, because
3179 previously such lookups could never work.
3181 The situation is now rather anomalous, since one *can* have colons in
3182 ordinary lsearch keys. However, making the change in all cases is
3183 incompatible and would probably break a number of configurations.
3185 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3188 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3189 conversion specification without a maximum field width, thereby enabling
3190 a rogue spamd server to cause a buffer overflow. While nobody in their
3191 right mind would setup Exim to query an untrusted spamd server, an
3192 attacker that gains access to a server running spamd could potentially
3193 exploit this vulnerability to run arbitrary code as the Exim user.
3195 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3196 $primary_hostname instead of what libspf2 thinks the hosts name is.
3198 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3199 a directory entry by the name of the lookup key. Previously, if a
3200 symlink pointed to a non-existing file or a file in a directory that
3201 Exim lacked permissions to read, a lookup for a key matching that
3202 symlink would fail. Now it is enough that a matching directory entry
3203 exists, symlink or not. (Bugzilla 503.)
3205 PH/03 The body_linecount and body_zerocount variables are now exported in the
3208 PH/04 Added the $dnslist_matched variable.
3210 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3211 This means they are set thereafter only if the connection becomes
3214 PH/06 Added the client_condition to authenticators so that some can be skipped
3215 by clients under certain conditions.
3217 PH/07 The error message for a badly-placed control=no_multiline_responses left
3218 "_responses" off the end of the name.
3220 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3222 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3223 (without spaces) instead of just copying the configuration text.
3225 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3227 PH/11 Added $max_received_linelength.
3229 PH/12 Added +ignore_defer and +include_defer to host lists.
3231 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3232 way in which PCRE > 7.0 is built.
3234 PH/14 Implemented queue_only_load_latch.
3236 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3237 MAIL command. The effect was to mangle the value on 64-bit systems.
3239 PH/16 Another patch from the Sieve maintainer.
3241 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3243 PH/18 If a system quota error occurred while trying to create the file for
3244 a maildir delivery, the message "Mailbox is full" was not appended to the
3245 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3246 only to a quota excession during the actual writing of the file.
3248 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3249 characters?) which causes problems in log lines. The DN values are now
3250 passed through string_printing() before being added to log lines.
3252 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3253 and InterBase are left for another time.)
3255 PH/21 Added message_body_newlines option.
3257 PH/22 Guard against possible overflow in moan_check_errorcopy().
3259 PH/23 POSIX allows open() to be a macro; guard against that.
3261 PH/24 If the recipient of an error message contained an @ in the local part
3262 (suitably quoted, of course), incorrect values were put in $domain and
3263 $local_part during the evaluation of errors_copy.
3269 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3270 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3273 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3274 issue a MAIL command.
3276 PH/02 In an ACL statement such as
3278 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3280 if a client was not listed at all, or was listed with a value other than
3281 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3282 the condition was not true (as it should be), so access was not denied.
3283 The bug was that the ! inversion was incorrectly passed on to the second
3284 item. This has been fixed.
3286 PH/03 Added additional dnslists conditions == and =& which are different from
3287 = and & when the dns lookup returns more than one IP address.
3289 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3290 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3292 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3293 FSYNC, which compiles an option called disable_fsync that allows for
3294 bypassing fsync(). The documentation is heavily laced with warnings.
3296 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3298 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3299 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3300 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3301 including adding "make clean"; (3) Added -fPIC when compiling the test
3302 dynamically loaded module, to get rid of a warning.
3304 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3305 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3306 = 0s. The bug is otherwise harmless.
3308 PH/07 There was a bug in the dovecot authenticator such that the value of
3309 $auth1 could be overwritten, and so not correctly preserved, after a
3310 successful authentication. This usually meant that the value preserved by
3311 the server_setid option was incorrect.
3313 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3315 PH/09 Installed PCRE release 7.0.
3317 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3318 run for batched SMTP input. It is now run at the start of every message
3319 in the batch. While fixing this I discovered that the process information
3320 (output by running exiwhat) was not always getting set for -bs and -bS
3321 input. This is fixed, and it now also says "batched" for BSMTP.
3323 PH/11 Added control=no_pipelining.
3325 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3326 patch, slightly modified), and move the expansion of helo_data till after
3327 the connection is made in the smtp transport (so it can use these
3330 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3332 PH/14 Added log_selector = +pid.
3334 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3336 PH/16 Add ${if forany and ${if forall.
3338 PH/17 Added dsn_from option to vary the From: line in DSNs.
3340 PH/18 Flush SMTP output before performing a callout, unless control =
3341 no_callout_flush is set.
3343 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3344 was true (the default) a successful delivery failed to delete the retry
3345 item, thus causing premature timeout of the address. The bug is now
3348 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3350 PH/21 Long custom messages for fakedefer and fakereject are now split up
3351 into multiline responses in the same way that messages for "deny" and
3352 other ACL rejections are.
3354 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3355 with slight modification.
3357 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3358 draft, changing the syntax and factoring some duplicate code".
3360 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3361 for deliveries of the second and subsequent messages over the same SMTP
3364 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3365 ${reduce, with only minor "tidies".
3367 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3369 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3370 expansion side effects.
3372 PH/27 When a message times out after an over-quota error from an Exim-imposed
3373 quota, the bounce message says "mailbox is full". This message was not
3374 being given when it was a system quota that was exceeded. It now should
3377 MH/03 Made $recipients available in local_scan(). local_scan() already has
3378 better access to the recipient list through recipients_list[], but
3379 $recipients can be useful in postmaster-provided expansion strings.
3381 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3382 in the case of a MAIL command with additional options following the
3383 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3384 were accidentally chopped off.
3386 PH/29 SMTP synchronization checks are implemented when a command is read -
3387 there is a check that no more input is waiting when there shouldn't be
3388 any. However, for some commands, a delay in an ACL can mean that it is
3389 some time before the response is written. In this time, more input might
3390 arrive, invalidly. So now there are extra checks after an ACL has run for
3391 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3392 pipelining has not been advertised.
3394 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3396 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3397 returned in the SMTP response when smtp_return_error_details was false.
3398 This has been fixed.
3400 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3401 instead of the C I/O that was originally supplied, because problems were
3402 reported on Solaris.
3404 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3405 Exim which did not show up earlier: it was assuming that a call to
3406 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3407 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3408 was a macro that became an assignment, so it seemed to work. This has
3409 changed to a proper function call with a void return, hence the compile
3410 error. Exim's code has been fixed.
3412 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3415 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3417 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3419 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3420 because it turns out to be too big or there is a timeout) there is no
3421 "Completed" line in the log. When some messages of this type were
3422 selected by exigrep, they were listed as "not completed". Others were
3423 picked up by some special patterns. I have improved the selection
3424 criteria to be more general.
3426 PH/38 The host_find_failed option in the manualroute router can now be set
3427 to "ignore", to completely ignore a host whose IP address cannot be
3428 found. If all hosts are ignored, the behaviour is controlled by the new
3429 host_all_ignored option.
3431 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3432 homing or because of multiple MX records with /mx) generated more than
3433 one IP address, and the following item turned out to be the local host,
3434 all the secondary addresses of the first item were incorrectly removed
3435 from the list, along with the local host and any following hosts (which
3436 is what is supposed to happen).
3438 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3439 whoever called Exim into the -H file. In the case of the daemon it was
3440 behaving confusingly. When first started, it used values for whoever
3441 started the daemon, but after a SIGHUP it used the Exim user (because it
3442 calls itself on a restart). I have changed the code so that it now always
3445 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3446 message are rejected with the same error (e.g. no authentication or bad
3447 sender address), and a DATA command is nevertheless sent (as can happen
3448 with PIPELINING or a stupid MUA), the error message that was given to the
3449 RCPT commands is included in the rejection of the DATA command. This is
3450 intended to be helpful for MUAs that show only the final error to their
3453 PH/42 Another patch from the Sieve maintainer.
3455 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3456 Eximstats - Fixed some broken HTML links and added missing column headers
3458 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3459 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3461 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3463 PH/43 Yet another patch from the Sieve maintainer.
3465 PH/44 I found a way to check for a TCP/IP connection going away before sending
3466 the response to the final '.' that terminates a message, but only in the
3467 case where the client has not sent further data following the '.'
3468 (unfortunately, this is allowed). However, in many cases there won't be
3469 any further data because there won't be any more messages to send. A call
3470 to select() can be used: if it shows that the input is "ready", there is
3471 either input waiting, or the socket has been closed. An attempt to read
3472 the next input character can distinguish the two cases. Previously, Exim
3473 would have sent an OK response which the client would never have see.
3474 This could lead to message repetition. This fix should cure that, at
3475 least in a lot of common cases.
3477 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3478 advertised in response to EHLO.
3484 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3485 fixed by 4.65/MH/01 (is this a record?) are fixed:
3487 (i) An empty string was always treated as zero by the numeric comparison
3488 operators. This behaviour has been restored.
3490 (ii) It is documented that the numeric comparison operators always treat
3491 their arguments as decimal numbers. This was broken in that numbers
3492 starting with 0 were being interpreted as octal.
3494 While fixing these problems I realized that there was another issue that
3495 hadn't been noticed. Values of message_size_limit (both the global option
3496 and the transport option) were treated as octal if they started with 0.
3497 The documentation was vague. These values are now always treated as
3498 decimal, and I will make that clear in the documentation.
3504 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3505 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3508 MH/01 Don't check that the operands of numeric comparison operators are
3509 integers when their expansion is in "skipping" mode (fixes bug
3510 introduced by 4.64-PH/07).
3512 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3513 child addresses, Exim now panics and dies. Previously, because the count
3514 is held in a short int, deliveries were likely to be lost. As such a
3515 large number of recipients for a single message is ridiculous
3516 (performance will be very, very poor), I have chosen to impose a limit
3517 rather than extend the field.
3523 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3524 leftover -K file (the existence of which was triggered by #402).
3525 While we were at it, introduced process PID as part of the -K
3526 filename. This should rule out race conditions when creating
3529 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3530 processing considerably. Previous code took too long for large mails,
3531 triggering a timeout which in turn triggers #401.
3533 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3534 in the DK code in transports.c. sendfile() is not really portable,
3535 hence the _LINUX specificness.
3537 TF/01 In the add_headers option to the mail command in an Exim filter,
3538 there was a bug that Exim would claim a syntax error in any
3539 header after the first one which had an odd number of characters
3542 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3543 callout verification, Exim cached a "reject" for the entire domain. This
3544 is correct for most verifications, but it is not correct for a recipient
3545 verification with use_sender or use_postmaster set, because in that case
3546 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3547 case of MAIL FROM:<> rejection from other early rejections (e.g.
3548 rejection of HELO). When verifying a recipient using a non-null MAIL
3549 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3550 Whatever the result of the callout, the value of the domain cache is
3551 left unchanged (for any other kind of callout, getting as far as trying
3552 RCPT means that the domain itself is ok).
3554 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3557 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3558 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3559 session, the daemon ignores ECONNECT errors and logs others; it now
3560 ignores EPIPE as well.
3562 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3563 (quoted-printable decoding).
3565 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3566 later the small subsequent patch to fix an introduced bug.
3568 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3570 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3572 PH/08 An error is now given if message_size_limit is specified negative.
3574 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3575 to be given (somewhat) arbitrary names.
3577 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3580 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3581 miscellaneous code fixes
3583 PH/10 Added the log_reject_target ACL modifier to specify where to log
3586 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3587 hostname. This is wrong, because it relates to the incoming message (and
3588 probably the interface on which it is arriving) and not to the outgoing
3589 callout (which could be using a different interface). This has been
3590 changed to use the value of the helo_data option from the smtp transport
3591 instead - this is what is used when a message is actually being sent. If
3592 there is no remote transport (possible with a router that sets up host
3593 addresses), $smtp_active_hostname is used.
3595 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3596 tweaks were necessary in order to get it to work (see also 21 below):
3597 (a) The code assumed that strncpy() returns a negative number on buffer
3598 overflow, which isn't the case. Replaced with Exim's string_format()
3600 (b) There were several signed/unsigned issues. I just did the minimum
3601 hacking in of casts. There is scope for a larger refactoring.
3602 (c) The code used strcasecmp() which is not a standard C function.
3603 Replaced with Exim's strcmpic() function.
3604 (d) The code set only $1; it now sets $auth1 as well.
3605 (e) A simple test gave the error "authentication client didn't specify
3606 service in request". It would seem that Dovecot has changed its
3607 interface. Fortunately there's a specification; I followed it and
3608 changed what the client sends and it appears to be working now.
3610 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3613 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3614 address A is aliased to B and C, where B exists and C does not. Without
3615 -v the output is "A verified" because verification stops after a
3616 successful redirection if more than one address is generated. However,
3617 with -v the child addresses are also verified. Exim was outputting "A
3618 failed to verify" and then showing the successful verification for C,
3619 with its parentage. It now outputs "B failed to verify", showing B's
3620 parentage before showing the successful verification of C.
3622 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3623 look up a TXT record in a specific list after matching in a combined
3626 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3627 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3628 they consult the DNS. I had assumed they would set it the way they
3629 wanted; and indeed my experiments on Linux seem to show that in some
3630 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3631 To be on the safe side, however, I have now made the interface to
3632 host_find_byname() similar to host_find_bydns(), with an argument
3633 containing the DNS resolver options. The host_find_byname() function now
3634 sets these options at its start, just as host_find_bydns() does. The smtp
3635 transport options dns_qualify_single and dns_search_parents are passed to
3636 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3637 of host_find_byname() use the default settings of RES_DEFNAMES
3638 (qualify_single) but not RES_DNSRCH (search_parents).
3640 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3641 spool_read_header() do less string testing, by means of a preliminary
3642 switch on the second character of optional "-foo" lines. (This is
3643 overdue, caused by the large number of possibilities that now exist.
3644 Originally there were few.) While I was there, I also converted the
3645 str(n)cmp tests so they don't re-test the leading "-" and the first
3646 character, in the hope this might squeeze out yet more improvement.
3648 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3649 flag allowing group syntax was set by the header_syntax check but not
3650 turned off, possible causing trouble later; (2) The flag was not being
3651 set at all for the header_verify test, causing "group"-style headers to
3652 be rejected. I have now set it in this case, and also caused header_
3653 verify to ignore an empty address taken from a group. While doing this, I
3654 came across some other cases where the code for allowing group syntax
3655 while scanning a header line wasn't quite right (mostly, not resetting
3656 the flag correctly in the right place). These bugs could have caused
3657 trouble for malformed header lines. I hope it is now all correct.
3659 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3660 with the "reply" argument non-NULL. The code, however (which originally
3661 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3662 but it didn't always do it. This confused somebody who was copying the
3663 code for some other use. I have removed all the tests.
3665 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3666 feature that was used to support insecure browsers during the U.S. crypto
3667 embargo. It requires special client support, and Exim is probably the
3668 only MTA that supported it -- and would never use it because real RSA is
3669 always available. This code has been removed, because it had the bad
3670 effect of slowing Exim down by computing (never used) parameters for the
3671 RSA_EXPORT functionality.
3673 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3674 authenticator to fail if there's a tab character in the incoming data
3675 (there should never be unless someone is messing about, as it's supposed
3676 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3677 if the connection is using TLS or if the remote IP is the same as the
3678 local IP, and the "valid-client-cert option" if a client certificate has
3681 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3682 authenticators. This can be used for authorization after authentication
3683 succeeds. (In the case of plaintext, it servers for both authentication
3686 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3687 if any retry times were supplied.
3689 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3690 connection, where there is no client IP address to check. In this
3691 situation, the verify now always succeeds.
3693 PH/25 Applied John Jetmore's -Mset patch.
3695 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3697 PH/27 In a string expansion for a processed (not raw) header when multiple
3698 headers of the same name were present, leading whitespace was being
3699 removed from all of them, but trailing whitespace was being removed only
3700 from the last one. Now trailing whitespace is removed from each header
3701 before concatenation. Completely empty headers in a concatenation (as
3702 before) are ignored.
3704 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3705 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3707 PH/29 [Removed. This was a change that I later backed out, and forgot to
3708 correct the ChangeLog entry (that I had efficiently created) before
3709 committing the later change.]
3711 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3712 address errors (4xx response to RCPT) over the same connection as other
3713 messages routed to the same hosts. Such deliveries are always "forced",
3714 so retry times are not inspected. This resulted in far too many retries
3715 for the affected addresses. The effect occurred only when there were more
3716 hosts than the hosts_max_try setting in the smtp transport when it had
3717 the 4xx errors. Those hosts that it had tried were not added to the list
3718 of hosts for which the message was waiting, so if all were tried, there
3719 was no problem. Two fixes have been applied:
3721 (i) If there are any address or message errors in an SMTP delivery, none
3722 of the hosts (tried or untried) are now added to the list of hosts
3723 for which the message is waiting, so the message should not be a
3724 candidate for sending over the same connection that was used for a
3725 successful delivery of some other message. This seems entirely
3726 reasonable: after all the message is NOT "waiting for some host".
3727 This is so "obvious" that I'm not sure why it wasn't done
3728 previously. Hope I haven't missed anything, but it can't do any
3729 harm, as the worst effect is to miss an optimization.
3731 (ii) If, despite (i), such a delivery is accidentally attempted, the
3732 routing retry time is respected, so at least it doesn't keep
3733 hammering the server.
3735 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3736 in ${readsocket because some servers need this prod.
3738 PH/32 Added some extra debug output when updating a wait-xxx database.
3740 PH/33 The hint "could be header name not terminated by colon", which has been
3741 given for certain expansion errors for a long time, was not being given
3742 for the ${if def:h_colon_omitted{... case.
3744 PH/34 The spec says: "With one important exception, whenever a domain list is
3745 being scanned, $domain contains the subject domain." There was at least
3746 one case where this was not true.
3748 PH/35 The error "getsockname() failed: connection reset by peer" was being
3749 written to the panic log as well as the main log, but it isn't really
3750 panic-worthy as it just means the connection died rather early on. I have
3751 removed the panic log writing for the ECONNRESET error when getsockname()
3754 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3755 runs only) independently of the message's sender address. This meant
3756 that, if the 4xx error was in fact related to the sender, a different
3757 message to the same recipient with a different sender could confuse
3758 things. In particular, this can happen when sending to a greylisting
3759 server, but other circumstances could also provoke similar problems.
3760 I have changed the default so that the retry time for these errors is now
3761 based a combination of the sender and recipient addresses. This change
3762 can be overridden by setting address_retry_include_sender=false in the
3765 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3766 remote server are returned as part of bounce messages. This was not
3767 happening for LMTP over a pipe (the lmtp transport), but now it is the
3768 same for both kinds of LMTP.
3770 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3771 in header lines that were in fact CNAMEs. This is no longer the case.
3773 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3774 and queue runs started by the daemon processed all messages. This has
3775 been fixed so that -R and -S can now usefully be given with -q<time>.
3777 PH/40 Import PCRE release 6.7 (fixes some bugs).
3779 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3781 PH/42 Give an error if -q is specified more than once.
3783 PH/43 Renamed the variables $interface_address and $interface_port as
3784 $received_ip_address and $received_port, to make it clear that these
3785 values apply to message reception, and not to the outgoing interface when
3786 a message is delivered. (The old names remain recognized, of course.)
3788 PH/44 There was no timeout on the connect() call when using a Unix domain
3789 socket in the ${readsocket expansion. There now is.
3791 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3792 be meaningful with "accept".
3794 SC/01 Eximstats V1.43
3795 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3797 SC/02 Eximstats V1.44
3798 Use a glob alias rather than an array ref in the generated
3799 parser. This improves both readability and performance.
3801 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3802 Collect SpamAssassin and rejection statistics.
3803 Don't display local sender or destination tables unless
3804 there is data to show.
3805 Added average volumes into the top table text output.
3807 SC/04 Eximstats V1.46
3808 Collect data on the number of addresses (recipients)
3809 as well as the number of messages.
3811 SC/05 Eximstats V1.47
3812 Added 'Message too big' to the list of mail rejection
3813 reasons (thanks to Marco Gaiarin).
3815 SC/06 Eximstats V1.48
3816 Mainlog lines which have GMT offsets and are too short to
3817 have a flag are now skipped.
3819 SC/07 Eximstats V1.49 (Alain Williams)
3820 Added the -emptyok flag.
3822 SC/08 Eximstats V1.50
3823 Fixes for obtaining the IP address from reject messages.
3825 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3826 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3827 whitespace changes from 4.64-PH/27
3829 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3832 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3833 are found, allow negative numbers in numeric criteria)
3835 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3837 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3838 to match changes made in 4.64-PH/43
3840 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3842 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3843 "domain" is actually an IP address, and force a failure. This locks out
3844 those revolvers/nameservers that support "A-for-A" lookups, in
3845 contravention of the specifications.
3847 PH/48 When a host name was looked up from an IP address, and the subsequent
3848 forward lookup of the name timed out, the host name was left in
3849 $sender_host_name, contrary to the specification.
3851 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3852 restricted to single-key lookups, Exim was not diagnosing an error if
3853 * or *@ was used with a query-style lookup.
3855 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3857 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3858 long ago, but noone interested enough thought of it. Let's just say that
3859 the "1.1" means that there are some new functions that weren't there at
3860 some point in the past.
3862 PH/51 Error processing for expansion failure of helo_data from an smtp
3863 transport during callout processing was broken.
3865 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3866 tested/used via the -bh/-bhc/-bs options.
3868 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3869 bug, fixed in subsequent PCRE releases).
3871 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3872 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3874 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3880 SC/01 Use a glob alias rather than an array ref in eximstats generated
3881 parser. This improves both readability and performance.
3883 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3884 Don't display local sender or destination tables in eximstats unless
3885 there is data to show.
3886 Added average volumes into the eximstats top table text output.
3888 SC/03 Collect data on the number of addresses (recipients) as well
3889 as the number of messages in eximstats.
3891 TF/01 Correct an error in the documentation for the redirect router. Exim
3892 does (usually) call initgroups() when daemonizing.
3894 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3895 with consistent privilege compared to when running as a daemon.
3897 TF/03 Note in the spec that $authenticated_id is not set for local
3898 submissions from trusted users.
3900 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3901 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3903 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3904 by adding some example configuration directives to the default
3905 configuration file. A little bit of work is required to uncomment the
3906 directives and define how usernames and passwords are checked, but
3907 there is now a framework to start from.
3909 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3910 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3911 without this. I don't know how relevant this is to other LDAP libraries.
3913 PH/02 Add the verb name to the "unknown ACL verb" error.
3915 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3917 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3919 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3920 directory not expanded when it should be if an expanded home directory
3921 was set for the address (which is overridden by the transport).
3923 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3926 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3927 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3928 because it is too late at that time, and has no effect.
3930 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3931 security issue with \' (bugzilla #107). I could not use the
3932 PQescapeStringConn() function, because it needs a PGconn value as one of
3935 PH/08 When testing addresses using -bt, indicate those final addresses that
3936 are duplicates that would not cause an additional delivery. At least one
3937 person was confused, thinking that -bt output corresponded to deliveries.
3938 (Suppressing duplicates isn't a good idea as you lose the information
3939 about possibly different redirections that led to the duplicates.)
3941 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3942 systems where poll() doesn't work, in particular OS X.
3944 PH/10 Added more information to debugging output for retry time not reached.
3946 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3947 operations in malware.c.
3949 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3952 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3953 syslog_duplication set false, log lines that would normally be written
3954 both the the main log and to the reject log were not written to syslog at
3957 PH/14 In the default configuration, change the use of "message" in ACL warn
3958 statements to "add_header".
3960 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3961 not followed by a command (e.g. "seen endif").
3963 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3964 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3967 PH/17 Added extra conditions to the default value of delay_warning_condition
3971 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3972 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3973 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3976 The Auto-Submitted: and various List- headers are standardised, whereas I
3977 don't think Precedence: ever was.
3979 PH/18 Refactored debugging code in route_finduser() to show more information,
3980 in particular, the error code if getpwnam() issues one.
3982 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3983 This is apparently needed in addition to the PH/07 change above to avoid
3984 any possible encoding problems.
3986 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3987 but not after initializing Perl.
3989 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3990 output them only if debugging. By default they are written stderr,
3991 apparently, which is not desirable.
3993 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3996 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3999 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4001 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4002 authenticated or an ident call has been made. Suppress the default
4003 values for $authenticated_id and $authenticated_sender (but permit -oMai
4004 and -oMas) when testing with -bh.
4006 PH/24 Re-jigged the order of the tests in the default configuration so that the
4007 tests for valid domains and recipients precede the DNS black list and CSA
4008 tests, on the grounds that those ones are more expensive.
4010 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4011 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4012 This bug exists in every version of Exim that I still have, right back to
4015 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4016 However, an attempt to turn on case-sensitivity in a regex key by
4017 including (?-i) didn't work because the subject string was already
4018 lowercased, and the effects were non-intuitive. It turns out that a
4019 one-line patch can be used to allow (?-i) to work as expected.
4025 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4026 other effects) broke the use of negated acl sub-conditions.
4028 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4031 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4032 "Deny" causes Exim to reject the incoming connection with a 554 error.
4033 Unfortunately, if there is a major crisis, such as a disk failure,
4034 tcp-wrappers gives "deny", whereas what one would like would be some
4035 kind of temporary error. A kludge has been added to help with this.
4036 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4037 554 error is used if errno is still zero or contains ENOENT (which occurs
4038 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4041 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4043 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4044 errors. Otherwise a message that provokes a temporary error (when other
4045 messages do not) can cause a whole host to time out.
4047 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4048 addresses were routed directly to files or pipes from a redirect router.
4049 File deliveries just didn't batch; pipe deliveries might have suffered
4052 PH/06 A failure to get a lock for a hints database would erroneously always say
4053 "Failed to get write lock", even when it was really a read lock.
4055 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4056 of 0600. This has been changed to use the value of the lockfile_mode
4057 option (which defaults to 0600).
4059 PH/08 Applied small patch from the Sieve maintainer.
4061 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4062 folder from quota calculations, a direct delivery into this folder messed
4063 up the contents of the maildirsize file. This was because the regex was
4064 used only to exclude .Trash (or whatever) when the size of the mailbox
4065 was calculated. There was no check that a delivery was happening into an
4066 excluded directory. This bug has been fixed by ignoring all quota
4067 processing for deliveries into excluded directories.
4069 PH/10 Added the maildirfolder_create_regex option to appendfile.
4075 PH/01 The code for finding all the local interface addresses on a FreeBSD
4076 system running IPv6 was broken. This may well have applied to all BSD
4077 systems, as well as to others that have similar system calls. The broken
4078 code found IPv4 interfaces correctly, but gave incorrect values for the
4079 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4080 that it would not match correctly against @[] and not recognize the IPv6
4083 PH/02 The ipliteral router was not recognizing addresses of the form user@
4084 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4086 PH/03 Added disable_ipv6.
4088 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4089 decoded form, because it is most often used to construct To: headers
4090 lines in autoreplies, and the decoded form may well be syntactically
4091 invalid. However, $reply_address has leading white space removed, and all
4092 newlines turned into spaces so that the autoreply transport does not
4095 PH/05 If group was specified without a user on a router, and no group or user
4096 was specified on a transport, the group from the router was ignored.
4098 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4099 for visible compile-time settings that can be used to change these
4100 numbers, for those that want even more. Backwards compatibility with old
4101 spool files has been maintained. However, going back to a previous Exim
4102 release will lost any variables that are in spool files.
4104 PH/07 Two small changes when running in the test harness: increase delay when
4105 passing a TCP/IP connection to a new process, in case the original
4106 process has to generate a bounce, and remove special handling of
4107 127.0.0.2 (sic), which is no longer necessary.
4109 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4110 be the same on different OS.
4112 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4115 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4116 whether --show-vars was specified or not
4118 JJ/02 exipick: Added support for new ACL variable spool format introduced
4121 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4122 syntactically invalid From: or Reply-to: line, and a filter used this to
4123 generate an autoreply, and therefore failed to obtain an address for the
4124 autoreply, Exim could try to deliver to a non-existent relative file
4125 name, causing unrelated and misleading errors. What now happens is that
4126 it logs this as a hard delivery error, but does not attempt to create a
4129 PH/11 The exinext utility has a -C option for testing purposes, but although
4130 the given file was scanned by exinext itself; it wasn't being passed on
4131 when Exim was called.
4133 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4134 an end-of-file indication when reading a command response.
4136 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4137 compiled. In many other places in Exim, IPv6 addresses are always
4138 recognized, so I have changed this. It also means that IPv4 domain
4139 literals of the form [IPV4:n.n.n.n] are now always recognized.
4141 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4142 used if the router is not running as root, for example, when verifying at
4143 ACL time, or when using -bh. The debugging output from this situation was
4144 non-existent - all you got was a failure to exec. I have made two
4147 (a) Failures to set uid/gid, the current directory, or a process leader
4148 in a subprocess such as that created by queryprogram now generate
4149 suitable debugging output when -d is set.
4151 (b) The queryprogram router detects when it is not running as root,
4152 outputs suitable debugging information if -d is set, and then runs
4153 the subprocess without attempting to change uid/gid.
4155 PH/15 Minor change to Makefile for building test_host (undocumented testing
4158 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4159 additional section of a DNS packet that returns MX or SRV records.
4160 Instead, it always explicitly searches for A/AAAA records. This avoids
4161 major problems that occur when a DNS server includes only records of one
4162 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4163 fixed another bug: if SRV records were looked up and the corresponding
4164 address records were *not* found in the additional section, the port
4165 values from the SRV records were lost.
4167 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4168 using the correct key (the original address) when searching the retry
4169 rules in order to find which one to use for generating the retry hint.
4171 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4172 adding the default one. Similarly, if it contains a Reply-To: header, the
4173 errors_reply_to option, if set, is not used.
4175 PH/19 When calculating a retry time, Exim used to measure the "time since
4176 failure" by looking at the "first failed" field in the retry record. Now
4177 it does not use this if it is later than than the arrival time of the
4178 message. Instead it uses the arrival time. This makes for better
4179 behaviour in cases where some deliveries succeed, thus re-setting the
4180 "first failed" field. An example is a quota failure for a huge message
4181 when small messages continue to be delivered. Without this change, the
4182 "time since failure" will always be short, possible causing more frequent
4183 delivery attempts for the huge message than are intended.
4184 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4186 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4187 $1, $2, $3) because the numerical variables can be reset during some
4188 expansion items (e.g. "match"), thereby losing the authentication data.
4190 PH/21 Make -bV show the size of off_t variables so that the test suite can
4191 decide whether to run tests for quotas > 2G.
4193 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4194 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4195 filecount value is greater than 2G or if a quota value is greater than 2G
4196 on a system where the size of off_t is not greater than 4, a panic error
4199 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4200 never match. The debug and -bh output now contains an explicit error
4201 message indicating a malformed IPv4 address or mask.
4203 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4204 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4205 PH/23 above applies.
4207 PH/25 Do not write to syslog when running in the test harness. The only
4208 occasion when this arises is a failure to open the main or panic logs
4209 (for which there is an explicit test).
4211 PH/26 Added the /no_tell option to "control=freeze".
4213 PH/27 If a host name lookup failed very early in a connection, for example, if
4214 the IP address matched host_lookup and the reverse lookup yielded a name
4215 that did not have a forward lookup, an error message of the form "no IP
4216 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4217 could be logged. Now it outputs the IP address instead of "NULL".
4219 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4220 allows the sender and the authenticated sender to be set when
4221 submitting a message from within Exim. Since child_open_exim() is
4222 documented for local_scan(), the new function should be too.
4224 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4225 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4226 results in an empty string is now treated as unset.
4228 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4230 PH/31 Added sender_verify_fail logging option.
4232 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4233 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4234 tidied the source and removed it altogether.
4236 PH/33 When a queue run was abandoned because the load average was too high, a
4237 log line was always written; now it is written only if the queue_run log
4238 selector is set. In addition, the log line for abandonment now contains
4239 information about the queue run such as the pid. This is always present
4240 in "start" and "stop" lines but was omitted from the "abandon" line.
4242 PH/34 Omit spaces between a header name and the colon in the error message that
4243 is given when verify = headers_syntax fails (if there are lots of them,
4244 the message gets confusing).
4246 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4247 names, as there are now some PTR records that contain slashes. This check
4248 is only to protect against broken name servers that fall over on strange
4249 characters, so the fact that it applies to all lookups doesn't matter.
4251 PH/36 Now that the new test suite is complete, we can remove some of the
4252 special code in Exim that was needed for the old test suite. For example,
4253 sorting DNS records because real resolvers return them in an arbitrary
4254 order. The new test suite's fake resolver always returns records in the
4257 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4258 bounces) except when queue_only is set, to avoid logging races between
4259 the different processes.
4261 PH/38 Panic-die if .include specifies a non-absolute path.
4263 PH/39 A tweak to the "H" retry rule from its user.
4265 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4266 a label. They prevented compilation on older perls.
4268 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4269 a warning to be raised on newish perls.
4271 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4272 on queue. Changes to match documented behaviour of showing count of
4273 messages matching specified criteria.
4275 PH/40 Changed the default ident timeout from 30s to 5s.
4277 PH/41 Added support for the use of login_cap features, on those BSD systems
4278 that have them, for controlling the resources used by pipe deliveries.
4280 PH/42 The content-scanning code uses fopen() to create files in which to put
4281 message data. Previously it was not paying any attention to the mode of
4282 the files. Exim runs with umask(0) because the rest of the code creates
4283 files with open(), and sets the required mode explicitly. Thus, these
4284 files were ending up world-writeable. This was not a big issue, because,
4285 being within the spool directory, they were not world-accessible. I have
4286 created a function called modefopen, which takes an additional mode
4287 argument. It sets umask(777), creates the file, chmods it to the required
4288 mode, then resets the umask. All the relevant calls to fopen() in the
4289 content scanning code have been changed to use this function.
4291 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4292 to 24 hours. This avoids potential overflow problems when processing G
4293 and H retry rules. I suspect nobody ever tinkers with this value.
4295 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4297 PH/45 When the plaintext authenticator is running as a client, the server's
4298 challenges are checked to ensure they are valid base64 strings. By
4299 default, the authentication attempt is cancelled if an invalid string is
4300 received. Setting client_ignore_invalid_base64 true ignores these errors.
4301 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4302 they are received. Thus, the responses can be made to depend on the
4303 challenges. If an invalid string is ignored, an empty string is placed in
4306 PH/46 Messages that are created by the autoreply transport now contains a
4307 References: header, in accordance with RFCs 2822 and 3834.
4309 PH/47 Added authenticated_sender_force to the smtp transport.
4311 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4313 PH/49 Installed latest patch from the Sieve maintainer.
4315 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4316 was also set, the appendfile transport was unnecessarily scanning a
4317 directory of message files (e.g. for maildir delivery) to find the count
4318 of files (along with the size), even though it did not need this
4319 information. It now does the scan only if it needs to find either the
4320 size of the count of files.
4322 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4324 PH/52 Two bugs concerned with error handling when the smtp transport is
4327 (i) Exim was not creating retry information for temporary errors given
4328 for individual recipients after the DATA command when the smtp transport
4329 was used in LMTP mode. This meant that they could be retried too
4330 frequently, and not timed out correctly.
4332 (ii) Exim was setting the flag that allows error details to be returned
4333 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4334 recipients that were returned after the DATA command.
4336 PH/53 This is related to PH/52, but is more general: for any failing address,
4337 when detailed error information was permitted to be returned to the
4338 sender, but the error was temporary, then after the final timeout, only
4339 "retry timeout exceeded" was returned. Now it returns the full error as
4340 well as "retry timeout exceeded".
4342 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4343 do this, and (what is worse) MTAs that accept it.
4345 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4346 will now be deprecated.
4348 PH/56 New os.c-cygwin from the Cygwin maintainer.
4350 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4351 formats (previously only available in exim formats via -bpr, -bpru,
4352 and -bpra. Now also available in native and exiqgrep formats)
4354 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4355 with very large, slow to parse queues
4357 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4359 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4361 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4362 responses to authentication challenges, though it was showing the
4363 challenges; (ii) I've removed the CR characters from the debug output for
4366 PH/58 Allow for the insertion of a newline as well as a space when a string
4367 is turned into more than one encoded-word during RFC 2047 encoding. The
4368 Sieve code now uses this.
4370 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4371 data_4xx, lost_connection, tls_required.
4373 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4374 message was being sent as an SMTP response.
4376 PH/61 Add -l and -k options to exicyclog.
4378 PH/62 When verifying, if an address was redirected to one new address, so that
4379 verification continued, and the new address failed or deferred after
4380 having set something in $address_data, the value of $address_data was not
4381 passed back to the ACL. This was different to the case when no
4382 redirection occurred. The value is now passed back in both cases.
4384 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4385 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4386 use login_cap.h, so on its own it isn't the distinguishing feature. The
4387 new name refers directly to the setclassresources() function.
4389 PH/65 Added configuration files for NetBSD3.
4391 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4393 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4394 is preferred over IPv4.
4396 PH/68 The bounce_return_message and bounce_return_body options were not being
4397 honoured for bounces generated during the reception of non-SMTP messages.
4398 In particular, this applied to messages rejected by the ACL. This bug has
4399 been fixed. However, if bounce_return_message is true and bounce_return_
4400 body is false, the headers that are returned for a non-SMTP message
4401 include only those that have been read before the error was detected.
4402 (In the case of an ACL rejection, they have all been read.)
4404 PH/69 The HTML version of the specification is now built in a directory called
4405 spec_html instead of spec.html, because the latter looks like a path with
4406 a MIME-type, and this confuses some software.
4408 PH/70 Catch two compiler warnings in sieve.c.
4410 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4411 function verify_get_ident() calls ip_connect() to connect a socket, but
4412 if the "connect()" function timed out, ip_connect() used to close the
4413 socket. However, verify_get_ident() also closes the socket later, and in
4414 between Exim writes to the log, which may get opened at this point. When
4415 the socket was closed in ip_connect(), the log could get the same file
4416 descriptor number as the socket. This naturally causes chaos. The fix is
4417 not to close the socket in ip_connect(); the socket should be closed by
4418 the function that creates it. There was only one place in the code where
4419 this was missing, in the iplookup router, which I don't think anybody now
4420 uses, but I've fixed it anyway.
4422 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4423 well as to direct DNS lookups. Otherwise the handling of names in host
4424 lists is inconsistent and therefore confusing.
4430 PH/01 Two changes to the default runtime configuration:
4432 (1) Move the checks for relay_from_hosts and authenticated clients from
4433 after to before the (commented out) DNS black list checks.
4435 (2) Add control=submission to the relay_from_hosts and authenticated
4436 clients checks, on the grounds that messages accepted by these
4437 statements are most likely to be submissions.
4439 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4441 (1) Generate an error if the third argument for the ${prvs expansion is
4444 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4447 (3) Reset the variables that are obtained from the first argument of
4448 ${prvscheck and used in the second argument before leaving the code,
4449 because their memory is reclaimed, so using them afterwards may do
4452 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4453 one (it's much easier than Tom thought :-).
4455 (5) Because of (4), we can now allow for the use of $prvscheck_result
4456 inside the third argument.
4458 PH/03 For some reason, the default setting of PATH when running a command from
4459 a pipe transport was just "/usr/bin". I have changed it to
4462 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4463 anything to be listed in the output from -bV.
4465 PH/05 When a filter generated an autoreply, the entire To: header line was
4466 quoted in the delivery log line, like this:
4468 => >A.N.Other <ano@some.domain> <original@ddress> ...
4470 This has been changed so that it extracts the operative address. There
4471 may be more than one such address. If so, they are comma-separated, like
4474 => >ano@some.domain,ona@other.domain <original@ddress> ...
4476 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4477 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4478 not being looked up in the rDNS to get a host name, Exim was showing the
4479 IP address twice in Received: lines, even though the IP addresses were
4480 identical. For example:
4482 Received: from [1.2.3.4] (helo=[1.2.3.4])
4484 However, if the real host name was known, it was omitting the HELO data
4485 if it matched the actual IP address. This has been tidied up so that it
4486 doesn't show the same IP address twice.
4488 PH/07 When both +timestamp and +memory debugging was on, the value given by
4489 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4490 called by the debug printing, thereby overwriting the timestamp buffer.
4491 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4493 PH/08 When the original message was included in an autoreply transport, it
4494 always said "this is a copy of the message, including all the headers",
4495 even if body_only or headers_only was set. It now gives an appropriate
4498 PH/09 Applied a patch from the Sieve maintainer which:
4500 o fixes some comments
4501 o adds the (disabled) notify extension core
4502 o adds some debug output for the result of if/elsif tests
4503 o points to the current vacation draft in the documentation
4504 and documents the missing references header update
4508 o fixes a bug in processing the envelope test (when testing
4509 multiple envelope elements, the last element determined the
4512 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4513 Electronic Mail") by including:
4515 Auto-submitted: auto-generated
4517 in the messages that it generates (bounce messages and others, such as
4518 warnings). In the case of bounce messages for non-SMTP messages, there was
4519 also a typo: it was using "Auto_submitted" (underscore instead of
4520 hyphen). Since every message generated by Exim is necessarily in response
4521 to another message, thes have all been changed to:
4523 Auto-Submitted: auto-replied
4525 in accordance with these statements in the RFC:
4527 The auto-replied keyword:
4529 - SHOULD be used on messages sent in direct response to another
4530 message by an automatic process,
4532 - MUST NOT be used on manually-generated messages,
4534 - MAY be used on Delivery Status Notifications (DSNs) and Message
4535 Disposition Notifications (MDNs),
4537 - MUST NOT be used on messages generated by automatic or periodic
4538 processes, except for messages which are automatic responses to
4541 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4542 to the default Received: header definition.
4544 PH/12 Added log selector acl_warn_skipped (default on).
4546 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4547 variables because (a) they are in the wrong storage pool and (b) even if
4548 they were copied, it wouldn't work properly because of the caching.
4550 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4551 checking when decoding. Apparently there are clients that generate
4552 overlong encoded strings. Why am I not surprised?
4554 PH/15 If the first argument of "${if match_address" was not empty, but did not
4555 contain an "@" character, Exim crashed. Now it writes a panic log message
4556 and treats the condition as false.
4558 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4560 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4561 extension "envelope-auth". The code is finished and in agreement with
4562 other implementations, but there is no documentation so far and in fact,
4563 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4564 not changing the active code.
4566 Print executed "if" and "elsif" statements when debugging is used. This
4567 helps a great deal to understand what a filter does.
4569 Document more things not specified clearly in RFC3028. I had all this
4570 sorted out, when out of a sudden new issues came to my mind. Oops."
4572 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4575 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4576 canonical form (as documented). However, after a host name lookup from
4577 the IP address, check_host() was doing a simple string comparison with
4578 addresses acquired from the DNS when checking that the found name did
4579 have the original IP as one of its addresses. Since any found IPv6
4580 addresses are likely to be in abbreviated form, the comparison could
4581 fail. Luckily, there already exists a function for doing the comparison
4582 by converting both addresses to binary, so now that is used instead of
4583 the text comparison.
4585 PH/20 There was another similar case to PH/19, when a complete host name was
4586 given in a host list; looking up its IP address could give an abbreviated
4587 form, whereas the current host's name might or might not be abbreviated.
4588 The same fix has been applied.
4594 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4595 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4598 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4599 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4601 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4603 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4604 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4605 Also, for "Auto-Submitted", treat anything other than "no" as
4606 non-personal, in accordance with RFC 3834. (Previously it treated
4607 anything starting "auto-" as non-personal.)
4609 TF/01 The control=submission/name=... option had a problem with syntax
4610 errors if the name included a slash character. The /name= option
4611 now slurps the rest of the string, so it can include any characters
4612 but it must come last in the list of options (after /sender_retain
4615 PH/05 Some modifications to the interface to the fake nameserver for the new
4623 TK/01 Added the "success_on_redirect" address verification option. See
4624 NewStuff for rationale and an example.
4626 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4628 PH/02 Patch to exigrep to allow it to work on syslog lines.
4630 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4631 fread() to skip over the body file's header line, because in Cygwin the
4632 header line is locked and is inaccessible.
4634 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4635 co-exist for some time) to make it clear that it is the Exim ID that is
4636 referenced, not the Message-ID: header line.
4638 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4639 string_format() function, because snprintf() does not exist on all
4642 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4643 ${stat: expansion item.
4645 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4646 protocol synchronization error", to keep the pedants happy.
4648 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4649 well as for IRIX systems, when gcc is being used. See the host.c source
4652 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4654 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4657 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4658 transport and to the smtp transport in LMTP mode.
4660 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4662 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4663 run a filter in a subprocess. This could lead to confusion in subsequent
4664 lookups in the parent process. There should also be a search_tidyup() at
4665 the end of the subprocess.
4667 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4668 only if the host matched helo_try_verify_hosts, which caused the
4669 verification to occur when the EHLO/HELO command was issued. The ACL just
4670 tested the remembered result. Now, if a previous verification attempt has
4671 not happened, "verify = helo" does it there and then.
4673 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4675 TK/03 Fix log output including CR from clamd.
4677 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4678 exist provoked a memory error which could cause a segfault.
4680 PH/15 Installed PCRE 6.2
4682 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4684 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4685 of the problem. Specifically, suggested +O2 rather than +O1 for the
4688 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4690 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4691 record, the port was not correctly logged when the outgoing_port log
4692 selector was set (it logged the transort's default port).
4694 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4695 fallback_hosts, and "hosts" in the smtp transport.
4697 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4698 host errors such as "Connection refused".
4700 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4701 authentication with radiusclient 0.4.9:
4703 - Error returned from rc_read_config was caught wrongly
4704 - Username/password not passed on to radius server due to wrong length.
4706 The presumption is that some radiusclient API changes for 4.51/PH/17
4707 were not taken care of correctly. The code is still untested by me (my
4708 Linux distribution still has 0.3.2 of radiusclient), but it was
4709 contributed by a Radius user.
4711 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4712 expanding the "port" option of the smtp transport.
4714 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4715 while reading a MIME header. Thanks to Tom Hughes for a patch.
4717 PH/24 Include config.h inside local_scan.h so that configuration settings are
4720 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4721 that in an ACL for RCPT (for example), you can examine exactly what was
4724 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4725 commands, but it was not correctly comparing the address with the actual
4726 client host address. Thus, it would show the EHLO address in Received:
4727 header lines when this was not necessary.
4729 PH/27 Added the % operator to ${eval:}.
4731 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4732 it should be ignoring failures (because with -C, for example, it has lost
4733 privilege). It wasn't ignoring creation failures other than "already
4736 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4739 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4740 because an input line was too long, either on its own, or by virtue of
4741 too many continuations, the temporary file was not being removed, and the
4742 return code was incorrect.
4744 PH/31 Missing "BOOL" in function definition in filtertest.c.
4746 PH/32 Applied Sieve patches from the maintainer.
4748 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4750 PH/33 Added "verify = not_blind".
4752 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4753 Local/Makefile (with some defaults set). These are used in built scripts
4754 such as exicyclog, but they have never been used in the exim_install
4755 script (though there are many overriding facilities there). I have
4756 arranged that the exim_install script now takes note of these two
4759 PH/35 Installed configuration files for Dragonfly.
4761 PH/36 When a locally submitted message by a trusted user did not contain a
4762 From: header, and the sender address was obtained from -f or from an SMTP
4763 MAIL command, and the trusted user did not use -F to supply a sender
4764 name, $originator_name was incorrectly used when constructing a From:
4765 header. Furthermore, $originator_name was used for submission mode
4766 messages from external hosts without From: headers in a similar way,
4767 which is clearly wrong.
4769 PH/37 Added control=suppress_local_fixups.
4771 PH/38 When log_selector = +received_sender was set, and the addition of the
4772 sender made the log line's construction buffer exactly full, or one byte
4773 less than full, an overflow happened when the terminating "\n" was
4776 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4777 when the result of a list match is failure because a DNS lookup failed.
4779 PH/40 RM_COMMAND is now used in the building process.
4781 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4782 the "build-* directories that it finds.
4784 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4785 address in a domain literal was a prefix of an interface address.
4787 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4788 when verifying a sender address, unless rewrite_headers is false.
4790 PH/44 Wrote a long comment about why errors_to addresses are verified as
4791 recipients, not senders.
4793 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4794 the ratelimit ACL was added.
4796 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4798 PH/46 Added extra information about PostgreSQL errors to the error string.
4800 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4801 avoiding the need to install special zones in a real server. This is
4802 backwards compatible; if it can't find the fake resolver, it drops back.
4803 Thus, both old and new test suites can be run.
4805 TF/02 Added util/ratelimit.pl
4807 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4808 clock is set back in time.
4810 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4811 Candler <B.Candler@pobox.com>.
4813 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4814 OK for addresses that are the result of redirections.
4816 PH/48 A number of further additions for the benefit of the new test suite,
4817 including a fake gethostbyname() that interfaces to the fake DNS resolver
4820 TF/06 The fix for widen_domains has also been applied to qualify_single and
4821 search_parents which are the other dnslookup options that can cause
4824 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4827 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4829 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4830 service" if it encountered an MX record with an empty target hostname.
4831 The message is now "an MX or SRV record indicated no SMTP service".
4833 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4834 if the DNS of the sending site is misconfigured. This is quite a
4835 common situation. This change restores the behaviour of treating a
4836 helo verification defer as a failure.
4838 PH/51 If self=fail was set on a router, the bounce message did not include the
4839 actual error message.
4845 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4847 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4848 command itself ended in error, the underlying message about the transport
4849 filter timeout was being overwritten with the pipe command error. Now the
4850 underlying error message should be appended to the second error message.
4852 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4854 PH/02 Reduce the amount of output that "make" produces by default. Full output
4855 can still be requested.
4857 PH/03 The warning log line about a condition test deferring for a "warn" verb
4858 was being output only once per connection, rather than after each
4859 occurrence (because it was using the same function as for successful
4860 "warn" verbs). This seems wrong, so I have changed it.
4862 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4863 it should not have, which might have caused a crash in the right
4864 circumstances, but probably never did.
4866 PH/04 Installed a modified version of Tony Finch's patch to make submission
4867 mode fix the return path as well as the Sender: header line, and to
4868 add a /name= option so that you can make the user's friendly name appear
4871 TF/03 Added the control = fakedefer ACL modifier.
4873 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4874 Mark Lowes for thorough testing.
4876 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4878 TK/03 Merged latest SRS patch from Miles Wilton.
4880 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4881 with the definition in sysexits.h (which is #included earlier).
4882 Fortunately, Exim does not actually use EX_OK. The code used to try to
4883 preserve the sysexits.h value, by assuming that macro definitions were
4884 scanned for macro replacements. I have been disabused of this notion,
4885 so now the code just undefines EX_OK before #including unistd.h.
4887 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4888 in the smtp transport. When a block could not be written in a single
4889 write() function, the timeout was being re-applied to each part-write.
4890 This seems wrong - if the receiver was accepting one byte at a time it
4891 would take for ever. The timeout is now adjusted when this happens. It
4892 doesn't have to be particularly precise.
4894 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4895 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4897 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4898 without a domain if the check to <postmaster@domain> fails.
4900 SC/01 Eximstats: added -xls and the ability to specify output files
4901 (patch written by Frank Heydlauf).
4903 SC/02 Eximstats: use FileHandles for outputting results.
4905 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4907 SC/04 Eximstats: fixed display of large numbers with -nvr option
4909 SC/05 Eximstats: fixed merging of reports with empty tables.
4911 SC/06 Eximstats: added the -include_original_destination flag
4913 SC/07 Eximstats: removed tabs and trailing whitespace.
4915 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4917 TK/06 MBOX spool code: Add real "From " MBOX separator line
4918 so the .eml file is really in mbox format (even though
4919 most programs do not really care). Patch from Alex Miller.
4921 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4922 The latter is generated from $received_to and is only set if the
4923 message has one envelope recipient. SA can use these headers,
4924 obviously out-of-the-box. Patch from Alex Miller.
4926 PH/08 The ${def test on a variable was returning false if the variable's
4927 value was "0", contrary to what the specification has always said!
4928 The result should be true unless the variable is empty.
4930 PH/09 The syntax error of a character other than { following "${if
4931 def:variable_name" (after optional whitespace) was not being diagnosed.
4932 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4933 accidental colon was present, for example, could give incorrect results.
4935 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4936 result is used (not including appendfile, where other changes are about
4939 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4940 This involved changing a lot of size variables from int to off_t. It
4941 should work with maildirs and everything.
4943 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4944 spamd dying while we are connected to it.
4946 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4949 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4950 function for BDB 4.3.
4952 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4954 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4955 the sources for PCRE have been re-arranged and more files are now
4958 PH/15 The code I had for printing potentially long long variables in PH/11
4959 above was not the best (it lost precision). The length of off_t variables
4960 is now inspected at build time, and an appropriate printing format (%ld
4961 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4962 to be "long long int" or "long int". This is needed for the internal
4963 formatting function string_vformat().
4965 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4966 the configuration file to be ":syslog", then the script "guesses" where
4967 the logs files are, rather than using the compiled in default. In our
4968 case the guess is not the same as the compiled default, so the script
4969 suddenly stopped working when I started to use syslog. The patch checks
4970 to see if log_file_path is "". If so, it attempts to read it from exim
4971 with no configuration file to get the compiled in version, before it
4972 falls back to the previous guessing code."
4974 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4975 implementing BATV in an Exim configuration. See NewStuff for the gory
4978 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4979 Makefile that are specific to HP-UX.
4981 PH/18 If the "use_postmaster" option was set for a recipient callout together
4982 with the "random" option, the postmaster address was used as the MAIL
4983 FROM address for the random test, but not for the subsequent recipient
4984 test. It is now used for both.
4986 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4987 patch removes a few documentation additions to RFC 3028, because the
4988 latest draft now contains them. It adds the new en;ascii-case comparator
4989 and a new error check for 8bit text in MIME parts. Comparator and
4990 require names are now matched exactly. I enabled the subaddress
4991 extension, but it is not well tested yet (read: it works for me)."
4993 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4994 rework some of the code of TK/09 above to avoid the hardwired use of
4995 "%lld" and "long long". Replaced the call to snprintf() with a call to
4998 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4999 records point to non-existent hosts", "retry timeout exceeded", and
5000 "retry time not reached for any host after a long failure period".
5002 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5003 experimental DomainKeys support:
5005 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5006 (2) On an error such as an illegally used "control", the wrong name for
5007 the control was given.
5009 These problems did NOT occur unless DomainKeys support was compiled.
5011 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5013 PH/24 Added ${if match_ip condition.
5015 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5016 databases so that it will be absolutely obvious if a crash occurs in the
5017 DB library. This is a regular occurrence (often caused by mis-matched
5020 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5021 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5022 picked up on a user's system that detects such things. There doesn't seem
5023 to be a gcc warning option for this - only an attribute that has to be
5024 put on the function's prototype. It seems that in Fedora Core 4 they have
5025 set this on a number of new functions. No doubt there will be more in due
5028 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5029 specify a transport. However, if an address that was verified by such a
5030 router was the subject of a callout, Exim crashed because it tried to
5031 read the rcpt_include_affixes from the non-existent transport. Now it
5032 just assumes that the setting of that option is false. This bug was
5033 introduced by 4.51/PH/31.
5035 PH/28 Changed -d+all to exclude +memory, because that information is very
5036 rarely of interest, but it makes the output a lot bigger. People tend to
5037 do -d+all out of habit.
5039 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5040 code in os-type was giving problems when libc.so lives in lib64, like on
5043 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5044 aren't the modern standard, and it seems that some systems' include files
5045 don't always have them. Exim was already checking for some of the newer
5046 ones like T_AAAA, and defining it itself. I've added checks for all the
5047 record types that Exim uses.
5049 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5050 not automatically generating a new one, as it is supposed to. This
5051 prevented TLS from working. If the file did exist, but contained invalid
5052 data, a new version was generated, as expected. It was only the case of a
5053 non-existent file that was broken.
5055 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5056 with a change in libdomainkeys > 0.64.
5058 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5059 from DNS. If the selector record carries the flag, it now has
5060 precedence over the domain-wide flag.
5062 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5064 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5065 the use of a port name that isn't defined in /etc/services) occurred, the
5066 message was deferred as in a normal delivery, and thus remained on the
5067 spool, instead of being failed because of the mua_wrapper setting. This
5068 is now fixed, and I tidied up some of the mua_wrapper messages at the
5071 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5072 the messages in a hash of arrays rather than using individual hashes.
5073 This is a bit cleaner and results in dramatic memory savings, albeit
5074 at a slight CPU cost.
5076 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5077 as requested by Marc Sherman.
5079 SC/10 Eximstats: added histograms for user specified patterns as requested
5082 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5084 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5085 fopen() in the content-scanning modules that did not already have it.
5091 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5092 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5094 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5096 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5098 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5099 to test Sieve filters that use "vacation".
5101 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5102 that changes the way the GnuTLS parameters are stored in the cache file.
5103 The new format can be generated externally. For backward compatibility,
5104 if the data in the cache doesn't make sense, Exim assumes it has read an
5105 old-format file, and it generates new data and writes a new file. This
5106 means that you can't go back to an older release without removing the
5109 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5110 work if there are any delivery delays because "one_time" forces the
5111 parent to be marked "delivered", so its unseen clone is never tried
5112 again. For this reason, Exim now forbids the simultaneous setting of
5115 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5116 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5117 entry if you want to know the details. The fix, however, neglected to
5118 consider the case when local delivery batching is involved. The test for
5119 "previously delivered" was not happening when checking to see if an
5120 address could be batched with a previous (undelivered) one; under
5121 certain circumstances this could lead to multiple deliveries to the same
5124 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5125 in its include files, and this causes problems building Exim.
5127 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5128 header_syntax) but Exim was just ignoring anything given after a slash.
5129 In particular, this caused confusion with an attempt to use "verify =
5130 reverse_host_lookup/defer_ok". An error is now given when options are
5131 supplied for verify items that do not have them. (Maybe reverse_host_
5132 lookup should have a defer_ok option, but that's a different point.)
5134 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5135 defined by RFC 821) to 2048, because there were problems with some AUTH
5136 commands, and RFC 1869 says the size should be increased for extended
5137 SMTP commands that take arguments.
5139 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5142 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5143 "unknown" error; now it says that the functionality isn't in the binary.
5145 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5146 an address' error message when a string expansion fails (syntax or
5147 whatever). Otherwise the password may appear in the log. Following change
5148 PH/42 below, there is no longer a chance of it appearing in a bounce
5151 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5153 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5154 of its addresses was ever tried. (Bugzilla bug #2.)
5156 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5157 the result incorrectly in the debug output. (It correctly added a newline
5158 to what was transported.)
5160 TF/01 Added $received_time.
5162 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5163 commented out examples of how to interface to a virus scanner and to
5164 SpamAssassin. Also added commented examples of av_scanner and
5165 spamd_address settings.
5167 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5168 and controls are allowed in which ACLs. There were a couple of minor
5169 errors. Some of the entries in the conditions table (which is a table of
5170 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5171 negation of where the condition IS allowed.
5173 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5175 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5176 header file does not have a version number, so I've had to invent a new
5177 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5178 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5179 radiusclient), but it was contributed by a Radius user.
5181 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5182 files or directories, for OpenSSL.
5184 PH/19 When an Exim process that is running as root has to create an Exim log
5185 file, it does so in a subprocess that runs as exim:exim so as to get the
5186 ownership right at creation (otherwise, other Exim processes might see
5187 the file with the wrong ownership). There was no test for failure of this
5188 fork() call, which would lead to the process getting stuck as it waited
5189 for a non-existent subprocess. Forks do occasionally fail when resources
5190 run out. I reviewed all the other calls to fork(); they all seem to check
5193 PH/20 When checking for unexpected SMTP input at connect time (before writing
5194 the banner), Exim was not dealing correctly with a non-positive return
5195 from the read() function. If the client had disconnected by this time,
5196 the result was a log entry for a synchronization error with an empty
5197 string after "input=" when read() returned zero. If read() returned -1
5198 (an event I could not check), uninitialized data bytes were printed.
5199 There were reports of junk text (parts of files, etc) appearing after
5202 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5204 PH/22 Added support for macro redefinition, and (re)definition in between
5205 driver and ACL definitions.
5207 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5208 forgetting to use the resulting value; it was using the unexpanded value.
5210 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5211 hadn't been configured. The fix is from Juergen Kreileder, who
5212 understands it better than I do:
5214 "Here's what I see happening with three configured cyrus_sasl
5215 authenticators configured (plain, login, cram-md5):
5217 On startup auth_cyrus_sasl_init() gets called for each of these.
5218 This means three calls to sasl_listmech() without a specified mech_list.
5219 => SASL tests which mechs of all available mechs actually work
5220 => three warnings about OTP not working
5221 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5223 With the patch, sasl_listmech() also gets called three times. But now
5224 SASL's mech_list option is set to the server_mech specified in the the
5225 authenticator. Or in other words, the answer from sasl_listmech()
5226 gets limited to just the mech you're testing for (which is different
5228 => the return list contains just 'plain' or 'login', 'cram-md5' or
5229 nothing depending on the value of ob->server_mech.
5231 I've just tested the patch: Authentication still works fine,
5232 unavailable mechs specified in the exim configuration are still
5233 caught, and the auth.log warnings about OTP are gone."
5235 PH/25 When debugging is enabled, the contents of the command line are added
5236 to the debugging output, even when log_selector=+arguments is not
5239 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5240 answer is "GNU", and only if the return is "GNU/something" is the answer
5243 PH/27 $acl_verify_message is now set immediately after the failure of a
5244 verification in an ACL, and so is available in subsequent modifiers. In
5245 particular, the message can be preserved by coding like this:
5247 warn !verify = sender
5248 set acl_m0 = $acl_verify_message
5250 Previously, $acl_verify_message was set only while expanding "message"
5251 and "log_message" when a very denied access.
5253 PH/28 Modified OS/os.c-Linux with
5255 -#ifndef OS_LOAD_AVERAGE
5256 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5258 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5259 nomenclature these days.)
5261 PH/29 Installed patch from the Sieve maintainer that adds the options
5262 sieve_useraddress and sieve_subaddress to the redirect router.
5264 PH/30 In these circumstances:
5265 . Two addresses routed to the same list of hosts;
5266 . First host does not offer TLS;
5267 . First host accepts first address;
5268 . First host gives temporary error to second address;
5269 . Second host offers TLS and a TLS session is established;
5270 . Second host accepts second address.
5271 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5272 and peerdn, if requested) that were in fact used only for the second
5275 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5276 attention to any local part prefix or suffix that was matched by the
5277 router that accepted the address. It now behaves in the same way as it
5278 does for delivery: the affixes are removed from the local part unless
5279 rcpt_include_affixes is set on the transport.
5281 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5282 timeout during the DATA phase of an incoming message.
5284 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5285 applied a patch sanctioned by the Sieve maintainer.
5287 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5288 the uid or gid is negative. A case of a negative gid caused this to be
5289 noticed. The fix allows for either to be negative.
5291 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5292 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5295 PH/36 The change PH/12 above was broken. Fixed it.
5297 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5298 the grounds that routing the same address twice would always produce the
5299 same answer. This might have been true once, but it is certainly no
5300 longer true now. Routing a child address may depend on the previous
5301 routing that produced that child. Some complicated redirection strategies
5302 went wrong when messages had multiple recipients, and made Exim's
5303 behaviour dependent on the order in which the addresses were given.
5305 I have moved the duplicate checking until after the routing is complete.
5306 Exim scans the addresses that are assigned to local and remote
5307 transports, and removes any duplicates. This means that more work will be
5308 done, as duplicates will always all be routed, but duplicates are
5309 presumably rare, so I don't expect this is of any significance.
5311 For deliveries to pipes, files, and autoreplies, the duplicate checking
5312 still happens during the routing process, since they are not going to be
5315 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5316 It corrects a timeout issue with spamd. This is Ian's comment: "The
5317 background is that sometimes spamd either never reads data from a
5318 connection it has accepted, or it never writes response data. The exiscan
5319 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5320 blindly assumes that writes won't block so it may never time out."
5322 PH/39 Allow G after quota size as well as K and M.
5324 PH/40 The value set for $authenticated_id in an authenticator may not contain
5325 binary zeroes or newlines because the value is written to log lines and
5326 to spool files. There was no check on this. Now the value is run through
5327 the string_printing() function so that such characters are converted to
5328 printable escape sequences.
5330 PH/41 $message_linecount is a new variable that contains the total number of
5331 lines in the message. Compare $body_linecount, which is the count for the
5334 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5335 bounce and delay warning messages, except in certain special cases, which
5338 (a) An SMTP error message from a remote host;
5339 (b) A message specified in a :fail: redirection;
5340 (c) A message specified in a "fail" command in a system filter;
5341 (d) A message specified in a FAIL return from the queryprogram router;
5342 (e) A message specified by the cannot_route_message router option.
5344 In these cases only, Exim does include the error details in bounce and
5345 warning messages. There are also a few cases where bland messages such
5346 as "unrouteable address" or "local delivery error" are given.
5348 PH/43 $value is now also set for the "else" part of a ${run expansion.
5350 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5351 being worked on, but at least Exim now implements the latest version to
5354 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5355 process to complete was treated as a delivery failure, a timeout while
5356 writing the message to the pipe was logged, but erroneously treated as a
5357 successful delivery. Such timeouts include transport filter timeouts. For
5358 consistency with the overall process timeout, these timeouts are now
5359 treated as errors, giving rise to delivery failures by default. However,
5360 there is now a new Boolean option for the pipe transport called
5361 timeout_defer, which, if set TRUE, converts the failures into defers for
5362 both kinds of timeout. A transport filter timeout is now identified in
5365 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5366 systems where "make" and "gmake" are different, calling "gmake" at top
5367 level broke things. I've arranged for the value of $(MAKE) to be passed
5368 from the Makefile to this script so that it can call the same version of
5372 A note about Exim versions 4.44 and 4.50
5373 ----------------------------------------
5375 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5376 changes of various kinds. As a consequence, a big documentation update was
5377 needed. This delayed the release for rather longer than seemed good, especially
5378 in the light of a couple of (minor) security issues. Therefore, the changes
5379 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5380 release. So 4.44 and 4.50 are in effect two different branches that both start
5383 I have left the 4.50 change log unchanged; it contains all the changes since
5384 4.43. The change log for 4.44 is below; many of its items are identical to
5385 those for 4.50. This seems to be the most sensible way to preserve the
5386 historical information.
5392 1. Minor wording change to the doc/README.SIEVE file.
5394 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5395 computation of the current number of files was incorrect.
5397 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5398 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5401 4. Give more explanation in the error message when the command for a transport
5402 filter fails to execute.
5404 5. There are several places where Exim runs a non-Exim command in a
5405 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5406 was being done only for the command run by the queryprogram router. It is
5407 now done for all such subprocesses. The other cases are: ${run, transport
5408 filters, and the commands run by the lmtp and pipe transports.
5410 6. Added CONFIGURE_GROUP build-time option.
5412 7. Some older OS have a limit of 256 on the maximum number of file
5413 descriptors. Exim was using setrlimit() to set 1000 as a large value
5414 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5415 systems. I've change it so that if it can't get 1000, it tries for 256.
5417 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5418 was an oversight, and furthermore, ever since the addition of extra
5419 controls (e.g. 4.43/32), the checks on when to allow different forms of
5420 "control" were broken. There should now be diagnostics for all cases when a
5421 control that does not make sense is encountered.
5423 9. Added the /retain_sender option to "control=submission".
5425 10. $recipients is now available in the predata ACL (oversight).
5427 11. Tidy the search cache before the fork to do a delivery from a message
5428 received from the command line. Otherwise the child will trigger a lookup
5429 failure and thereby defer the delivery if it tries to use (for example) a
5430 cached ldap connection that the parent has called unbind on.
5432 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5433 of $address_data from the recipient verification was clobbered by the
5434 sender verification.
5436 13. The value of address_data from a sender verification is now available in
5437 $sender_address_data in subsequent conditions in the ACL statement.
5439 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5441 15. Added a new option "connect=<time>" to callout options, to set a different
5444 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5445 was its contents. (It was OK if the option was not defined at all.)
5447 17. A "Completed" log line is now written for messages that are removed from
5448 the spool by the -Mrm option.
5450 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5451 information about exactly what failed.
5453 19. Added -dd to debug only the daemon process.
5455 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5456 handles timeouts, both on the server side and network timeouts. Renamed the
5457 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5459 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5460 It is now set to "smtps".
5462 22. $host_address is now set to the target address during the checking of
5463 ignore_target_hosts.
5465 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5466 being passed; this would have caused $sender_host_name to have been used if
5467 matching the list had actually called for a host name (not very likely,
5468 since this list is usually IP addresses). A host name is now passed as
5471 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5472 code that specifies a non-restarting handler (typically sigaction() in
5473 modern systems) in an attempt to fix a rare and obscure crash bug.
5475 25. Narrowed the window for a race in the daemon that could cause it to ignore
5476 SIGCHLD signals. This is not a major problem, because they are used only to
5477 wake it up if nothing else does.
5479 26. A malformed maildirsize file could cause Exim to calculate negative values
5480 for the mailbox size or file count. Odd effects could occur as a result.
5481 The maildirsize information is now recalculated if the size or filecount
5484 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5485 support for a long time. Removed HAVE_SYS_VFS_H.
5487 28. Installed the latest version of exipick from John Jetmore.
5489 29. In an address list, if the pattern was not a regular expression, an empty
5490 subject address (from a bounce message) matched only if the pattern was an
5491 empty string. Non-empty patterns were not even tested. This was the wrong
5492 because it is perfectly reasonable to use an empty address as part of a
5493 database query. An empty address is now tested by patterns that are
5494 lookups. However, all the other forms of pattern expect the subject to
5495 contain a local part and a domain, and therefore, for them, an empty
5496 address still always fails if the pattern is not itself empty.
5498 30. Exim went into a mad DNS loop when attempting to do a callout where the
5499 host was specified on an smtp transport, and looking it up yielded more
5500 than one IP address.
5502 31. Re-factored the code for checking spool and log partition space into a
5503 function that finds that data and another that does the check. The former
5504 is then used to implement four new variables: $spool_space, $log_space,
5505 $spool_inodes, and $log_inodes.
5507 32. The RFC2047 encoding function was originally intended for short strings
5508 such as real names; it was not keeping to the 75-character limit for
5509 encoded words that the RFC imposes. It now respects the limit, and
5510 generates multiple encoded words if necessary. To be on the safe side, I
5511 have increased the buffer size for the ${rfc2047: expansion operator from
5514 33. It is now permitted to omit both strings after an "if" condition; if the
5515 condition is true, the result is "true". As before, when the second string
5516 is omitted, a false condition yields an empty string. This makes it less
5517 cumbersome to write custom ACL and router conditions.
5519 34. Failure to deliver a bounce message always caused it to be frozen, even if
5520 there was an errors_to setting on the router. The errors_to setting is now
5523 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5524 canonical form (fully expanded) before being placed in
5525 $sender_host_address.
5527 36. The table in the code that translates DNS record types into text (T_A to
5528 "A" for instance) was missing entries for NS and CNAME. It is just possible
5529 that this could have caused confusion if both these types were looked up
5530 for the same domain, because the text type is used as part of Exim's
5531 per-process caching. But the chance of anyone hitting this buglet seems
5534 37. The dnsdb lookup has been extended in a number of ways.
5536 (1) There is a new type, "zns", which walks up the domain tree until it
5537 finds some nameserver records. It should be used with care.
5539 (2) There is a new type, "mxh", which is like "mx" except that it returns
5540 just the host names, not the priorities.
5542 (3) It is now possible to give a list of domains (or IP addresses) to be
5543 looked up. The behaviour when one of the lookups defers can be
5544 controlled by a keyword.
5546 (4) It is now possible to specify the separator character for use when
5547 multiple records are returned.
5549 38. The dnslists ACL condition has been extended: it is now possible to supply
5550 a list of IP addresses and/or domains to be looked up in a particular DNS
5553 39. Added log_selector=+queue_time_overall.
5555 40. When running the queue in the test harness, wait just a tad after forking a
5556 delivery process, to get repeatability of debugging output.
5558 41. Include certificate and key file names in error message when GnuTLS fails
5559 to set them up, because the GnuTLS error message doesn't include the name
5560 of the failing file when there is a problem reading it.
5562 42. Allow both -bf and -bF in the same test run.
5564 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5566 44. The "Exiscan patch" is now merged into the mainline Exim source.
5568 45. Sometimes the final signoff response after QUIT could fail to get
5569 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5570 < 0 before doing a fflush(). This bug looks as though it goes back to the
5571 introduction of TLS in release 3.20, but "sometimes" must have been rare
5572 because the tests only now provoked it.
5574 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5575 (this can affect the format of dates).
5577 47. exim_tidydb, when checking for the continued existence of a message for
5578 which it has found a message-specific retry record, was not finding
5579 messages that were in split spool directories. Consequently, it was
5580 deleting retry records that should have stayed in existence.
5582 48. Steve fixed some bugs in eximstats.
5584 49. The SPA authentication driver was not abandoning authentication and moving
5585 on to the next authenticator when an expansion was forced to fail,
5586 contradicting the general specification for all authenticators. Instead it
5587 was generating a temporary error. It now behaves as specified.
5589 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5590 (the order specifies the preference for clients). The order is now AES256,
5591 AES128, 3DES, ARCFOUR128.
5593 51. Small patch to Sieve code - explicitly set From: when generating an
5596 52. Exim crashed if a remote delivery caused a very long error message to be
5597 recorded - for instance if somebody sent an entire SpamAssassin report back
5598 as a large number of 550 error lines. This bug was coincidentally fixed by
5599 increasing the size of one of Exim's internal buffers (big_buffer) that
5600 happened as part of the Exiscan merge. However, to be on the safe side, I
5601 have made the code more robust (and fixed the comments that describe what
5604 53. Now that there can be additional text after "Completed" in log lines (if
5605 the queue_time_overall log selector is set), a one-byte patch to exigrep
5606 was needed to allow it to recognize "Completed" as not the last thing in
5609 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5610 patch that reportedly fixes this has been added. I am not expert enough to
5611 create a test for it. This is what the patch creator wrote:
5613 "I found a little strange behaviour of ldap code when working with
5614 Windows 2003 AD Domain, where users was placed in more than one
5615 Organization Units. When I tried to give exim partial DN, the exit code
5616 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5617 But simultaneously result of request was absolutely normal ldap result,
5618 so I produce this patch..."
5620 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5621 REFERENCE, so I have modified the code to exclude the patch when that macro
5624 55. Some experimental protocols are using DNS PTR records for new purposes. The
5625 keys for these records are domain names, not reversed IP addresses. The
5626 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5627 leaves it alone. Component reversal etc. now happens only for IP addresses.
5630 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5632 57. Double the size of the debug message buffer (to 2048) so that more of very
5633 long debug lines gets shown.
5635 58. The exicyclog utility now does better if the number of log files to keep
5636 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5638 59. Two changes related to the smtp_active_hostname option:
5640 (1) $smtp_active_hostname is now available as a variable.
5641 (2) The default for smtp_banner uses $smtp_active_hostname instead
5642 of $primary_hostname.
5644 60. The host_aton() function is supposed to be passed a string that is known
5645 to be a valid IP address. However, in the case of IPv6 addresses, it was
5646 not checking this. This is a hostage to fortune. Exim now panics and dies
5647 if the condition is not met. A case was found where this could be provoked
5648 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5649 components; fortuitously, this particular loophole had already been fixed
5650 by change 4.50/55 above.
5652 If there are any other similar loopholes, the new check in host_aton()
5653 itself should stop them being exploited. The report I received stated that
5654 data on the command line could provoke the exploit when Exim was running as
5655 exim, but did not say which command line option was involved. All I could
5656 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5657 running as the user.
5660 61. There was a buffer overflow vulnerability in the SPA authentication code
5661 (which came originally from the Samba project). I have added a test to the
5662 spa_base64_to_bits() function which I hope fixes it.
5665 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5666 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5668 63. The daemon start-up calls getloadavg() while still root for those OS that
5669 need the first call to be done as root, but it missed one case: when
5670 deliver_queue_load_max is set with deliver_drop_privilege. This is
5671 necessary for the benefit of the queue runner, because there is no re-exec
5672 when deliver_drop_privilege is set.
5674 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5675 This has been fixed.
5677 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5678 was in use, was not putting the data itself into the right store pool;
5679 consequently, it could be overwritten for a subsequent message in the same
5680 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5683 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5685 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5686 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5687 the function were treating the return as a boolean value, which happened to
5688 work because 0=false and not-0=true, but is not correct code.
5690 68. The host_aton() function was not handling scoped IPv6 addresses (those
5691 with, for example, "%eth0" on the end) correctly.
5693 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5694 negated items (that is, ~something) in unsigned ints. Some compilers
5695 apparently mutter when there is no cast.
5697 70. If an address verification called from an ACL failed, and did not produce a
5698 user-specific message (i.e. there was only a "system" message), nothing was
5699 put in $acl_verify_message. In this situation, it now puts the system
5702 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5703 session; change 4.31/43 added the unwanted input to the log line - except
5704 that it did not do this in the start of session case. It now does.
5706 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5707 This is wrong and can cause the other end to generate a synchronization
5708 error if it is another Exim or anything else that does the synchronization
5709 check. A QUIT command is no longer sent after a timeout.
5711 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5712 during host lookups.
5714 74. The defer_ok option of callout verification was not working if it was used
5715 when verifying addresses in header lines, that is, for this case:
5717 verify = header_sender/callout=defer_ok
5719 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5720 those file descriptors could be used for SMTP connections. If anything
5721 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5722 could be sent to the SMTP client, causing chaos. The daemon now opens
5723 stdin, stdout, and stderr to /dev/null when it puts itself into the
5726 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5727 log by default. The user can override this with suitable Perl magic.
5729 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5730 the log message when discard triggers, was not working for the DATA ACL or
5731 for the non-SMTP ACL.
5733 78. Error message wording change in sieve.c.
5735 79. If smtp_accept_max_per_host was set, the number of connections could be
5736 restricted to fewer than expected, because the daemon was trying to set up
5737 a new connection before checking whether the processes handling previous
5738 connections had finished. The check for completed processes is now done
5739 earlier. On busy systems, this bug wouldn't be noticed because something
5740 else would have woken the daemon, and it would have reaped the completed
5743 80. If a message was submitted locally by a user whose login name contained one
5744 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5745 It caused a spool format error. I have fixed the spool reading code. A
5746 related problem was that the "from" clause in the Received: line became
5747 illegal because of the space(s). It is now covered by ${quote_local_part.
5749 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5752 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5754 83. Patch from Sieve maintainer to fix unterminated string problem in
5755 "vacation" handling.
5757 84. Some minor changes to the Linux configuration files to help with other
5758 OS variants using glibc.
5760 85. One more patch for Sieve to update vacation handling to latest spec.
5763 ----------------------------------------------------
5764 See the note above about the 4.44 and 4.50 releases.
5765 ----------------------------------------------------
5771 1. Change 4.43/35 introduced a bug that caused file counts to be
5772 incorrectly computed when quota_filecount was set in an appendfile
5775 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5776 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5779 3. Give more explanation in the error message when the command for a transport
5780 filter fails to execute.
5782 4. There are several places where Exim runs a non-Exim command in a
5783 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5784 was being done only for the command run by the queryprogram router. It is
5785 now done for all such subprocesses. The other cases are: ${run, transport
5786 filters, and the commands run by the lmtp and pipe transports.
5788 5. Some older OS have a limit of 256 on the maximum number of file
5789 descriptors. Exim was using setrlimit() to set 1000 as a large value
5790 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5791 systems. I've change it so that if it can't get 1000, it tries for 256.
5793 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5794 was an oversight, and furthermore, ever since the addition of extra
5795 controls (e.g. 4.43/32), the checks on when to allow different forms of
5796 "control" were broken. There should now be diagnostics for all cases when a
5797 control that does not make sense is encountered.
5799 7. $recipients is now available in the predata ACL (oversight).
5801 8. Tidy the search cache before the fork to do a delivery from a message
5802 received from the command line. Otherwise the child will trigger a lookup
5803 failure and thereby defer the delivery if it tries to use (for example) a
5804 cached ldap connection that the parent has called unbind on.
5806 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5807 of $address_data from the recipient verification was clobbered by the
5808 sender verification.
5810 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5811 was its contents. (It was OK if the option was not defined at all.)
5813 11. A "Completed" log line is now written for messages that are removed from
5814 the spool by the -Mrm option.
5816 12. $host_address is now set to the target address during the checking of
5817 ignore_target_hosts.
5819 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5820 being passed; this would have caused $sender_host_name to have been used if
5821 matching the list had actually called for a host name (not very likely,
5822 since this list is usually IP addresses). A host name is now passed as
5825 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5826 code that specifies a non-restarting handler (typically sigaction() in
5827 modern systems) in an attempt to fix a rare and obscure crash bug.
5829 15. Narrowed the window for a race in the daemon that could cause it to ignore
5830 SIGCHLD signals. This is not a major problem, because they are used only to
5831 wake it up if nothing else does.
5833 16. A malformed maildirsize file could cause Exim to calculate negative values
5834 for the mailbox size or file count. Odd effects could occur as a result.
5835 The maildirsize information is now recalculated if the size or filecount
5838 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5839 support for a long time. Removed HAVE_SYS_VFS_H.
5841 18. Updated exipick to current release from John Jetmore.
5843 19. Allow an empty sender to be matched against a lookup in an address list.
5844 Previously the only cases considered were a regular expression, or an
5847 20. Exim went into a mad DNS lookup loop when doing a callout where the
5848 host was specified on the transport, if the DNS lookup yielded more than
5851 21. The RFC2047 encoding function was originally intended for short strings
5852 such as real names; it was not keeping to the 75-character limit for
5853 encoded words that the RFC imposes. It now respects the limit, and
5854 generates multiple encoded words if necessary. To be on the safe side, I
5855 have increased the buffer size for the ${rfc2047: expansion operator from
5858 22. Failure to deliver a bounce message always caused it to be frozen, even if
5859 there was an errors_to setting on the router. The errors_to setting is now
5862 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5863 canonical form (fully expanded) before being placed in
5864 $sender_host_address.
5866 24. Updated eximstats to version 1.33
5868 25. Include certificate and key file names in error message when GnuTLS fails
5869 to set them up, because the GnuTLS error message doesn't include the name
5870 of the failing file when there is a problem reading it.
5872 26. Expand error message when OpenSSL has problems setting up cert/key files.
5875 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5876 (this can affect the format of dates).
5878 28. exim_tidydb, when checking for the continued existence of a message for
5879 which it has found a message-specific retry record, was not finding
5880 messages that were in split spool directories. Consequently, it was
5881 deleting retry records that should have stayed in existence.
5883 29. eximstats updated to version 1.35
5884 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5885 1.35 - bugfix such that pie charts by volume are generated correctly
5887 30. The SPA authentication driver was not abandoning authentication and moving
5888 on to the next authenticator when an expansion was forced to fail,
5889 contradicting the general specification for all authenticators. Instead it
5890 was generating a temporary error. It now behaves as specified.
5892 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5893 (the order specifies the preference for clients). The order is now AES256,
5894 AES128, 3DES, ARCFOUR128.
5896 31. Small patch to Sieve code - explicitly set From: when generating an
5899 32. Exim crashed if a remote delivery caused a very long error message to be
5900 recorded - for instance if somebody sent an entire SpamAssassin report back
5901 as a large number of 550 error lines. This bug was coincidentally fixed by
5902 increasing the size of one of Exim's internal buffers (big_buffer) that
5903 happened as part of the Exiscan merge. However, to be on the safe side, I
5904 have made the code more robust (and fixed the comments that describe what
5907 33. Some experimental protocols are using DNS PTR records for new purposes. The
5908 keys for these records are domain names, not reversed IP addresses. The
5909 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5910 leaves it alone. Component reversal etc. now happens only for IP addresses.
5913 34. The host_aton() function is supposed to be passed a string that is known
5914 to be a valid IP address. However, in the case of IPv6 addresses, it was
5915 not checking this. This is a hostage to fortune. Exim now panics and dies
5916 if the condition is not met. A case was found where this could be provoked
5917 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5918 components; fortuitously, this particular loophole had already been fixed
5919 by change 4.50/55 or 4.44/33 above.
5921 If there are any other similar loopholes, the new check in host_aton()
5922 itself should stop them being exploited. The report I received stated that
5923 data on the command line could provoke the exploit when Exim was running as
5924 exim, but did not say which command line option was involved. All I could
5925 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5926 running as the user.
5929 35. There was a buffer overflow vulnerability in the SPA authentication code
5930 (which came originally from the Samba project). I have added a test to the
5931 spa_base64_to_bits() function which I hope fixes it.
5934 36. The daemon start-up calls getloadavg() while still root for those OS that
5935 need the first call to be done as root, but it missed one case: when
5936 deliver_queue_load_max is set with deliver_drop_privilege. This is
5937 necessary for the benefit of the queue runner, because there is no re-exec
5938 when deliver_drop_privilege is set.
5940 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5941 was in use, was not putting the data itself into the right store pool;
5942 consequently, it could be overwritten for a subsequent message in the same
5943 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5946 38. Sometimes the final signoff response after QUIT could fail to get
5947 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5948 < 0 before doing a fflush(). This bug looks as though it goes back to the
5949 introduction of TLS in release 3.20, but "sometimes" must have been rare
5950 because the tests only now provoked it.
5956 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5957 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5958 Now it returns an integer. A number of calls were still expecting a T/F
5959 return. Fortuitously, in all cases, the tests worked in OK situations,
5960 which is the norm. However, things would have gone wrong on any write
5961 failures on the smtp file descriptor. This function is used when sending
5962 messages over SMTP and also when doing verify callouts.
5964 2. When Exim is called to do synchronous delivery of a locally submitted
5965 message (the -odf or -odi options), it no longer closes stderr before doing
5968 3. Implemented the mua_wrapper option.
5970 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5972 5. Implemented the functions header_remove(), header_testname(),
5973 header_add_at_position(), and receive_remove_recipient(), and exported them
5976 6. If an ACL "warn" statement specified the addition of headers, Exim already
5977 inserted X-ACL-Warn: at the start if there was no header name. However, it
5978 was not making this test for the second and subsequent header lines if
5979 there were newlines in the string. This meant that an invalid header could
5980 be inserted if Exim was badly configured.
5982 7. Allow an ACL "warn" statement to add header lines at the start or after all
5983 the Received: headers, as well as at the end.
5985 8. Added the rcpt_4xx retry error code.
5987 9. Added postmaster_mailfrom=xxx to callout verification option.
5989 10. Added mailfrom=xxxx to the callout verification option, for verify=
5992 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5993 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5995 12. Inserted some casts to stop certain compilers warning when using pointer
5996 differences as field lengths or precisions in printf-type calls (mostly
5997 affecting debugging statements).
5999 13. Added optional readline() support for -be (dynamically loaded).
6001 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6002 same clock tick as a message's arrival, so that its received time was the
6003 same as the "first fail" time on the retry record, and that message
6004 remained on the queue past the ultimate address timeout, every queue runner
6005 would try a delivery (because it was past the ultimate address timeout) but
6006 after another failure, the ultimate address timeout, which should have then
6007 bounced the address, did not kick in. This was a "< instead of <=" error;
6008 in most cases the first failure would have been in the next clock tick
6009 after the received time, and all would be well.
6011 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6012 being recognized when the domain list was tested by the match_domain
6013 condition in an expansion string.
6015 16. Added the ${str2b64: operator.
6017 17. Exim was always calling setrlimit() to set a large limit for the number of
6018 processes, without checking whether the existing limit was already
6019 adequate. (It did check for the limit on file descriptors.) Furthermore,
6020 errors from getrlimit() and setrlimit() were being ignored. Now they are
6021 logged to the main and panic logs, but Exim does carry on, to try to do its
6022 job under whatever limits there are.
6024 18. Imported PCRE 5.0.
6026 19. Trivial typo in log message " temporarily refused connection" (the leading
6029 20. If the log selector return_path_on_delivery was set and an address was
6030 redirected to /dev/null, the delivery process crashed because it assumed
6031 that a return path would always be set for a "successful" delivery. In this
6032 case, the whole delivery is bypassed as an optimization, and therefore no
6035 21. Internal re-arrangement: the function for sending a challenge and reading
6036 a response while authentication was assuming a zero-terminated challenge
6037 string. It's now changed to take a pointer and a length, to allow for
6038 binary data in such strings.
6040 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6042 23. Exim was not respecting finduser_retries when seeking the login of the
6043 uid under which it was called; it was always trying 10 times. (The default
6044 setting of finduser_retries is zero.) Also, it was sleeping after the final
6045 failure, which is pointless.
6047 24. Implemented tls_on_connect_ports.
6049 25. Implemented acl_smtp_predata.
6051 26. If the domain in control=submission is set empty, Exim assumes that the
6052 authenticated id is a complete email address when it generates From: or
6053 Sender: header lines.
6055 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6056 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6057 chown and chgrp in /bin and hostname in /usr/bin.
6059 28. Exim was keeping the "process log" file open after each use, just as it
6060 does for the main log. This opens the possibility of it remaining open for
6061 long periods when the USR1 signal hits a daemon. Occasional processlog
6062 errors were reported, that could have been caused by this. Anyway, it seems
6063 much more sensible not to leave this file open at all, so that is what now
6066 29. The long-running daemon process does not normally write to the log once it
6067 has entered its main loop, and it closes the log before doing so. This is
6068 so that log files can straightforwardly be renamed and moved. However,
6069 there are a couple of unusual error situations where the daemon does write
6070 log entries, and I had neglected to close the log afterwards.
6072 30. The text of an SMTP error response that was received during a remote
6073 delivery was being truncated at 512 bytes. This is too short for some of
6074 the long messages that one sometimes sees. I've increased the limit to
6077 31. It is now possible to make retry rules that apply only when a message has a
6078 specific sender, in particular, an empty sender.
6080 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6081 it possible to be selective about when SMTP synchronization is enforced.
6083 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6085 32. Implemented hosts_connection_nolog.
6087 33. Added an ACL for QUIT.
6089 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6092 35. Added mailbox_size and mailbox_filecount to appendfile.
6094 36. Added control = no_multiline_responses to ACLs.
6096 37. There was a bug in the logic of the code that waits for the clock to tick
6097 in the case where the clock went backwards by a substantial amount such
6098 that the microsecond fraction of "now" was more than the microsecond
6099 fraction of "then" (but the whole seconds number was less).
6101 38. Added support for the libradius Radius client library this is found on
6102 FreeBSD (previously only the radiusclient library was supported).
6108 1. When certain lookups returned multiple values in the form name=value, the
6109 quoting of the values was not always being done properly. Specifically:
6110 (a) If the value started with a double quote, but contained no whitespace,
6112 (b) If the value contained whitespace other than a space character (i.e.
6113 tabs or newlines or carriage returns) it was not quoted.
6114 This fix has been applied to the mysql and pgsql lookups by writing a
6115 separate quoting function and calling it from the lookup code. The fix
6116 should probably also be applied to nisplus, ibase and oracle lookups, but
6117 since I cannot test any of those, I have not disturbed their existing code.
6119 2. A hit in the callout cache for a specific address caused a log line with no
6120 reason for rejecting RCPT. Now it says "Previous (cached) callout
6121 verification failure".
6123 3. There was an off-by-one bug in the queryprogram router. An over-long
6124 return line was truncated at 256 instead of 255 characters, thereby
6125 overflowing its buffer with the terminating zero. As well as fixing this, I
6126 have increased the buffer size to 1024 (and made a note to document this).
6128 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6129 when Exim is waiting for an SMTP response from a remote server, Exim
6130 restarts its select() call on the socket, thereby resetting its timeout.
6131 This is not a problem when such interrupts are rare. Somebody set up a cron
6132 job to run exiwhat every 2 minutes, which is less than the normal select()
6133 timeout (5 or 10 minutes). This meant that the select() timeout never
6134 kicked in because it was always reset. I have fixed this by comparing the
6135 time when an interrupt arrives with the time at the start of the first call
6136 to select(). If more time than the timeout has elapsed, the interrupt is
6137 treated as a timeout.
6139 5. Some internal re-factoring in preparation for the addition of Sieve
6140 extensions (by MH). In particular, the "personal" test is moved to a
6141 separate function, and given an option for scanning Cc: and Bcc: (which is
6142 not set for Exim filters).
6144 6. When Exim created an email address using the login of the caller as the
6145 local part (e.g. when creating a From: or Sender: header line), it was not
6146 quoting the local part when it contained special characters such as @.
6148 7. Installed new OpenBSD configuration files.
6150 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6151 try to make them clearer.
6153 9. Callout options, other than the timeout value, were being ignored when
6154 verifying sender addresses in header lines. For example, when using
6156 verify = header_sender/callout=no_cache
6158 the cache was (incorrectly) being used.
6160 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6161 only the Cygwin environment.
6163 11. When return_path_on_delivery was set as a log selector, if different remote
6164 addresses in the same message used different return paths and parallel
6165 remote delivery occurred, the wrong values would sometimes be logged.
6166 (Whenever a remote delivery process finished, the return path value from
6167 the most recently started remote delivery process was logged.)
6169 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6170 lines when AUTH and/or TLS are in use. This is the "received protocol"
6171 field. Exim used to use "asmtp" for authenticated SMTP, without any
6172 indication (in the protocol name) for TLS use. Now it follows the RFC and
6173 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6174 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6175 names appear in log lines as well as in Received: header lines.
6177 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6178 extensions, and comparison tests, and to fix some bugs.
6180 14. Changes to the "personal" filter test:
6182 (1) The test was buggy in that it was just doing the equivalent of
6183 "contains" tests on header lines. For example, if a user's address was
6184 anne@some.where, the "personal" test would incorrectly be true for
6186 To: susanne@some.where
6188 This test is now done by extracting each address from the header in turn,
6189 and checking the entire address. Other tests that are part of "personal"
6190 are now done using regular expressions (for example, to check local parts
6191 of addresses in From: header lines).
6193 (2) The list of non-personal local parts in From: addresses has been
6194 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6195 taken from the Sieve specification recommendations.
6197 (3) If the message contains any header line starting with "List-" it is
6198 treated as non-personal.
6200 (4) The test for "circular" in the Subject: header line has been removed
6201 because it now seems ill-conceived.
6203 15. Minor typos in src/EDITME comments corrected.
6205 16. Installed latest exipick from John Jetmore.
6207 17. If headers_add on a router specified a text string that was too long for
6208 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6209 of string_sprintf() is now avoided.
6211 18. $message_body_size was not set (it was always zero) when running the DATA
6212 ACL and the local_scan() function.
6214 19. For the "mail" command in an Exim filter, no default was being set for
6215 the once_repeat time, causing a random time value to be used if "once" was
6216 specified. (If the value happened to be <= 0, no repeat happened.) The
6217 default is now 0s, meaning "never repeat". The "vacation" command was OK
6218 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6219 (I found it when inspecting the code).
6221 20. There is now an overall timeout for performing a callout verification. It
6222 defaults to 4 times the callout timeout, which applies to individual SMTP
6223 commands during the callout. The overall timeout applies when there is more
6224 than one host that can be tried. The timeout is checked before trying the
6225 next host. This prevents very long delays if there are a large number of
6226 hosts and all are timing out (e.g. when the network connections are timing
6227 out). The value of the overall timeout can be changed by specifying an
6228 additional sub-option for "callout", called "maxwait". For example:
6230 verify = sender/callout=5s,maxwait=20s
6232 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6233 to the end before writing, but this should make it even safer).
6235 22. Exim was forgetting that it had advertised PIPELINING for the second and
6236 subsequent messages on an SMTP connection. It was also not resetting its
6237 memory on STARTTLS and an internal HELO.
6239 23. When Exim logs an SMTP synchronization error within a session, it now
6240 records whether PIPELINING has been advertised or not.
6242 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6243 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6244 rather than long int.
6246 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6248 26. Added the never_mail option to autoreply.
6254 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6255 crash if the getsockname() call failed; this can happen if a connection is
6256 closed very soon after it is established. The problem was simply in the
6257 order in which certain operations were done, causing Exim to try to write
6258 to the SMTP stream before it had set up the file descriptor. The bug has
6259 been fixed by making things happen in the correct order.
6265 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6266 before the connection was closed, thus losing the rejection response.
6268 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6269 some early Solaris releases, but causes trouble in current releases where
6270 socklen_t is defined.
6272 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6275 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6278 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6279 that releasing the top of it at the end releases what was used for sub-
6280 expansions (unless the block got too big). However, discard this block if
6281 the first thing is a variable or header, so that we can use its block when
6282 it is dynamic (useful for very large $message_headers, for example).
6284 6. Lookups now cache *every* query, not just the most recent. A new, separate
6285 store pool is used for this. It can be recovered when all lookup caches are
6286 flushed. Lookups now release memory at the end of their result strings.
6287 This has involved some general refactoring of the lookup sources.
6289 7. Some code has been added to the store_xxx() functions to reduce the amount
6290 of flapping under certain conditions.
6292 8. log_incoming_interface used to affect only the <= reception log lines. Now
6293 it causes the local interface and port to be added to several more SMTP log
6294 lines, for example "SMTP connection from", and rejection lines.
6296 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6298 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6300 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6301 could be overwritten at the end of the current message (or the start of a
6302 new message if it was set in a HELO ACL). The value is now preserved for
6303 the duration of the SMTP connection.
6305 12. If a transport had a headers_rewrite setting, and a matching header line
6306 contained an unqualified address, that address was qualified, even if it
6307 did not match any rewriting rules. The underlying bug was that the values
6308 of the flags that permit the existence of unqualified sender and recipient
6309 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6310 non-local messages, and by -bnq for local messages) were not being
6311 preserved with the message after it was received.
6313 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6314 "next input=" as part of the text comprising the host identity instead of
6315 the correct text. The code was using the same buffer for two different
6316 strings. However, depending on which order the printing function evaluated
6317 its arguments, the bug did not always show up. Under Linux, for example, my
6318 test suite worked just fine.
6320 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6321 doesn't work with some older versions of Perl. It has been changed to "my",
6322 which in any case is probably the better facility to use.
6324 15. A really picky compiler found some instances of statements for creating
6325 error messages that either had too many or two few arguments for the format
6328 16. The size of the buffer for calls to the DNS resolver has been increased
6329 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6330 for addresses that have a lot of PTR records. This alleviates a problem; it
6331 does not fully solve it.
6333 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6334 buffer now truncates the list and logs the incident, which is the same
6335 action as happens when Exim is looking up a host name and its aliases.
6336 Previously in this situation something unpredictable would happen;
6337 sometimes it was "internal error: store_reset failed".
6339 18. If a server dropped the connection unexpectedly when an Exim client was
6340 using GnuTLS and trying to read a response, the client delivery process
6341 crashed while trying to generate an error log message.
6343 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6344 string, for example:
6346 warn message = H1: something\nH2: something
6348 the text was added as a single header line from Exim's point of view
6349 though it ended up OK in the delivered message. However, searching for the
6350 second and subsequent header lines using $h_h2: did not work. This has been
6351 fixed. Similarly, if a system filter added multiple headers in this way,
6352 the routers could not see them.
6354 20. Expanded the error message when iplsearch is called with an invalid key to
6355 suggest using net-iplsearch in a host list.
6357 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6358 ACLs are no longer actually imposed (and a message to that effect is
6361 22. If a "gecos" field in a passwd entry contained escaped characters, in
6362 particular, if it contained a \" sequence, Exim got it wrong when building
6363 a From: or a Sender: header from that name. A second bug also caused
6364 incorrect handling when an unquoted " was present following a character
6365 that needed quoting.
6367 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6368 was not being matched caselessly.
6370 24. Arranged for all hyphens in the exim.8 source to be escaped with
6373 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6374 properly. Recipient callout cache records were still being keyed to include
6375 the sender, even when use_sender was set false. This led to far more
6376 callouts that were necessary. The sender is no longer included in the key
6377 when use_sender is false.
6379 26. Added "control = submission" modifier to ACLs.
6381 27. Added the ${base62d: operator to decode base 62 numbers.
6383 28. dnsdb lookups can now access SRV records.
6385 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6386 the configuration file.
6388 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6389 (-v) mode. This makes the output for a verbose queue run more intelligible.
6391 31. Added a use_postmaster feature to recipient callouts.
6393 32. Added the $body_zerocount variable, containing the number of binary zero
6394 bytes in the message body.
6396 33. The time of last modification of the "new" subdirectory is now used as the
6397 "mailbox time last read" when there is a quota error for a maildir
6400 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6402 35. Added +ignore_unknown as a special item in host lists.
6404 36. Code for decoding IPv6 addresses in host lists is now included, even if
6405 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6406 address was recognized as an IP address, but was then not correctly decoded
6407 into binary, causing unexpected and incorrect effects when compared with
6414 1. Very minor rewording of debugging text in manualroute to say "list of
6415 hosts" instead of "hostlist".
6417 2. If verify=header_syntax was set, and a header line with an unqualified
6418 address (no domain) and a large number of spaces between the end of the
6419 name and the colon was received, the reception process suffered a buffer
6420 overflow, and (when I tested it) crashed. This was caused by some obsolete
6421 code that should have been removed. The fix is to remove it!
6423 3. When running in the test harness, delay a bit after writing a bounce
6424 message to get a bit more predictability in the log output.
6426 4. Added a call to search_tidyup() just before forking a reception process. In
6427 theory, someone could use a lookup in the expansion of smtp_accept_max_
6428 per_host which, without the tidyup, could leave open a database connection.
6430 5. Added the variables $recipient_data and $sender_data which get set from a
6431 lookup success in an ACL "recipients" or "senders" condition, or a router
6432 "senders" option, similar to $domain_data and $local_part_data.
6434 6. Moved the writing of debug_print from before to after the "senders" test
6437 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6438 problems for message scanning, either using a data ACL, or using
6439 local_scan() because the Received: header was not generated till after they
6440 were called (in order to set the time as the time of reception completion).
6441 I have revised the way this works. The header is now generated after the
6442 body is received, but before the ACL or local_scan() are called. After they
6443 are run, the timestamp in the header is updated.
6449 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6450 before starting a queue runner without re-exec. This happened only when
6451 deliver_drop_privilege was set or when the Exim user was set to root. The
6452 effect of the bug was that timeouts during subsequent deliveries caused
6453 crashes instead of being properly handled. The handler is now left at its
6454 default (and expected) setting.
6456 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6457 message, again when deliver_drop_privilege is set or Exim is run as root.
6458 The bug described in (1) was not present in this case, but the tidying up
6459 of the other signals was missing. I have made the two cases consistent.
6461 3. The ignore_target_hosts setting on a manualroute router was being ignored
6462 for hosts that were looked up using the /MX notation.
6464 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6467 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6468 operated on the sender address. After changing the $sender_address to <>
6469 for the sender address verify, Exim was re-instated it as the original
6470 (before rewriting) address, but remembering that it had rewritten it, so it
6471 wasn't rewriting it again. This bug also had the effect of breaking the
6472 sender address verification caching when the sender address was rewritten.
6474 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6475 This has been changed so that if the ip literal address matches
6476 ignore_target_hosts, the router declines.
6478 7. Added expansion conditions match_domain, match_address, and match_local_
6479 part (NOT match_host).
6481 8. The placeholder for the Received: header didn't have a length field set.
6483 9. Added code to Exim itself and to exim_lock to test for a specific race
6484 condition that could lead to file corruption when using MBX delivery. The
6485 issue is with the lockfile that is created in /tmp. If this file is removed
6486 after a process has opened it but before that process has acquired a lock,
6487 there is the potential for a second process to recreate the file and also
6488 acquire a lock. This could lead to two Exim processes writing to the file
6489 at the same time. The added code performs the same test as UW imapd; it
6490 checks after acquiring the lock that its file descriptor still refers to
6491 the same named file.
6493 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6494 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6497 11. Added the smtp_active_hostname option. If used, this will typically be made
6498 to depend on the incoming interface address. Because $interface_address is
6499 not set up until the daemon has forked a reception process, error responses
6500 that can happen earlier (such as "too many connections") no longer contain
6503 12. If an expansion in a condition on a "warn" statement fails because a lookup
6504 defers, the "warn" statement is abandoned, and the next ACL statement is
6505 processed. Previously this caused the whole ACL to be aborted.
6507 13. Added the iplsearch lookup type.
6509 14. Added ident_timeout as a log selector.
6511 15. Added tls_certificate_verified as a log selector.
6513 16. Added a global option tls_require_ciphers (compare the smtp transport
6514 option of the same name). This controls incoming TLS connections.
6516 17. I finally figured out how to make tls_require_ciphers do a similar thing
6517 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6518 before starting the TLS session.
6520 18. Tabs are now shown as \t in -bP output.
6522 19. If the log selector return_path_on_delivery was set, Exim crashed when
6523 bouncing a message because it had too many Received: header lines.
6525 20. If two routers both had headers_remove settings, and the first one included
6526 a superfluous trailing colon, the final name in the first list and the
6527 first name in the second list were incorrectly joined into one item (with a
6528 colon in the middle).
6534 1. Added -C and -D options to the exinext utility, mainly to make it easier
6535 to include in the automated testing, but these could be helpful when
6536 multiple configurations are in use.
6538 2. The exinext utility was not formatting the output nicely when there was
6539 an alternate port involved in the retry record key, nor when there was a
6540 message id as well (for retries that were specific to a specific message
6541 and a specific host). It was also confused by IPv6 addresses, because of
6542 the additional colons they contain. I have fixed the IPv4 problem, and
6543 patched it up to do a reasonable job for IPv6.
6545 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6546 delivery, the log line now contains "pipelined" if PIPELINING was used.
6548 4. An SMTP transport process used to panic and die if the bind() call to set
6549 an explicit outgoing interface failed. This has been changed; it is now
6550 treated in the same way as a connect() failure.
6552 5. A reference to $sender_host_name in the part of a conditional expansion
6553 that was being skipped was still causing a DNS lookup. This no longer
6556 6. The def: expansion condition was not recognizing references to header lines
6557 that used bh_ and bheader_.
6559 7. Added the _cache feature to named lists.
6561 8. The code for checking quota_filecount in the appendfile transport was
6562 allowing one more file than it should have been.
6564 9. For compatibility with Sendmail, the command line option
6572 and sets the incoming protocol and host name (for trusted callers). The
6573 host name and its colon can be omitted when only the protocol is to be set.
6574 Note the Exim already has two private options, -pd and -ps, that refer to
6575 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6576 "s", but I don't think that's a major issue.
6578 10. A number of refactoring changes to the code, none of which should affect
6581 (a) The number of logging options was getting close to filling up the
6582 32-bit word that was used as a bit map. I have split them into two classes:
6583 those that are passed in the argument to log_write(), and those that are
6584 only ever tested independently outside of that function. These are now in
6585 separate 32-bit words, so there is plenty of room for expansion again.
6586 There is no change in the user interface or the logging behaviour.
6588 (b) When building, for example, log lines, the code previously used a
6589 macro that called string_cat() twice, in order to add two strings. This is
6590 not really sufficiently general. Furthermore, there was one instance where
6591 it was actually wrong because one of the argument was used twice, and in
6592 one call a function was used. (As it happened, calling the function twice
6593 did not affect the overall behaviour.) The macro has been replaced by a
6594 function that can join an arbitrary number of extra strings onto a growing
6597 (c) The code for expansion conditions now uses a table and a binary chop
6598 instead of a serial search (which was left over from when there were very
6599 few conditions). Also, it now recognizes conditions like "pam" even when
6600 the relevant support is not compiled in: a suitably worded error message is
6601 given if an attempt is made to use such a condition.
6603 11. Added ${time_interval:xxxxx}.
6605 12. A bug was causing one of the ddress fields not to be passed back correctly
6606 from remote delivery subprocesses. The field in question was not being
6607 subsequently used, so this caused to problems in practice.
6609 13. Added new log selectors queue_time and deliver_time.
6611 14. Might have fixed a bug in maildirsizefile handling that threw up
6612 "unexpected character" debug warnings, and recalculated the data
6613 unnecessarily. In any case, I expanded the warning message to give more
6616 15. Added the message "Restricted characters in address" to the statements in
6617 the default ACL that block characters like @ and % in local parts.
6619 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6620 Three changes have been made:
6622 (a) There was a serious bug; a negative response to MAIL caused the whole
6623 recipient domain to be cached as invalid, thereby blocking all messages
6624 to all local parts at the same domain, from all senders. This bug has
6625 been fixed. The domain is no longer cached after a negative response to
6626 MAIL if the sender used is not empty.
6628 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6631 (c) A new callout option, "use_sender" has been added for people who want
6632 the modified behaviour.
6638 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6641 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6642 indeed breaks things for older releases.
6644 3. Added additional logging to the case where there is a problem reading data
6645 from a filter that is running in a subprocess using a pipe, in order to
6646 try to track down a specific problem.
6648 4. Testing facility fudge: when running in the test harness and attempting
6649 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6650 getting "No route to host". Convert this to a timeout.
6652 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6655 6. Some OS don't have socklen_t but use size_t instead. This affects the
6656 fifth argument of getsockopt() amongst other things. This is now
6657 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6658 can be set for individual OS. I have set it for SunOS5, OSF1, and
6659 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6660 some earlier ones do not.
6662 7. Change 4.30/15 was not doing the test caselessly.
6664 8. The standard form for an IPv6 address literal was being rejected by address
6665 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6666 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6667 this, as well as the form without the "IPv6" on the front (but only when
6668 address literals are enabled, of course).
6670 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6672 10. Exim crashed if a message with an empty sender address specified by -f
6673 encountered a router with an errors_to setting. This could be provoked only
6674 by a command such as
6678 where an empty string was supplied; "<>" did not hit this bug.
6680 11. Installed PCRE release 4.5.
6682 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6683 remained set. It is now erased.
6685 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6686 times from message ids (which are base 36 rather than the normal 62).
6688 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6689 were being counted as actual protocol errors, and logged if the log
6690 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6691 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6692 and DATA following a set of rejected RCPTs do not count as protocol errors.
6693 In other words, Exim assumes they were pipelined, though this may not
6694 actually be the case. Of course, in all cases the client gets an
6695 appropriate error code.
6697 15. If a lookup fails in an ACL condition, a message about the failure may
6698 be available; it is used if testing the ACL cannot continue, because most
6699 such messages specify what the cause of the deferral is. However, some
6700 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6701 that caused an old message to be retained and used if a later statement
6702 caused a defer, replacing the real cause of the deferral.
6704 16. If an IP address had so many PTR records that the DNS lookup buffer
6705 was not large enough to hold them, Exim could crash while trying to process
6706 the truncated data. It now detects and logs this case.
6708 17. Further to 4.21/58, another change has been made: if (and only if) the
6709 first line of a message (the first header line) ends with CRLF, a bare LF
6710 in a subsequent header line has a space inserted after it, so as not to
6711 terminate the header.
6713 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6714 unnecessarily, used atoi() instead of strtol(), and didn't check the
6715 termination when getting file sizes from file names by regex.
6717 19. Completely re-implemented the support for maildirsize files, in the light
6718 of a number of problems with the previous contributed implementation
6719 (4.30/29). In particular:
6721 . If the quota is zero, the maildirsize file is maintained, but no quota is
6724 . If the maildir directory does not exist, it is created before any attempt
6725 to write a maildirsize file.
6727 . The quota value in the file is just a cache; if the quota is changed in
6728 the transport, the new value overrides.
6730 . A regular expression is available for excluding directories from the
6733 20. The autoreply transport checks the characters in options that define the
6734 message's headers; it allows continued headers, but it was checking with
6735 isspace() after an embedded newline instead of explicitly looking for a
6738 21. If all the "regular" hosts to which an address was routed had passed their
6739 expiry times, and had not reached their retry times, the address was
6740 bounced, even if fallback hosts were defined. Now Exim should go on to try
6743 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6744 equivalent code in the SMTP transport. Some hosts send humungous responses
6745 to HELO/EHLO, more than 1024 it seems.
6747 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6748 gives compiler warnings in some environments. I've now done it "properly",
6751 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6752 (because of problems with the built-in one) was declared to return uschar *
6753 instead of char *, causing compiler failure.
6755 25. Fixed a file descriptor leak when processing alias/forward files.
6757 26. Fixed a minor format string issue in dbfn.c.
6759 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6761 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6762 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6765 29. When a sender address is verified, it is cached, to save repeating the test
6766 when there is more than one recipient in a message. However, when the
6767 verification involves a callout, it is possible for different callout
6768 options to be set for different recipients. It is too complicated to keep
6769 track of this in the cache, so now Exim always runs a verification when a
6770 callout is required, relying on the callout cache for the optimization.
6771 The overhead is duplication of the address routing, but this should not be
6774 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6775 to be verified with callout=postmaster, and the main callout worked but the
6776 postmaster check failed, the verification correctly failed. However, if a
6777 subsequent RCPT command asked for sender verification *without* the
6778 postmaster check, incorrect caching caused this verification also to fail,
6781 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6782 it was not caching the DNS options (qualify_single, search_parents) that
6783 were used when the lookup failed. A subsequent lookup with different
6784 options therefore always gave the same answer, though there were cases
6785 where it should not have. (Example: a "domains = !$mx_any" option on a
6786 dnslookup router: the "domains" option is always processed without any
6787 widening, but the router might have qualify_single set.) Now Exim uses the
6788 cached value only when the same options are set.
6790 32. Added John Jetmore's "exipick" utility to the distribution.
6792 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6793 than a timeout (e.g. a certificate is required, and is not provided), an
6794 Exim server now closes the connection immediately. Previously it waited for
6795 the client to close - but if the client is SSL, it seems that they each
6796 wait for each other, leading to a delay before one of them times out.
6798 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6799 maintained 0.8.x compatibility because I don't think many are using it, and
6800 it is clearly obsolete.
6802 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6805 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6806 client certificate was expired. A simple patch fixes this, though I don't
6807 understand the full logic of why the verify callback is called multiple
6810 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6811 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6812 which causes problems with some clients (such as the Certicom SSL Plus
6813 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6814 disables the coutermeasure allowing Eudora to connect."
6816 38. Exim was not checking that a write() to a log file succeeded. This could
6817 lead to Bad Things if a log got too big, in particular if it hit a file
6818 size limit. Exim now panics and dies if it cannot write to a log file, just
6819 as it does if it cannot open a log file.
6821 39. Modified OS/Makefile-Linux so that it now contains
6823 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6825 The two -D definitions ensure that Exim is compiled with large file
6826 support, which makes it possible to handle log files that are bigger than
6829 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6830 instance) a domain was checked against a named list that involved a lookup,
6831 causing $domain_data to be set, then another domain was checked against the
6832 same list, then the first domain was re-checked, the value of $domain_data
6833 after the final check could be wrong. In particular, if the second check
6834 failed, it could be set empty. This bug probably also applied to
6837 41. The strip_trailing_dot option was not being applied to the address given
6838 with the -f command-line option.
6840 42. The code for reading a message's header from the spool was incrementing
6841 $received_count, but never initializing it. This meant that the value was
6842 incorrect (doubled) while delivering a message in the same process in which
6843 it was received. In the most common configuration of Exim, this never
6844 happens - a fresh exec is done - but it can happen when
6845 deliver_drop_privilege is set.
6847 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6848 it now includes up to 150 characters of the unexpected data in the log
6851 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6852 and building data strings. The size of both of these buffers was 10 000
6853 bytes - far larger than anybody would *ever* want, thought I. Needless to
6854 say, somebody hit the limit. I have increased the maximum line length to
6855 20 000 and the maximum data length of concatenated lines to 100 000. I have
6856 also fixed two bugs, because there was no checking on these buffers. Tsk,
6857 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6858 buffer is too small.
6860 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6861 lsearch lookups. Now it does.
6863 46. When parsing a route_list item in a manualroute router, a fixed-length
6864 buffer was used for the list of hosts. I made this 1024 bytes long,
6865 thinking that nobody would ever have a list of hosts that long. Wrong.
6866 Somebody had a whole pile of complicated expansion conditions, and the
6867 string was silently truncated, leading to an expansion error. It turns out
6868 that it is easier to change to an unlimited length (owing to other changes
6869 that have happened since this code was originally written) than to build
6870 structure for giving a limitation error. The length of the item that
6871 expands into the list of hosts is now unlimited.
6873 47. The lsearch lookup could not handle data where the length of text line was
6874 more than 4095 characters. Such lines were truncated, leading to shortened
6875 data being returned. It should now handle lines of any length.
6877 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6878 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6881 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6882 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6883 no longer happens in comment lines. A list of replacements is now placed
6884 at the head of all of the source files, except those whose only change is
6885 to replace PERL_COMMAND in the very first #! line.
6887 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6888 messages on the queue, with a bottom-up merge sort, using code contributed
6889 by Michael Haardt. This should make operations like -bp somewhat faster on
6890 large queues. It won't affect queue runners, except when queue_run_in_order
6893 51. Installed eximstats 1.31 in the distribution.
6895 52. Added support for SRV lookups to the dnslookup router.
6897 53. If an ACL referred to $message_body or $message_body_end, the value was not
6898 reset for any messages that followed in the same SMTP session.
6900 54. The store-handling optimization for building very long strings was not
6901 differentiating between the different store pools. I don't think this
6902 actually made any difference in practice, but I've tidied it.
6904 55. While running the routers to verify a sender address, $sender_address
6905 was still set to the sender address. This is wrong, because when routing to
6906 send a bounce to the sender, it would be empty. Therefore, I have changed
6907 it so that, while verifying a sender address, $sender_address is set to <>.
6908 (There is no change to what happens when verifying a recipient address.)
6910 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6911 target A or AAAA records (if not already returned) without resetting the
6912 qualify_single or search_parents options of the DNS resolver. These are
6913 inappropriate in this case because the targets of MX and SRV records must
6914 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6915 target that, when qualified, matched something in the local domain. These
6916 two options are now turned off when doing these lookups.
6918 57. It seems that at least some releases of Reiserfs (which does not have the
6919 concept of a fixed number of inodes) returns zero and not -1 for the
6920 number of available inodes. This interacted badly with check_spool_inodes,
6921 which assumed that -1 was the "no such thing" setting. What I have done is
6922 to check that the total number of inodes is greater than zero before doing
6923 the test of how many are available.
6925 58. When a "warn" ACL statement has a log_message modifier, the message is
6926 remembered, and not repeated. This is to avoid a lot of repetition when a
6927 message has many recipients that cause the same warning to be written.
6928 However, Exim was preserving the list of already written lines for an
6929 entire SMTP session, which doesn't seem right. The memory is now reset if a
6930 new message is started.
6932 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6933 debugging output unless log_rewrite was also set.
6935 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6936 of (int)(handle) when we know that handle contains (void *)(-1).
6938 61. The Exim daemon panic-logs an error return when it closes the incoming
6939 connection. However "connection reset by peer" seems to be common, and
6940 isn't really an error worthy of noting specially, so that particular error
6943 62. When Exim is trying to find all the local interfaces, it used to panic and
6944 die if the ioctl to get the interface flags failed. However, it seems that
6945 on at least one OS (Solaris 9) it is possible to have an interface that is
6946 included in the list of interfaces, but for which you get a failure error
6947 for this call. This happens when the interface is not "plumbed" into a
6948 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6949 failure of the "get flags" call assumes that the interface is down.
6951 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6952 makes life easier for people who are doing arithmetic on fields extracted
6953 from dates, where you often get leading zeros that should not be
6954 interpreted as octal.
6956 64. Added qualify_domain to the redirect router, to override the global
6959 65. If a pathologically long header line contained very many addresses (the
6960 report of this problem mentioned 10 000) and each of them was rewritten,
6961 Exim could use up a very large amount of memory. (It kept on making new
6962 copies of the header line as it rewrote, and never released the old ones.)
6963 At the expense of a bit more processing, the header rewriting function has
6964 been changed so that it no longer eats memory in this way.
6966 66. The generation of the Received: header has been moved from the time that a
6967 message starts to be received, to the time that it finishes. The timestamp
6968 in the Received: header should now be very close to that of the <= log
6969 line. There are two side-effects of this change:
6971 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6972 logged header lines no longer include the local Received: line, because
6973 it has not yet been created. The same applies to a copy of the message
6974 that is returned to a non-SMTP sender when a message is rejected.
6976 (b) When a filter file is tested using -bf, no additional Received: header
6977 is added to the test message. After some thought, I decided that this
6980 This change does not affect the value of $received_for. It is still set
6981 after address rewriting, but before local_scan() is called.
6983 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6985 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6986 gave an unhelpful panic error message, and a defer error. I have managed to
6987 change this behaviour so that it now rejects any supplied certificate,
6988 which seems right, as the list of acceptable certificates is empty.
6990 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6991 gave an unhelpful defer error. I have not managed to make this reject any
6992 supplied certificates, but the error message it gives is "no certificate
6993 supplied", which is not helpful.
6995 70. exigrep's output now also includes lines that are not associated with any
6996 message, but which match the given pattern. Implemented by a patch from
6997 Martin Sluka, which also tidied up the Perl a bit.
6999 71. Recipient callout verification, like sender verification, was using <> in
7000 the MAIL FROM command. This isn't really the right thing, since the actual
7001 sender may affect whether the remote host accepts the recipient or not. I
7002 have changed it to use the actual sender in the callout; this means that
7003 the cache record is now keyed on a recipient/sender pair, not just the
7004 recipient address. There doesn't seem to be a real danger of callout loops,
7005 since a callout by the remote host to check the sender would use <>.
7006 [SEE ABOVE: changed after hitting problems.]
7008 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7009 temporary errors. However, in the case of such a code being given after
7010 the end of a data transmission (i.e. after ".") Exim was failing to write
7011 a retry record for the message. (Yes, there was some broken host that was
7012 actually sending 8xx at this point.)
7014 73. An unknown lookup type in a host list could cause Exim to panic-die when
7015 the list was checked. (An example that provoked this was putting <; in the
7016 middle of a list instead of at the start.) If this happened during a DATA
7017 ACL check, a -D file could be left lying around. This kind of configuration
7018 error no longer causes Exim to die; instead it causes a defer error. The
7019 incident is still logged to the main and panic logs.
7021 74. Buglet left over from Exim 3 conversion. The message "too many messages
7022 in one connection" was written to the rejectlog but not the mainlog, except
7023 when address rewriting (yes!) was being logged.
7025 75. Added write_rejectlog option.
7027 76. When a system filter was run not as root (that is, when system_filter_user
7028 was set), the values of the $n variables were not being returned to the
7029 main process; thus, they were not subsequently available in the $sn
7032 77. Added +return_path_on_delivery log selector.
7034 78. A connection timeout was being treated differently from recipients deferred
7035 when testing hosts_max_try with a message that was older than the host's
7036 retry timeout. (The host should not be counted, thus allowing all hosts to
7037 be tried at least once before bouncing.) This may have been the cause of an
7038 occasionally reported bug whereby a message would remain on the queue
7039 longer than the retry timeout, but would be bounced if a delivery was
7040 forced. I say "may" because I never totally pinned down the problem;
7041 setting up timeout/retry tests is difficult. See also the next item.
7043 79. The ultimate address timeout was not being applied to errors that involved
7044 a combination of host plus message (for example, a timeout on a MAIL
7045 command). When an address resolved to a number of possible hosts, and they
7046 were not all tried for each delivery (e.g. because of hosts_max_try), a
7047 message could remain on the queue longer than the retry timeout.
7049 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7052 81. Fixed an obscure SMTP outgoing bug which required at least the following
7053 conditions: (a) there was another message waiting for the same server;
7054 (b) the server returned 5xx to all RCPT commands in the first message so
7055 that the message was not completed; (c) the server dropped the connection
7056 or gave a negative response to the RSET that Exim sends to abort the
7057 transaction. The observed case was a dropped connection after DATA that had
7058 been sent in pipelining mode. That is, the server had advertised PIPELINING
7059 but was not implementing it correctly. The effect of the bug was incorrect
7060 behaviour, such as trying another host, and this could lead to a crash.
7066 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7067 and daemon.c were passed as pointers to ints; they should have been
7068 pointers to socklen_t variables (which are typically unsigned ints).
7070 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7073 3. Fixed a really odd bug that affected only the testing scheme; patching a
7074 certain fixed string in the binary changed the value of another string that
7075 happened to be identical to the end of the original first string.
7077 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7078 name", it returns that address as the IP address. On some operating
7079 systems (e.g. Solaris), it also passes back the IP address string as the
7080 "host name". However, on others (e.g. Linux), it passes back an empty
7081 string. Exim wasn't checking for this, and was changing the host name to an
7082 empty string, assuming it had been canonicalized.
7084 5. Although rare, it is permitted to have more than one PTR record for a given
7085 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7086 all the names associated with an address, because they do in Solaris.
7087 However, it seems that they do not in Linux for data that comes from the
7088 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7089 I found this out when I moved to a new Linux workstation and tried to run
7090 the Exim test suite.
7092 To get round this problem I have changed the code so that it now does its
7093 own call to the DNS to look up PTR records when searching for a host name.
7094 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7095 addresses that are only in /etc/hosts are still found.
7097 This behaviour is, however, controlled by an option called host_lookup_
7098 order, which defaults to "bydns:byaddr". If people want to use the other
7099 order, or indeed, just use one or the other means of lookup, they can
7100 specify it in this variable.
7102 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7103 some operating systems, this comes back from gethostbyaddr() as an empty
7104 string, and this is what Exim used to test for. However, it seems that in
7105 other systems, "." is yielded. Exim now tests for this case too.
7107 7. The values of check_spool_space and check_log_space are now held internally
7108 as a number of kilobytes instead of an absolute number of bytes. If a
7109 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7110 kilobyte. This means that much larger values can be stored.
7112 8. Exim monitor: an attempt to get the action menu when not actually pointing
7113 at a message produces an empty menu entitled "No message selected". This
7114 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7115 no entries in it ("Shell widget menu has zero width and/or height"). So I
7116 have added a single, blank menu entry in this case.
7118 9. Added ${quote_local_part.
7120 10. MIME decoding is now applied to the contents of Subject: header lines when
7123 11. Now that a reference to $sender_host_address automatically causes a reverse
7124 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7125 host lookup before query-style lookups in lists that might use this
7126 variable. This has therefore been abolished, and the "net-" prefix is no
7127 longer necessary for query-style lookups.
7129 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7130 have been a typo for LFLAGS, so it has been changed.
7132 13. The install script calls Exim with "-C /dev/null" in order to find the
7133 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7134 to be output. However, since Exim outputs its version number before the
7135 error, it didn't break the script. It just looked ugly. I fixed this by
7136 always allowing "-C /dev/null" if the caller is root.
7138 14. Ignore overlarge ACL variable number when reading spool file - insurance
7139 against a later release with more variables having written the file.
7141 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7142 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7143 this, as well as the form without the "IPv6" on the front.
7145 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7146 OS/Makefile-Darwin file.
7148 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7149 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7151 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7152 variable. However, some OS use ints for inodes. I've added cast to long int
7153 to get rid of the compiler warning.
7155 19. I had forgotten to lock out "/../" in configuration file names when
7156 ALT_CONFIG_PREFIX was set.
7158 20. Routers used for verification do not need to specify transports. However,
7159 if such a router generated a host list, and callout was configured, Exim
7160 crashed, because it could not find a port number from the (non-existent)
7161 transport. It now assumes port 25 in this circumstance.
7163 21. Added the -t option to exigrep.
7165 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7166 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7167 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7168 them is set, LOOKUP_LSEARCH is forced.
7170 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7172 24. Added sender and host information to the "rejected by local_scan()" log
7173 line; previously there was no indication of these.
7175 25. Added .include_if_exists.
7177 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7178 Linux. It turns out that not all file systems support this. Apparently some
7179 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7180 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7181 not supported on the file descriptor, is now ignored when Exim is trying to
7182 sync a directory. This applies only to Linux.
7184 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7186 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7187 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7190 29. Added support for maildirsize files from supplied patch (modified a bit).
7192 30. The use of :fail: followed by an empty string could lead Exim to respond to
7193 sender verification failures with (e.g.):
7195 550 Verification failed for <xxx>
7196 550 Sender verify failed
7198 where the first response line was missing the '-' that indicates it is not
7199 the final line of the response.
7201 31. The loop for finding the name of the user that called Exim had a hardwired
7202 limit of 10; it now uses the value of finduser_retries, which is used for
7203 all other user lookups.
7205 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7208 33. Exim was neglecting to zero errno before one call of strtol() when
7209 expanding a string and expecting an integer value. On some systems this
7210 resulted in spurious "integer overflow" errors. Also, it was casting the
7211 result into an int without checking.
7213 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7214 did not work. The code looks as if it has *never* worked, though it appears
7215 to have been documented since at least release 1.62. I have made it work.
7217 35. The "timeout_DNS" error in retry rules, also documented since at least
7218 1.62, also never worked. As it isn't clear exactly what this means, and
7219 clearly it isn't a major issue, I have abolished the feature by treating it
7220 as "timeout", and writing a warning to the main and panic logs.
7222 36. The display of retry rules for -brt wasn't always showing the error code
7225 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7226 timeout_connect_A, timeout_connect_MX.
7228 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7229 to the empty sender.
7231 39. The daemon was not analysing the content of -oX till after it had closed
7232 stderr and disconnected from the controlling terminal. This meant that any
7233 syntax errors were only noted on the panic log, and the return code from
7234 the command was 0. By re-arranging the code a little, I've made the
7235 decoding happen first, so such errors now appear on stderr, and the return
7236 code is 1. However, the actual setting up of the sockets still happens in
7237 the disconnected process, so errors there are still only recorded on the
7240 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7241 connections (as happens on some IP stacks) was logged at start up time as
7242 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7243 it from "IPv6 and IPv4", which means that two separate sockets are being
7246 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7247 says whether AF_INET or AF_INET6 was passed as an argument.
7249 42. Exiwhat output was messed up when time zones were included in log
7252 43. Exiwhat now gives more information about the daemon's listening ports,
7253 and whether -tls-on-connect was used.
7255 44. The "port" option of the smtp transport is now expanded.
7257 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7258 silently ignored. Now an error message is written to the main and panic
7261 46. There's a new ACL modifier called "logwrite" which writes to a log file
7262 as soon as it is encountered.
7264 47. Added $local_user_uid and $local_user_gid at routing time.
7266 48. Exim crashed when trying to verify a sender address that was being
7269 49. Exim was recognizing only a space character after ".include". It now also
7270 recognizes a tab character.
7272 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7273 extracting the relevant information from the specification. The man page no
7274 longer contains scrambled data for the -d option, and I've added a section
7275 at the front about calling Exim under different names.
7277 51. Added "extra_headers" argument to the "mail" command in filter files.
7279 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7282 53. Installed eximstats 1.29.
7284 54. Added transport_filter_timeout as a generic transport option.
7286 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7287 Cc: header lines. This was required by RFC 822, but it not required by RFC
7290 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7291 incoming messages that did not have them. Now it does so only if the
7292 message originates locally, that is, if there is no associated remote host
7293 address. When Resent- header lines are present, this applies to the Resent-
7294 lines rather than the non-Resent- lines.
7296 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7297 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7299 58. Messages for configuration errors now include the name of the main
7300 configuration file - useful now that there may be more than one file in a
7301 list (.included file names were always shown).
7303 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7304 for those rare installations that do not start the daemon as root or run it
7305 setuid root. I've cut out the call to initgroups() if the daemon is not
7308 60. The Exim user and group can now be bound into the binary as text strings
7309 that are looked up at the start of Exim's processing.
7311 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7313 62. Added $mailstore_basename variable.
7315 63. Installed patch to sieve.c from Michael Haardt.
7317 64. When Exim failed to open the panic log after failing to open the main log,
7318 the original message it was trying to log was written to stderr and debug
7319 output, but if they were not available (the usual case in production), it
7320 was lost. Now it is written to syslog before the two lines that record the
7321 failures to open the logs.
7323 65. Users' Exim filters run in subprocesses under the user's uid. It is
7324 possible for a "deliver" command or an alias in a "personal" command to
7325 provoke an address rewrite. If logging of address rewriting is configured,
7326 this fails because the process is not running as root or exim. There may be
7327 a better way of dealing with this, but for the moment (because 4.30 needs
7328 to be released), I have disabled address rewrite logging when running a
7329 filter in a non-root, non-exim process.
7335 1. The buildconfig auxiliary program wasn't quoting the value set for
7336 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7337 not defined. This bug was masked in 4.22 by the effect that was fixed in
7340 2. Some messages that were rejected after a message id was allocated were
7341 shown as "incomplete" by exigrep. It no longer does this for messages that
7342 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7344 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7345 have allow_domain_literals set, the ID did not get logged in the <= line.
7346 Domain literals are now always recognized in Message-ID: header lines.
7348 4. The first argument for a ${extract expansion item is the key name or field
7349 number. Leading and trailing spaces in this item were not being ignored,
7350 causing some misleading effects.
7352 5. When deliver_drop_privilege was set, single queue runner processes started
7353 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7354 same command in the process it spins off) were not dropping privilege.
7356 6. When the daemon running as "exim" started a queue runner, it always
7357 re-executed Exim in the spun-off process. This is a waste of effort when
7358 deliver_drop_privilege is set. The new process now just calls the
7359 queue-runner function directly.
7365 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7368 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7369 ignored. Though the use of -oP was forcing the writing of a pid file, it
7370 was always written to the default place.
7372 3. If the message "no IP address found for host xxxx" is generated during
7373 incoming verification, it is now followed by identification of the incoming
7374 connection (so you can more easily find what provoked it).
7376 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7378 5. Added some features to "harden" Exim a bit more against certain attacks:
7380 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7381 be put in Local/Makefile. This is like the never_users runtime option,
7382 but it cannot be overridden. The default setting is "root".
7384 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7385 prefix string with which any file named in a -C command line option
7388 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7389 is retained for -C and -D only if the caller of Exim is root. Without
7390 it, the exim user may also use -C and -D and retain privilege.
7392 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7393 command line option is disabled.
7395 6. Macro names set by the -D option must start with an upper case letter, just
7396 like macro names defined in the configuration file.
7398 7. Added "dereference=" facility to LDAP.
7400 8. Two instances of the typo "uknown" in the source files are fixed.
7402 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7403 the Configure-Makefile script screwed up while processing it.
7405 10. Incorporated PCRE 4.4.
7407 11. The SMTP synchronization check was not operating right at the start of an
7408 SMTP session. For example, it could not catch a HELO sent before the client
7409 waited for the greeting. There is now a check for outstanding input at the
7410 point when the greeting is written. Because of the duplex, asynchronous
7411 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7412 way, but not yet received, when the check is performed.
7414 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7415 on TCP/IP sockets, because this apparently causes some broken clients to
7418 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7419 unchanged) from the Cygwin maintainer.
7421 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7422 was supported instead of testing for mbx. Effectively a typo.
7424 15. The spa authenticator server code was not checking that the input it
7425 received was valid base64.
7427 16. The debug output line for the "set" modifier in ACLs was not showing the
7428 name of the variable that was being set.
7430 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7432 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7433 lookup on its own. Something else was needed to trigger the lookup. For
7434 example, a match in host_lookup or the need for a host name in a host list.
7435 Now, if $sender_host_name is referenced and the host name has not yet been
7436 looked up, a lookup is performed. If the lookup fails, the variable remains
7437 empty, and $host_lookup_failed is set to "1".
7439 19. Added "eqi" as a case-independent comparison operator.
7441 20. The saslauthd authentication condition could segfault if neither service
7442 nor realm was specified.
7444 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7445 error message that was logged was misleading, and incoming SMTP
7446 connections were dropped. The message is now more accurate, and temporary
7447 errors are given to SMTP connections.
7449 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7450 (except RSET) with a 421 error, until QUIT is received. However, it was
7451 failing to send a response to QUIT.
7453 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7454 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7455 ACL is now run after the helo_try_verify_hosts code.
7457 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7458 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7459 case-independent, so other case variants are also recognized). Apparently
7460 some systems use these upper case variants.
7462 25. If more than two messages were waiting for the same host, and a transport
7463 filter was specified for the transport, Exim sent two messages over the
7464 same TCP/IP connection, and then failed with "socket operation on non-
7465 socket" when it tried to send the third.
7467 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7469 27. The extern definition of crypt16() in expand.c was not being excluded when
7470 the OS had its own crypt16() function.
7472 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7473 as a preferred synonym for return_size_limit, both as an option and as an
7476 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7478 30. Changed the default configuration ACL to relax the local part checking rule
7479 for addresses that are not in any local domains. For these addresses,
7480 slashes and pipe symbols are allowed within local parts, but the sequence
7481 /../ is explicitly forbidden.
7483 31. SPA server authentication was not clearing the challenge buffer before
7486 32. log_message in a "warn" ACL statement was writing to the reject log as
7487 well as to the main log, which contradicts the documentation and doesn't
7488 seem right (because no rejection is happening). So I have stopped it.
7490 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7491 However, I am unable to do any testing of this.
7493 34. Fixed an infelicity in the appendfile transport. When checking directories
7494 for a mailbox, to see if any needed to be created, it was accidentally
7495 using path names with one or more superfluous leading slashes; tracing
7496 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7498 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7499 contents are added to the log line that is written for every discarded
7500 recipient. (Previously a log_message setting was ignored.)
7502 36. The ${quote: operator now quotes the string if it is empty.
7504 37. The install script runs exim in order to find its version number. If for
7505 some reason other than non-existence or emptiness, which it checks, it
7506 could not run './exim', it was installing it with an empty version number,
7507 i.e. as "exim-". This error state is now caught, and the installation is
7510 38. An argument was missing from the function that creates an error message
7511 when Exim fails to connect to the socket for saslauthd authentication.
7512 This could cause Exim to crash, or give a corrupted message.
7514 39. Added isip, isip4, and isip6 to ${if conditions.
7516 40. The ACL variables $acl_xx are now saved with the message, and can be
7517 accessed later in routers, transports, and filters.
7519 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7520 strings in the file are not string-expanded.
7522 42. If a MAIL command specified a SIZE value that was too large to fit into an
7523 int variable, the check against message_size_limit failed. Such values are
7524 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7525 day this will have to be increased, but I don't think I want to be around
7526 when emails are that large.
7533 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7534 iconv() is not standard in FreeBSD.
7536 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7537 IPv6 enabled. The observed symptom was a segmentation fault on return from
7538 the function os_common_find_running_interfaces() in src/os.c.
7540 3. In the check_special_case() function in daemon.c I had used "errno" as an
7541 argument name, which causes warnings on some systems. This was basically a
7542 typo, since it was named "eno" in the comments!
7544 4. The code that waits for the clock to tick (at a resolution of some fraction
7545 of a second) so as to ensure message-id uniqueness was always waiting for
7546 at least one whole tick, when it could have waited for less. [This is
7547 almost certainly not relevant at current processor speeds, where it is
7548 unlikely to ever wait at all. But we try to future-proof.]
7550 5. The function that sleeps for a time interval that includes fractions of a
7551 second contained a race. It did not block SIGALRM between setting the
7552 timer, and suspending (a couple of lines later). If the interval was short
7553 and the sigsuspend() was delayed until after it had expired, the suspension
7554 never ended. On busy systems this could lead to processes getting stuck for
7557 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7558 process, before it forks any delivery processes. The open lookup caching
7559 mechanism meant that the open file or database connection was passed into
7560 the delivery process. The problem was that delivery processes always tidy
7561 up cached lookup data. This could cause a problem for the next delivery
7562 process started by the queue runner, because the external queue runner
7563 process does not know about the closure. So the next delivery process
7564 still has data in the lookup cache. In the case of a file lookup, there was
7565 no problem because closing a file descriptor in a subprocess doesn't affect
7566 the parent. However, if the lookup was caching a connection to a database,
7567 the connection was closed, and the second delivery process was likely to
7568 see errors such as "PGSQL: query failed: server closed the connection
7569 unexpectedly". The problem has been fixed by closing all cached lookups
7570 in a queue runner before running a delivery process.
7572 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7573 seem to have the "const" qualifier which it has on other OS. I've
7576 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7577 *of the same type* (client or server) with the same public name that an
7578 error should be diagnosed.
7580 9. When Exim looked up a host name for an IP address, but failed to find the
7581 original IP address when looking up the host name (a safety check), it
7582 output the message "<ip address> does not match any IP for NULL", which was
7583 confusing, to say the least. The bug was that the host name should have
7584 appeared instead of "NULL".
7586 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7587 user that is built into the binary, and the -C or -D options is used, root
7588 privilege is dropped before the configuration file is read. In addition,
7589 logging is switched to stderr instead of the normal log files. If the
7590 configuration then re-defines the Exim user, the unprivileged environment
7591 is probably not what is expected, so Exim logs a panic warning message (but
7594 However, if deliver_drop_privilege is set, the unprivileged state may well
7595 be exactly what is intended, so the warning has been cut out in that case,
7596 and Exim is allowed to try to write to its normal log files.
7602 1. smtp_return_error_details was not giving details for temporary sender
7603 or receiver verification errors.
7605 2. Diagnose a configuration error if two authenticators have the same public
7608 3. Exim used not to create the message log file for a message until the first
7609 delivery attempt. This could be confusing when incoming messages were held
7610 for policy or load reasons. The message log file is now created at the time
7611 the message is received, and an initial "Received" line is written to it.
7613 4. The automatically generated man page for command line options had a minor
7614 bug that caused no ill effects; however, a more serious problem was that
7615 the procedure for building the man page automatically didn't always
7616 operate. Consequently, release 4.20 contains an out-of-date version. This
7617 shouldn't happen again.
7619 5. When building Exim with embedded Perl support, the script that builds the
7620 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7621 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7623 6. The freeze_tell option was not being used for messages that were frozen on
7624 arrival, either by an ACL or by local_scan().
7626 7. Added the smtp_incomplete_transaction log selector.
7628 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7629 was accepting AUTH without a new EHLO.
7631 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7632 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7635 10. Logging of TCP/IP connections (when configured) now happens in the main
7636 daemon process instead of the child process, so that the TCP/IP connection
7637 count is more accurate (but it can never be perfect).
7639 11. The use of "drop" in a nested ACL was not being handled correctly in the
7640 outer ACL. Now, if condition failure induced by the nested "drop" causes
7641 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7642 or "require"), the connection is dropped.
7644 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7645 that yield "discard" can now be used with an "accept" or a "discard" verb,
7646 but an error is generated for any others (because I can't see a useful way
7647 to define what should happen).
7649 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7650 are now processed in the same way as lines in the Exim configuration file.
7651 In particular, continuation lines are supported.
7653 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7655 15. Added -ti meaning -t -i.
7657 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7658 domains, and warn by logging if others are found.
7660 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7661 returned from GIFCONF when Exim is trying to find the list of interfaces on
7662 a host. The code in os.c has been modified to copy each ifreq to an aligned
7663 structure in all cases.
7665 Also, in some cases, the returned ifreq's were being copied to a 'struct
7666 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7667 means the last couple of bytes of an IPv6 address could be chopped if the
7668 ifreq contained only a normal sockaddr (14 bytes storage).
7670 18. Named domain lists were not supported in the hosts_treat_as_local option.
7671 An entry such as +xxxx was not recognized, and was treated as a literal
7674 19. Ensure that header lines added by a DATA ACL are included in the reject log
7675 if the ACL subsequently rejects the message.
7677 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7678 MD5 (which is deprecated).
7680 21. When testing a filter file using -bf, Exim was writing a message when it
7681 took the sender from a "From " line in the message, but it was not doing so
7682 when it took $return_path from a Return-Path: header line. It now does.
7684 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7685 with a valid header line field name (a series of printing characters
7686 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7688 23. Changed "disc" in the source to "disk" to conform to the documentation and
7689 the book and for uniformity.
7691 24. Ignore Sendmail's -Ooption=value command line item.
7693 25. When execve() failed while trying to run a command in a pipe transport,
7694 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7695 could be confused with a return value of 69 from the command itself. This
7696 has been changed to 127, the value the shell returns if it is asked to run
7697 a non-existent command. The wording for the related log line suggests a
7698 non-existent command as the problem.
7700 26. If received_header_text expands to an empty string, do not add a Received:
7701 header line to the message. (Well, it adds a token one on the spool, but
7702 marks it "old" so that it doesn't get used or transmitted.)
7704 27. Installed eximstats 1.28 (addition of -nt option).
7706 28. There was no check for failure on the call to getsockname() in the daemon
7707 code. This can fail if there is a shortage of resources on the system, with
7708 ENOMEM, for example. A temporary error is now given on failure.
7710 29. Contrary to the C standard, it seems that in some environments, the
7711 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7712 program. Exim now does this explicitly; it affects the formatting of
7713 timestamps using strftime().
7715 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7716 complaints. I've now initialized all the variables, to avoid this.
7718 32. Header lines added by a system filter were not being "seen" during
7719 transport-time rewrites.
7721 33. The info_callback() function passed to OpenSSL is set up with type void
7722 (*)(SSL *, int, int), as described somewhere. However, when calling the
7723 function (actually a macro) that sets it up, the type void(*)() is
7724 expected. I've put in a cast to prevent warnings from picky compilers.
7726 34. If a DNS black list lookup found a CNAME record, but there were no A
7727 records associated with the domain it pointed at, Exim crashed.
7729 35. If a DNS black list lookup returned more than one A record, Exim ignored
7730 all but the first. It now scans all returned addresses if a particular IP
7731 value is being sought. In this situation, the contents of the
7732 $dnslist_value variable are a list of all the addresses, separated by a
7735 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7736 to accept a host name and all its aliases if the forward lookup for any of
7737 them yielded the IP address of the incoming connection. Now it accepts only
7738 those names whose forward lookup yields the correct IP address. Any other
7739 names are discarded. This closes a loophole whereby a rogue DNS
7740 administrator could create reverse DNS records to break through a
7741 wildcarded host restriction in an ACL.
7743 37. If a user filter or a system filter that ran in a subprocess used any of
7744 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7745 the wrong values were passed to the pipe command ($thisaddress had the
7746 value of $0, $0 had the value of $1, etc). This bug was introduced by
7747 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7749 38. Improved the line breaking for long SMTP error messages from ACLs.
7750 Previously, if there was no break point between 40 and 75 characters, Exim
7751 left the rest of the message alone. Two changes have been made: (a) I've
7752 reduced the minimum length to 35 characters; (b) if it can't find a break
7753 point between 35 and 75 characters, it looks ahead and uses the first one
7754 that it finds. This may give the occasional overlong line, but at least the
7755 remaining text gets split now.
7757 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7758 file descriptors might be low, and that setting 1000 would always raise it.
7759 It turns out that in some environments, the limit is already over 1000 and
7760 that lowering it causes trouble. So now Exim takes care not to decrease it.
7762 40. When delivering a message, the value of $return_path is set to $sender_
7763 address at the start of routing (routers may change the value). By an
7764 oversight, this default was not being set up when an address was tested by
7765 -bt or -bv, which affected the outcome if any router or filter referred to
7768 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7769 log only when "message" or "log_message" are set. However, if one of the
7770 conditions was an address verification, or a call to a nested ACL, the
7771 messages generated by the underlying test were being passed through. This
7772 no longer happens. The underlying message is available in $acl_verify_
7773 message for both "message" and "log_message" expansions, so it can be
7774 passed through if needed.
7776 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7777 new expansion $bh_ to give the encoded byte string without charset
7778 translation. Translation happens only if iconv() is available; HAVE_ICONV
7779 indicates this at build time. HEADERS_CHARSET gives the charset to
7780 translate to; headers_charset can change it in the configuration, and
7781 "headers charset" can change it in an individual filter file.
7783 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7784 that creates RFC 2047 encoded "words" labels them as that charset instead
7785 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7786 expansion operator; (ii) when Exim creates a From: line for a local
7787 message; (iii) when a header line is rewritten to include a "phrase" part.
7789 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7790 buggy, causing it to skip the first lines of messages whose message ID
7791 ended in 'D'. This would not have bitten before Exim release 4.14, because
7792 message IDs were unlikely to end in 'D' before then. The effect was to have
7793 incorrect size information for certain domains.
7795 45. #include "config.h" was missing at the start of the crypt16.c module. This
7796 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7799 46. If there was a timeout during a "random" callout check, Exim treated it as
7800 a failure of the random address, and carried on sending RSET and the real
7801 address. If the delay was just some slowness somewhere, the response to the
7802 original RCPT would be taken as a response to RSET and so on, causing
7803 mayhem of various kinds.
7805 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7806 when I implemented it. It didn't allow for the fact that some option values
7807 may legitimately be negative (e.g. size_addition), and it didn't even do
7808 the right test for positive values.
7810 48. Domain names in DNS records are case-independent. Exim always looks them up
7811 in lower case. Some resolvers return domain names in exactly the case they
7812 appear in the zone file, that is, they may contain uppercase letters. Not
7813 all resolvers do this - some return always lower case. Exim was treating a
7814 change of case by a resolver as a change of domain, similar to a widening
7815 of a domain abbreviation. This triggered its re-routing code and so it was
7816 trying to route what was effectively the same domain again. This normally
7817 caused routing to fail (because the router wouldn't handle the domain
7818 twice). Now Exim checks for this case specially, and just changes the
7819 casing of the domain that it ultimately uses when it transmits the message
7822 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7825 50. If a filter generated a file delivery with a non-absolute name (possible if
7826 no home directory exists for the router), the forbid_file option was not
7829 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7830 the existing equality matching.
7832 52. Exim was using ints instead of ino_t variables in some places where it was
7833 dealing with inode numbers.
7835 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7836 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7837 called TMPDIR, and if it finds it is different, it changes its value.
7839 54. The smtp_printf() function is now made available to local_scan() so
7840 additional output lines can be written before returning. There is also an
7841 smtp_fflush() function to enable the detection of a dropped connection.
7842 The variables smtp_input and smtp_batched_input are exported to
7845 55. Changed the default runtime configuration: the message "Unknown user"
7846 has been removed from the ACL, and instead placed on the localuser router,
7847 using the cannot_route_message feature. This means that any verification
7848 failures that generate their own messages won't get overridden. Similarly,
7849 the "Unrouteable address" message that was in the ACL for unverifiable
7850 relay addresses has also been removed.
7852 56. Added hosts_avoid_esmtp to the smtp transport.
7854 57. The exicyclog script was not checking for the esoteric option
7855 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7856 will work only if exicyclog is run under the appropriate euid.
7858 58. Following a discussion on the list, the rules by which Exim recognises line
7859 endings on incoming messages have been changed. The -dropcr and drop_cr
7860 options are now no-ops, retained only for backwards compatibility. The
7861 following line terminators are recognized: LF CRLF CR. However, special
7862 processing applies to CR:
7864 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7865 nor a local message in the state where . is a terminator.
7867 (ii) If a bare CR is encountered in a header line, an extra space is added
7868 after the line terminator so as not to end the header. The reasoning
7869 behind this is that bare CRs in header lines are most likely either
7870 to be mistakes, or people trying to play silly games.
7872 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7873 was being incorrectly given as 18 bytes larger than it should have been.
7874 This is a VOB (very old bug).
7876 60. This may never have affected anything current, but just in case it has:
7877 When the local host is found other than at the start of a list of hosts,
7878 the local host, those with the same MX, and any that follow, are discarded.
7879 When the list in question was part of a longer list of hosts, the following
7880 hosts (not currently being processed) were also being discarded. This no
7881 longer happens. I'm not sure if this situation could ever has previously
7884 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7887 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7888 header. This is something that is recommended in a new Internet Draft, and
7889 is something that is documented as being done by Sendmail. There are two
7890 possible values. For messages generated by the autoreply transport, Exim
7893 Auto-Submitted: auto-replied
7895 whereas for all other generated messages (e.g. bounces) it adds
7897 Auto-Submitted: auto-generated
7899 63. The "personal" condition in filters now includes a test for the
7900 Auto-Submitted: header. If it contains the string "auto-" the message it
7901 not considered personal.
7903 64. Added rcpt_include_affixes as a generic transport option.
7905 65. Added queue_only_override (default true).
7907 66. Added the syslog_duplication option.
7909 67. If what should have been the first header line of a message consisted of
7910 a space followed by a colon, Exim was mis-interpreting it as a header line.
7911 It isn't of course - it is syntactically invalid and should therefore be
7912 treated as the start of the message body. The misbehaviour could have
7913 caused a number of strange effects, including loss of data in subsequent
7914 header lines, and spool format errors.
7916 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7917 client host had authenticated. This control can now be exercised by an ACL
7918 for more flexibility.
7920 69. By default, callouts do not happen when testing with -bh. There is now a
7921 variant, -bhc, which does actually run the callout code, including
7922 consulting and updating the callout cache.
7924 70. Added support for saslauthd authentication, courtesy of Alexander
7927 71. If statvfs() failed on the spool or log directories while checking their
7928 size for availability, Exim confusingly gave the error "space shortage".
7929 Furthermore, in debugging mode it crashed with a floating point exception.
7930 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7931 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7932 serious problem, Exim now writes to the main and panic logs when this
7933 happens, with details of the failure. It then refuses to accept the
7934 incoming message, giving the message "spool directory problem" or "log
7935 directory problem" with a 421 code for SMTP messages.
7937 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7938 0, 1, and 2 exist, because some OS complain for execs without them (see
7939 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7940 descriptors. However, the code omitted to check that the open succeeded,
7941 causing mysterious errors if for some reason the permissions on /dev/null
7942 got screwed. Now Exim writes a message to the main and panic logs, and
7943 bombs out if it can't open /dev/null.
7945 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7946 interact so that it is all more flexible. It is supposed to remain
7947 backwards compatible. Also added extra_local_interfaces.
7949 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7950 to bomb out with an assertion failure - to the client this appears as a
7951 connection drop. This problem occurs in the part of the code that was taken
7952 from the Samba project. Fortunately, the assertion is in a very simple
7953 function, so I have fixed this by reproducing the function inline in the
7954 one place where it is called, and arranging for authentication to fail
7955 instead of killing the process with assert().
7957 75. The SPA client code was not working when the server requested OEM rather
7958 than Unicode encoding.
7960 76. Added code to make require_files with a specific uid setting more usable in
7961 the case where statting the file as root fails - usually a non-root-mounted
7962 NFS file system. When this happens and the failure is EACCES, Exim now
7963 forks a subprocess and does the per-uid checking as the relevant uid.
7965 77. Added process_log_path.
7967 78. If log_file_path was not explicitly set, a setting of check_log_space or
7968 check_log_inodes was ignored.
7970 79. If a space check for the spool or log partitions fails, the incident is now
7971 logged. Of course, in the latter case the data may get lost...
7973 80. Added the %p formatting code to string_format() so that it can be used to
7974 print addresses in debug_print(). Adjusted all the address printing in the
7975 debugging in store.c to use %p rather than %d.
7977 81. There was a concern that a line of code in smtp_in.c could overflow a
7978 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7979 initially expressed, the concern was not well-founded, because trailing
7980 spaces are removed early. However, if the trailing spaces were followed by
7981 a NULL, they did not get removed, so the overflow was possible. Two fixes
7984 (a) I re-wrote the offending code in a cleaner fashion.
7985 (b) If an incoming SMTP command contains a NULL character, it is rejected
7988 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7989 runs initgroups(), so that if the Exim user is in any additional groups,
7990 they will be used during message reception.
7996 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob