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.
172 JH/01 Avoid costly startup code when not strictly needed. This reduces time
173 for some exim process initialisations. It does mean that the logging
174 of TLS configuration problems is only done for the daemon startup.
176 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
178 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
179 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
182 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
183 DKIM signing is being done. Previously a transport_filter would always
184 disable CHUNKING, falling back to traditional DATA.
186 JH/05 Regard command-line receipients as tainted.
188 JH/06 Bug 340: Remove the daemon pid file on exit, whe due to SIGTERM.
190 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
191 PAM library frees one of the arguments given to it, despite the
192 documentation. Therefore a plain malloc must be used.
194 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
195 on-stack buffers were used, resulting in a taint trap when DSN information
196 copied from a received message was written into the buffer.
198 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
199 the ordering of its ARC headers. This caused a crash.
201 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
202 a new record was being constructed with information from the peer, a trap
205 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
206 installation would get error messages from DMARC verify, when it hit the
207 nonexistent file indicated by the default. Distros wanting DMARC enabled
208 should both provide the file and set the option.
209 Also enforce no DMARC verification for command-line sourced messages.
211 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
212 could, depending on the platform, hang at the STARTTLS response.
214 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
215 message on a connection. Previously if one message had ARC headers and
216 the following one did not, a crash could result when adding an
217 Authentication-Results: header.
219 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
220 Exim main code and Exim-related utities. The introduction of taint
221 tracking also did many adjustments to string handling. Since then, eximon
222 frequently terminated with an assert failure.
224 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
225 check for 452 responses. This slightly helps the inefficieny of doing
226 a large alias-expansion into a recipient-limited target. The max_rcpt
227 transport option still applies (and at the current default, will override
228 the new feature). The check is done for either cause of synch, and forces
229 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
230 connection. The new facility is not tunable at this time.
232 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
233 library live data was being used, so the results became garbage. Make
234 copies while it is still usable.
236 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
237 on delivery deferred (==) and failed (**) lines (if a delivery was
238 attemtped). Previously it was only on completion (=>) lines.
240 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
241 for the expansion of the server_scram_iter and server_scram_salt options.
243 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
244 are now specifically given a NO_DATA response without hitting the system
245 resolver. The library goes on to do the now-standard TXT lookup.
246 Use of dnsdb lookups is not affected.
248 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
249 only retrieve the errormessage once. Previously two calls to dlerror()
250 were used, and the second one (for mainlog/paniclog) retrieved null
253 JH/20 Taint checking: disallow use of tainted data for
254 - the appendfile transport file and directory options
255 - the pipe transport command
256 - the autoreply transport file, log and once options
257 - file names used by the redirect router (including filter files)
259 - paths used by single-key lookups
260 Previously this was permitted.
262 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
263 adjusted the size of a major service buffer; this failed because the
264 buffer was in use at the time. Change to a compile-time increase in the
265 buffer size, when this authenticator is compiled into exim.
267 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
268 previous fast-mode was untenable in the face of glibs using mmap to
269 support larger malloc requests.
271 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
272 New values supported, if defined on system where compiled:
273 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
274 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
276 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
277 running a limited number of proceses in parallel, a benefit is gained. The
278 amount varies with the platform hardware and load. The use of the option
279 queue_run_in_order means we cannot do this, as ordering becomes
282 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
283 had introduced a string-copy (for ensuring NUL-termination) which was not
284 appropriate for that case, which can include embedded NUL bytes in the
285 block of data. Investigation showed the copy to actually be needless, the
286 data being length-specified.
288 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
289 done during a receiving connection, and both used TLS, global info was
290 used rather than per-connection info for tracking the state of data
291 queued for transmission. This could result in a connection hang.
293 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
294 Previously, when delivering serveral messages down a single connection
295 only the first would provide a SIZE. This was due to the size information
296 not being properly tracked.
298 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
299 TAI (at 37 seconds currently), pretend to be in UTC for time-related
300 expansion and logging. Previously, spurious values such as a future
301 minute could be seen.
303 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
304 it could crash from a null-deref. This could also affect the
305 ${addresses: } operator and ${readsock } item.
307 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
308 message following a mime one, the variable was not reset.
310 JH/30 When an pipelined-connect fails at the first response, assume incorrect
311 cached capability (perhaps the peer reneged?) and immediately retry in
314 JH/31 Fix spurious detection of timeout while writing to transport filter.
316 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
317 an attempt to copy the string was made before checking it.
319 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
320 taint of the lookup key was maintained; we now regard the presence in the
321 filesystem as sufficient validation.
323 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
324 argument is supplied.
326 JH/35 The dsearch lookup now requires that the directory is an absolute path.
327 Previously this was not checked, and nonempty relative paths made an
328 access under Exim's current working directory.
330 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
331 Previously no event was raised.
333 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
334 parameter supplied by the sender MAIL FROM command. Previously it was
335 ignored, and only the check_spool_space option value for the required
338 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
339 the size of the signing public-key. Previously it was instead giving
340 the size of the signature hash.
342 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
343 the default. See the (new) dkim_verify_min_keysizes option.
345 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
346 and an ACL use a lookup for checking either the local_part or domain,
347 stale data could be accessed. Ensure that variable references are
348 dropped between messages.
350 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
351 by the client was not checked as pointing within response data before
352 being used. A malicious client could thus cause an out-of-bounds read and
353 possibly gain authentication. Fix by adding the check.
355 JH/42 Internationalisation: change the default for downconversion in the smtp
356 transport to be "if needed". Previously it was "as previously set" for
357 the message, which usually meant "if needed" for message-submission but
358 "no" for everything else. However, MTAs have been seen using SMTPUTF8
359 even when the envelope addresses did not need it, resulting in forwarding
360 failures to non-supporting MTAs. A downconvert in such cases will be
361 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
362 transport. The change does mean that addresses needing conversion will
363 be converted when previously a delivery failure would occur.
365 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
366 response was received it would be used unchecked in a fail-DSN, violating
367 standards on line-length limits. Truncate if needed.
369 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
370 doesn't work. (Additionally add a new main config option to configure the
377 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
378 side, to decode a TLS 1.3 packet capture.
380 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
381 Previously the default library behaviour applied, sending two, each in
384 JH/03 Debug output for ACL now gives the config file name and line number for
387 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
389 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
391 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
392 buffer overrun for (non-chunking) other transports.
394 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
395 TLS1.3, means that a server rejecting a client certificate is not visible
396 to the client until the first read of encrypted data (typically the
397 response to EHLO). Add detection for that case and treat it as a failed
398 TLS connection attempt, so that the normal retry-in-clear can work (if
399 suitably configured).
401 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
402 and/or domain. Found and fixed by Jason Betts.
404 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
405 configuration). If a CNAME target was not a wellformed name pattern, a
408 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
409 the OS reports them interleaved with other addresses.
411 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
412 used both for input and for a verify callout, both encrypted, SMTP
413 responses being sent by the server could be lost. This resulted in
414 dropped connections and sometimes bounces generated by a peer sending
417 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
418 string. Previously it was possible to cause undefined behaviour in a
419 library routine (usually a crash). Found by "zerons".
421 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
424 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
425 API was removed, so update to use the newer ones.
427 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
428 any timeout set, is taking a long time. Previously we would hang on to a
429 rotated logfile "forever" if the input was arriving with long gaps
430 (a previous attempt to fix addressed lack, for a long time, of initial
433 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
434 shared (NFS) environment. The length of the tempfile name is now
435 4 + 16 ("hdr.$message_exim_id") which might break on file
436 systems which restrict the file name length to lower values.
439 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
440 shared (NFS) environment.
442 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
443 did for all versions <4.90). Notably -M, -m, --invert, -I may be
446 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
447 on some platforms for bit 31.
449 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
450 to changes apparently associated with TLS1.3 handling some of the APIs
451 previously used were either nonfunctional or inappropriate. Strings
452 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
453 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
454 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
455 This affects log line X= elements, the $tls_{in,out}_cipher variables,
456 and the use of specific cipher names in the encrypted= ACL condition.
458 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
460 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
461 verification result was not updated unless hosts_require_ocsp applied.
463 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
464 queue_list_requires_admin set to false, non-admin users were denied the
467 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
468 directory-of-certs mode. Previously they were advertised despite the
471 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
472 A single TCP connection by a client will now hold a TLS connection open
473 for multiple message deliveries, by default. Previously the default was to
476 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
477 default. If built with the facility, DANE will be used. The facility
478 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
480 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
481 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
482 must be defined and you must still, unless you define DISABLE_TLS, manage
483 the the include-dir and library-file requirements that go with that
484 choice. Non-TLS builds are still supported.
486 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
489 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
490 default. If the platform supports and has the facility enabled, it will
491 be requested on all coneections.
493 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
494 controlled by the build-time option SUPPORT_PIPE_CONNECT.
496 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
498 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
499 success-DSN messages. Previously the From: header was always the default
500 one for these; the option was ignored.
502 JH/28 Fix the timeout on smtp response to apply to the whole response.
503 Previously it was reset for every read, so a teergrubing peer sending
504 single bytes within the time limit could extend the connection for a
505 long time. Credit to Qualsys Security Advisory Team for the discovery.
507 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
508 delivery address, which leaked information of the results of local
509 forwarding. Change to the original envelope recipient address, per
512 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
513 requested. Previously not bounce was generated and a log entry of
514 error ignored was made.
516 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
518 JH/32 Introduce a general tainting mechanism for values read from the input
519 channel, and values derived from them. Refuse to expand any tainted
520 values, to catch one form of exploit.
522 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
523 was unused and the unexpanded text used for the test. Found and
524 fixed by Ruben Jenster.
526 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
527 an attempt to use a TLS library read routine dereffed a nul pointer,
530 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
531 them in our smtp response.
533 JH/36 Have the general router option retry_use_local_part default to true when
534 any of the restrictive preconditions are set (to anything). Previously it
535 was only for check_local user. The change removes one item of manual
536 configuration which is required for proper retries when a remote router
537 handles a subset of addresses for a domain.
539 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
540 link count into consideration.
542 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
543 caused the extension of big_buffer, the following lines were ignored.
545 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
546 accordance with RFC 2308. Previously there was no expiry, so a longlived
547 receive process (eg. due to ACL delays) versus a short SOA value could
550 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
552 JH/39 Promote DMARC support to mainline.
554 JH/40 Bug 2452: Add a References: header to DSNs.
556 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
557 parameters. The relevant library call is documented as "Deprecated: This
558 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
559 3.6.0, DH parameters are negotiated following RFC7919."
561 HS/06 Change the default of dnssec_request_domains to "*"
563 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
564 carried on and emitted a BDAT command, even when PIPELINING was not
567 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
568 buffer was used for the filename, resulting in a trap when tainted
569 arguments (eg. $domain) were used.
571 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
572 recommended to avoid a possible server-load attack. The feature can be
573 re-enabled via the openssl_options main cofiguration option.
575 JH/45 local_scan API: documented the current smtp_printf() call. This changed
576 for version 4.90 - adding a "more data" boolean to the arguments.
577 Bumped the ABI version number also, this having been missed previously;
578 release versions 4.90 to 4.92.3 inclusive were effectively broken in
579 respect of usage of smtp_printf() by either local_scan code or libraries
580 accessed via the ${dlfunc } expansion item. Both will need coding
581 adjustment for any calls to smtp_printf() to match the new function
582 signature; a FALSE value for the new argument is always safe.
584 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
585 the file-offset (which the Linux syscall does, and exim expects); this
586 resulted in an indefinite loop.
588 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
589 to do ARC verification. The Authentication-Results: header line added
590 by the configuration then had no ARC item.
596 JH/01 Remove code calling the customisable local_scan function, unless a new
597 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
599 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
600 non-signal-safe functions being used.
602 JH/03 Bug 2269: When presented with a received message having a stupidly large
603 number of DKIM-Signature headers, disable DKIM verification to avoid
604 a resource-consumption attack. The limit is set at twenty.
606 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
607 report of oldest_pass in ${authres } in consequence, and separate out
608 some descriptions of reasons for verification fail.
610 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
611 files in the spool were present and unlocked. A queue-runner could spot
612 them, resulting in a duplicate delivery. Fix that by doing the unlock
613 after the unlink. Investigation by Tim Stewart. Take the opportunity to
614 add more error-checking on spoolfile handling while that code is being
617 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
618 No known attacks, no CVE, this is defensive hardening.
620 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
621 a queue-runner could start a delivery while other operations were ongoing.
622 Cutthrough delivery was a common victim, resulting in duplicate delivery.
623 Found and investigated by Tim Stewart. Fix by using the open message data
624 file handle rather than opening another, and not locally closing it (which
625 releases a lock) for that case, while creating the temporary .eml format
626 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
628 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
629 $sender_verify_failure/$recipient_verify_failure to "random".
631 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
634 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
635 Previously this would segfault.
637 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
640 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
641 like zero, since the resolver should be doing this for us, But we need one
642 as a CNAME but no MX presence gets the CNAME returned; we need to check
643 that doesn't point to an MX to declare it "no MX returned" rather than
644 "error, loop". A new main option is added so the older capability of
645 following some limited number of chain links is maintained.
647 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
649 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
650 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
651 now-deprecated earlier definitions used only specified the range up to TLS
652 1.2 (in the older-version library docs).
654 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
656 JH/15 Rework TLS client-side context management. Stop using a global, and
657 explicitly pass a context around. This enables future use of TLS for
658 connections to service-daemons (eg. malware scanning) while a client smtp
659 connection is using TLS; with cutthrough connections this is quite likely.
661 JH/16 Fix ARC verification to do AS checks in reverse order.
663 JH/17 Support a "tls" option on the ${readsocket } expansion item.
665 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
666 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
667 Previously the "utf8" would be re-prepended for every additional message.
669 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
670 Previously thery were accepted, resulting in issues when attempting to
671 forward messages to a non-supporting MTA.
673 PP/02 Let -n work with printing macros too, not just options.
675 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
676 one parent address was copied, and bogus data was used at delivery-logging
677 time. Either a crash (after delivery) or bogus log data could result.
678 Discovery and analysis by Tim Stewart.
680 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
681 Previously if the string ended mid-character, we did not insert the
682 promised '?' replacement.
684 PP/04 Documentation: current string operators work on bytes, not codepoints.
686 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
687 should pack well giving a smaller memory footprint so better caching and
688 therefore performance. Group the declarations where this can't be done so
689 that the byte-sized flag variables are not interspersed among pointer
690 variables, giving a better chance of good packing by the compiler.
692 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
693 non-null, to avoid issues with sites running BATV. Previously reports were
694 sent with an empty envelope sender so looked like bounces.
696 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
697 The ignore_error flag wasn't being returned from the filter subprocess so
698 was not set for later routers. Investigation and fix by Matthias Kurz.
700 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
701 and a msg:complete for the whole, when a message is manually removed using
702 -Mrm. Developement by Matthias Kurz, hacked on by JH.
704 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
705 a "Gnu special" function, asprintf() in the DB utility binary builds; I
706 hope that is portable enough.
708 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
709 requiring a known-CA anchor certificate; make it now rely entirely on the
710 TLSA as an anchor. Checking the name on the leaf cert against the name
711 on the A-record for the host is still done for TA (but not for EE mode).
713 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
714 included in delivery lines for non-proxied connections, when compiled with
715 SUPPORT_SOCKS and running with proxy logging enabled.
717 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
718 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
719 move the existing event to fire before the normal logging of message
720 failure so that custom logging is bracketed by normal logging.
722 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
723 msg:fail:internal event. Developement by Matthias Kurz.
725 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
726 far too small for todays use of crypto signatures stored there. Go all
727 the way to the max DNS message size of 64kB, even though this might be
728 overmuch for IOT constrained device use.
730 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
731 copy a string over itself. The library routine is documented as not
732 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
734 JH/32 For main options check_spool_space and check_inode_space, where the
735 platform supports 64b integers, support more than the previous 2^31 kB
736 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
737 the previous G, M, k.
739 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
740 $authenticated_fail_id variable on authentication failure. Previously
743 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
744 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
745 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
746 GNUTLS_SEC_PARAM_MEDIUM.
748 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
749 side. Previously we would continue as if no SNI had been received.
751 JH/36 Harden the handling of string-lists. When a list consisted of a sole
752 "<" character, which should be a list-separator specification, we walked
753 off past the nul-terimation.
755 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
756 causes) even when the retry time is not yet met. Previously they were
757 not, meaning that when (say) an account was over-quota and temp-rejecting,
758 and multiple senders' messages were queued, only one sender would get
759 notified on each configured delay_warning cycle.
761 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
763 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
764 cipher-suites, an error can be left on the stack even for a succeeding
765 accept; this results in impossible error messages when a later operation
768 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
769 return error codes indicating retry. Under TLS1.3 this becomes required.
771 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
772 it only wrote the new authenticators, resulting in a lack of tracking of
773 peer changes of ESMTP extensions until the next cache flush.
775 JH/41 Fix the loop reading a message header line to check for integer overflow,
776 and more-often against header_maxsize. Previously a crafted message could
777 induce a crash of the recive process; now the message is cleanly rejected.
779 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
780 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
786 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
787 When redis_servers is set to a list of > 1 element, and the Redis servers
788 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
789 case of MOVED into a DEFER case instead, thus moving the query onto the
790 next server in the list. For a cluster of N elements, all N servers must
791 be defined in redis_servers.
793 GF/02 Catch and remove uninitialized value warning in exiqsumm
794 Check for existence of @ARGV before looking at $ARGV[0]
796 JH/01 Replace the store_release() internal interface with store_newblock(),
797 which internalises the check required to safely use the old one, plus
798 the allocate and data copy operations duplicated in both (!) of the
799 extant use locations.
801 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
802 modifier. This matches the restriction on the commandline.
804 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
805 Previously only the last row was returned.
807 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
808 we assumed that tags in the header were well-formed, and parsed the
809 element content after inspecting only the first char of the tag.
810 Assumptions at that stage could crash the receive process on malformed
813 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
814 While running the DKIM ACL we operate on the Permanent memory pool so that
815 variables created with "set" persist to the DATA ACL. Also (at any time)
816 DNS lookups that fail create cache records using the Permanent pool. But
817 expansions release any allocations made on the current pool - so a dnsdb
818 lookup expansion done in the DKIM ACL releases the memory used for the
819 DNS negative-cache, and bad things result. Solution is to switch to the
820 Main pool for expansions.
821 While we're in that code, add checks on the DNS cache during store_reset,
822 active in the testsuite.
823 Problem spotted, and debugging aided, by Wolfgang Breyha.
825 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
826 When none of the hosts presented to a transport match an already-open
827 connection, close it and proceed with the list. Previously we would
828 queue the message. Spotted by Lena with Yahoo, probably involving
831 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
832 Previously a spurious "250 OK id=" response was appended to the proper
835 JH/08 The "support for" informational output now, which built with Content
836 Scanning support, has a line for the malware scanner interfaces compiled
837 in. Interface can be individually included or not at build time.
839 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
840 by the template makefile "src/EDITME". The "STREAM" support for an older
841 ClamAV interface method is removed.
843 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
844 rows affected is given instead).
846 JH/11 The runtime Berkeley DB library version is now additionally output by
847 "exim -d -bV". Previously only the compile-time version was shown.
849 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
850 SMTP connection. Previously, when one had more receipients than the
851 first, an abortive onward connection was made. Move to full support for
852 multiple onward connections in sequence, handling cutthrough connection
853 for all multi-message initiating connections.
855 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
856 routers. Previously, a multi-recipient message would fail to match the
857 onward-connection opened for the first recipient, and cause its closure.
859 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
860 a timeout on read on a GnuTLS initiating connection, resulting in the
861 initiating connection being dropped. This mattered most when the callout
862 was marked defer_ok. Fix to keep the two timeout-detection methods
865 JH/15 Relax results from ACL control request to enable cutthrough, in
866 unsupported situations, from error to silently (except under debug)
867 ignoring. This covers use with PRDR, frozen messages, queue-only and
870 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
872 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
873 metadata, resulting in a crash in free().
875 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
876 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
877 Broken also in d185889f4, with init system revamp.
879 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
880 Previously we abruptly closed the connection after reading a malware-
881 found indication; now we go on to read the "scan ok" response line,
884 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
885 ACL. Previously, a crash would result.
887 JH/19 Speed up macro lookups during configuration file read, by skipping non-
888 macro text after a replacement (previously it was only once per line) and
889 by skipping builtin macros when searching for an uppercase lead character.
891 JH/20 DANE support moved from Experimental to mainline. The Makefile control
892 for the build is renamed.
894 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
895 was allocated for every new TLS startup, meaning one per message. Fix
896 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
898 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
899 reported the original. Fix to report (as far as possible) the ACL
900 result replacing the original.
902 JH/23 Fix memory leak during multi-message connections using STARTTLS under
903 OpenSSL. Certificate information is loaded for every new TLS startup,
904 and the resources needed to be freed.
906 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
908 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
911 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
912 DATA response info to the (existing) per-recipient response info for
913 the "C=" log element. It can have useful tracking info from the
914 destination system. Patch from Simon Arlott.
916 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
917 length value. Previously this would segfault.
919 HS/02 Support Avast multiline protoocol, this allows passing flags to
920 newer versions of the scanner.
922 JH/28 Ensure that variables possibly set during message acceptance are marked
923 dead before release of memory in the daemon loop. This stops complaints
924 about them when the debug_store option is enabled. Discovered specifically
925 for sender_rate_period, but applies to a whole set of variables.
926 Do the same for the queue-runner and queue-list loops, for variables set
927 from spool message files. Do the same for the SMTP per-message loop, for
928 certain variables indirectly set in ACL operations.
930 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
931 as a multi-recipient message from a mailinglist manager). The coding had
932 an arbitrary cutoff number of characters while checking for more input;
933 enforced by writing a NUL into the buffer. This corrupted long / fast
934 input. The problem was exposed more widely when more pipelineing of SMTP
935 responses was introduced, and one Exim system was feeding another.
936 The symptom is log complaints of SMTP syntax error (NUL chars) on the
937 receiving system, and refused recipients seen by the sending system
938 (propating to people being dropped from mailing lists).
939 Discovered and pinpointed by David Carter.
941 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
942 replaced by the ${authresults } expansion.
944 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
946 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
947 allows proper process termination in container environments.
949 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
950 Previously the "final dot" had a newline after it; ensure it is CR,LF.
952 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
953 and "err_perm", deprecated since 4.83 when the RFC-defined words
954 "temperror" and "permerror" were introduced.
956 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
957 transport-filters or DKIM-signing. The restriction was lost in the
958 consolidation of verify-callout and delivery SMTP handling.
959 Extend the restriction to also cover ARC-signing.
961 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
962 in defer=pass mode supply a 450 to the initiator. Previously the message
965 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
966 tls_require_ciphers is used as before.
968 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
969 "pass_unscanned". Only tmpfails from the scanner are written to
970 the paniclog, as they may require admin intervention (permission
971 denied, license issues). Other scanner errors (like decompression
972 bombs) do not cause a paniclog entry.
974 JH/36 Fix reinitialisation of DKIM logging variable between messages.
975 Previously it was possible to log spurious information in receive log
978 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
979 triggered odd behaviour from Outlook Express clients.
981 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
984 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
985 since the IETF WG has not yet settled on that versus the original
986 "bare" representation.
988 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
989 Previously the millisecond value corrupted the output.
990 Fix also for syslog_pid=no and log_selector +pid, for which the pid
991 corrupted the output.
997 JH/01 Rework error string handling in TLS interface so that the caller in
998 more cases is responsible for logging. This permits library-sourced
999 string to be attached to addresses during delivery, and collapses
1000 pairs of long lines into single ones.
1002 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1003 during configuration. Wildcards are allowed and expanded.
1005 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1006 This permits better logging.
1008 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1009 do not close it down and have the child transport start it up again on
1010 the passed-on TCP connection. Instead, proxy the child (and any
1011 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1012 affected: the continued delivery log lines do not have any DNSSEC, TLS
1013 Certificate or OCSP information. TLS cipher information is still logged.
1015 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1016 identical IP addresses on different listening ports. Will also affect
1019 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1020 add noisy ifdef guards to special-case this sillyness.
1021 Patch from Bernd Kuhls.
1023 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1024 than 255 are no longer allowed.
1026 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1027 Disable the session-cache too, which might reduce our load. Since we
1028 currrectly use a new context for every connection, both as server and
1029 client, there is no benefit for these.
1030 GnuTLS appears to not support tickets server-side by default (we don't
1031 call gnutls_session_ticket_enable_server()) but client side is enabled
1032 by default on recent versions (3.1.3 +) unless the PFS priority string
1035 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1036 <https://reproducible-builds.org/specs/source-date-epoch/>.
1038 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1039 the check for any unsuccessful recipients did not notice the limit, and
1040 erroneously found still-pending ones.
1042 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1043 MSG_MORE. Only in-clear (not on TLS connections).
1045 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1046 transport-filter is involved we can buffer the headers in memory for
1047 creating the signature, and read the spool data file once for the
1048 signature and again for transmission.
1050 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1051 4.77 as the kernel support then wasn't solid, having issues in 64bit
1052 mode. Now, it's been long enough. Add support for FreeBSD also.
1054 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1055 case where the routing stage had gathered several addresses to send to
1056 a host before calling the transport for the first, we previously failed
1057 to close down TLS in the old transport process before passing the TCP
1058 connection to the new process. The new one sent a STARTTLS command
1059 which naturally failed, giving a failed delivery and bloating the retry
1060 database. Investigation and fix prototype from Wolfgang Breyha.
1062 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1063 false-negatives in the check that the sender had not preempted a response
1064 or prompt from Exim (running as a server), due to that code's lack of
1065 awareness of the SMTP input buffering.
1067 PP/04 Add commandline_checks_require_admin option.
1068 Exim drops privileges sanely, various checks such as -be aren't a
1069 security problem, as long as you trust local users with access to their
1070 own account. When invoked by services which pass untrusted data to
1071 Exim, this might be an issue. Set this option in main configuration
1072 AND make fixes to the calling application, such as using `--` to stop
1075 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1076 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1077 responses to those, into a single TLS record each way (this usually means
1078 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1081 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1082 affects you only if you're dancing at the edge of the param size limits.
1083 If you are, and this message makes sense to you, then: raise the
1084 configured limit or use OpenSSL 1.1. Nothing we can do for older
1087 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1088 cmdline element to get the documented default one. Previously it was
1091 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1092 are unsuable for non-message-specific reasons - eg. connection timeout,
1093 banner-time rejection.
1095 JH/16 Fix logging of delivery remote port, when specified by router, under
1098 PP/06 Repair manualroute's ability to take options in any order, even if one
1099 is the name of a transport.
1102 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1104 JH/17 Change the list-building routines interface to use the expanding-string
1105 triplet model, for better allocation and copying behaviour.
1107 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1108 Previously it was constructed the first time a possibly-matching string
1109 was met in the configuration file input during startup; now it is done
1112 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1113 compatible one, to avoid the (poorly documented) possibility of a config
1114 file in the working directory redirecting the DB files, possibly correpting
1115 some existing file. CVE-2017-10140 assigned for BDB.
1117 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1118 cache-hot. Previously, although the result was properly cached, the
1119 initial verify call returned a defer.
1121 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1122 the main verify for receipient in uncached-mode.
1124 JH/22 Retire historical build files to an "unsupported" subdir. These are
1125 defined as "ones for which we have no current evidence of testing".
1127 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1128 if present. Previously it was ignored.
1130 JH/24 Start using specified-initialisers in C structure init coding. This is
1131 a C99 feature (it's 2017, so now considered safe).
1133 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1134 if was a fixed-sized field and bitmask ops via macros; it is now more
1137 PP/07 GitHub PR 56: Apply MariaDB build fix.
1138 Patch provided by Jaroslav Škarvada.
1140 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1141 during Coverity cleanups [4.87 JH/47]
1142 Diagnosis and fix provided by Michael Fischer v. Mollard.
1144 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1145 the right size to place the terminating semicolon on its own folded
1146 line, the header hash was calculated to an incorrect value thanks to
1147 the (relaxed) space the fold became.
1149 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1150 and confused the parent.
1152 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1153 which could crash as a result. This could lead to undeliverable messages.
1155 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1158 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1159 matters on fast-turnover and PID-randomising systems, which were getting
1160 out-of-order delivery.
1162 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1163 a possibly-overlapping copy. The symptom was that "Remote host closed
1164 connection in response to HELO" was logged instead of the actual 4xx
1167 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1168 Previously only that bufferd was discarded, resulting in SYMTP command
1171 JH/32 DKIM: when a message has multiple signatures matching an identity given
1172 in dkim_verify_signers, run the dkim acl once for each. Previously only
1173 one run was done. Bug 2189.
1175 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1176 "panic the current process" to "deliberately defer". The panic log is
1177 still written with the problem list name; the mail and reject logs now
1178 get a temp-reject line for the message that was being handled, saying
1179 something like "domains check lookup or other defer". The SMTP 451
1180 message is still "Temporary local problem".
1182 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1183 A crafted sequence of BDAT commands could result in in-use memory beeing
1184 freed. CVE-2017-16943.
1186 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1187 from SMTP input. Previously it was always done; now only done for DATA
1188 and not BDAT commands. CVE-2017-16944.
1190 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1191 to the message (such as an overlong header line). Previously this was
1192 not done and we did not exit BDAT mode. Followon from the previous item
1193 though a different problem.
1199 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1200 than -2003 did; needs libidn2 in addition to libidn.
1202 JH/02 The path option on a pipe transport is now expanded before use.
1204 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1205 Patch provided by "Björn", documentation fix added too.
1207 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1208 missing a wire-to-host endian conversion.
1210 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1211 close after a BDAT command line could be taken as a following command,
1212 giving a synch failure. Fix by only checking for synch immediately
1213 before acknowledging the chunk.
1215 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1216 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1217 macro. Patches provided by Josh Soref.
1219 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1220 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1221 by IANA as a service extension. However, John Klensin suggests that we
1224 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1225 the dkim code may be unix-mode line endings rather than smtp wire-format
1226 CRLF, so prepend a CR to any bare LF.
1228 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1229 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1231 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1232 the first were themselves being wrongly included in the feed into dkim
1233 processing; with most chunk sizes in use this resulted in an incorrect
1234 body hash calculated value.
1236 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1237 DKIM signature block, for verification. Although advised against by
1238 standards it is specifically not ruled illegal.
1240 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1242 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1243 missing a body hash (the bh= tag).
1245 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1246 It seems that HAProxy sends the Proxy Protocol information in clear and
1247 only then does a TLS startup, so do the same.
1249 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1250 TCP connections (such as for Spamd) unless the daemon successfully set
1251 Fast Open mode on its listening sockets. This fixes breakage seen on
1252 too-old kernels or those not configured for Fast Open, at the cost of
1253 requiring both directions being enabled for TFO, and TFO never being used
1254 by non-daemon-related Exim processes.
1256 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1257 endings, at least on the first header line. Try to canonify any that get
1258 past that check, despite the cost.
1260 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1261 now limited to an arbitrary five deep, while parsing addresses with the
1262 strip_excess_angle_brackets option enabled.
1264 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1265 instead leave the unprompted TLS handshake in socket buffer for the
1266 TLS library to consume.
1268 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1270 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1272 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1273 region in one operation, for speed, and this leaves assigned pointers
1274 dangling. Add checks run only under the testsuite which checks all
1275 variables at a store-reset and panics on a dangling pointer; add code
1276 explicitly nulling out all the variables discovered. Fixes one known
1277 bug: a transport crash, where a dangling pointer for $sending_ip_address
1278 originally assigned in a verify callout, is re-used.
1280 PP/06 Drop '.' from @INC in various Perl scripts.
1282 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1284 PP/08 Reduce a number of compilation warnings under clang; building with
1285 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1286 should be warning-free.
1288 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1290 HS/01 Fix portability problems introduced by PP/08 for platforms where
1291 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1293 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1294 chunk. This allows us to accept broken chunked messages. We need a more
1295 general solution here.
1297 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1298 already-broken messages in the queue.
1300 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1302 JH/19 Fix reference counting bug in routing-generated-address tracking.
1308 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1309 supports it and a size is available (ie. the sending peer gave us one).
1311 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1312 years of being deprecated). The replacements are the ACLs
1313 acl_smtp_mime and acl_not_smtp_mime.
1315 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1316 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1317 or even an in-clear connection were permitted. Now, if the host lookup
1318 was dnssec and dane was requested then the host is only used if the
1319 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1320 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1321 if one fails this test.
1322 This means that a poorly-configured remote DNS will make it incommunicado;
1323 but it protects against a DNS-interception attack on it.
1325 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1326 non-noisy when a race steals the message being considered.
1328 JH/05 If main configuration option tls_certificate is unset, generate a
1329 self-signed certificate for inbound TLS connections.
1331 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1332 in rewrites and routers.
1334 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1335 and logged a warning sing 4.83; now they are a configuration file error.
1337 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1338 (lacking @domain). Apply the same qualification processing as RCPT.
1340 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1342 JH/10 Support ${sha256:} applied to a string (as well as the previous
1345 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1346 a cutthrough deliver is pending, as we always want to make a connection.
1347 This also avoids re-routing the message when later placing the cutthrough
1348 connection after a verify cache hit.
1349 Do not update it with the verify result either.
1351 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1352 when routing results in more than one destination address.
1354 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1355 signing (which inhibits the cutthrough capability). Previously only
1356 the presence of an option was tested; now an expansion evaluating as
1357 empty is permissible (obviously it should depend only on data available
1358 when the cutthrough connection is made).
1360 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1361 the relevant preceding SMTP command did not note the pipelining mode.
1363 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1364 Previously they were not counted.
1366 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1367 as one having no matching records. Previously we deferred the message
1368 that needed the lookup.
1370 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1371 distinguished as "(=".
1373 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1374 for missing MX records. Previously it only worked for missing A records.
1376 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1378 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1379 after the data-go-ahead and data-ack. Patch from Jason Betts.
1381 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1382 even for a "none" policy. Patch from Tony Meyer.
1384 JH/22 Fix continued use of a connection for further deliveries. If a port was
1385 specified by a router, it must also match for the delivery to be
1388 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1389 When one of the recipients of a message was unsuitable for the connection
1390 (has no matching addresses), we lost track of needing to mark it
1391 deferred. As a result mail would be lost.
1393 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1395 JH/25 Decoding ACL controls is now done using a binary search; the source code
1396 takes up less space and should be simpler to maintain. Merge the ACL
1397 condition decode tables also, with similar effect.
1399 JH/26 Fix problem with one_time used on a redirect router which returned the
1400 parent address unchanged. A retry would see the parent address marked as
1401 delivered, so not attempt the (identical) child. As a result mail would
1404 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1405 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1406 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1409 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1410 Main config options check_{log,spool}_{inodes,space} are now
1411 100 inodes, 10MB unless set otherwise in the configuration.
1413 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1414 Previously it only applied to the main-section connection policy
1417 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1419 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1420 by me. Added RFC7919 DH primes as an alternative.
1422 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1423 are not in the system include path.
1425 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1426 GnuTLS, when a session startup failed (eg because the client disconnected)
1427 Exim did stdio operations after fclose. This was exposed by a recent
1428 change which nulled out the file handle after the fclose.
1430 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1431 signed directly by the cert-signing cert, rather than an intermediate
1432 OCSP-signing cert. This is the model used by LetsEncrypt.
1434 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1436 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1437 an incoming connection.
1439 HS/02 Bug 1802: Do not half-close the connection after sending a request
1442 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1443 fallback to "prime256v1".
1445 JH/34 SECURITY: Use proper copy of DATA command in error message.
1446 Could leak key material. Remotely exploitable. CVE-2016-9963.
1452 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1453 and 3.4.4 - once the server is enabled to respond to an OCSP request
1454 it does even when not requested, resulting in a stapling non-aware
1455 client dropping the TLS connection.
1457 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1458 support variable-length bit vectors. No functional change.
1460 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1461 The I= interface field on outgoing lines is now after the H= remote
1462 host field, same as incoming lines. There is a separate
1463 outgoing_interface log selector which allows you to disable the
1466 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1467 If not running log_selector +smtp_connection the mainlog would be held
1468 open indefinitely after a "too many connections" event, including to a
1469 deleted file after a log rotate. Leave the per net connection logging
1470 leaving it open for efficiency as that will be quickly detected by the
1471 check on the next write.
1473 HS/01 Bug 1671: Fix post transport crash.
1474 Processing the wait-<transport> messages could crash the delivery
1475 process if the message IDs didn't exist for some reason. When
1476 using 'split_spool_directory=yes' the construction of the spool
1477 file name failed already, exposing the same netto behaviour.
1479 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1480 mime_regex ACL conditions.
1482 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1483 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1484 to HELO, local diagnostic string.
1486 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1487 log line to debug. Even when configured with a tls authenticator many
1488 client connections are expected to not authenticate in this way, so
1489 an authenticate fail is not an error.
1491 HS/02 Add the Exim version string to the process info. This way exiwhat
1492 gives some more detail about the running daemon.
1494 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1495 matter for fast-change records such as DNSBLs.
1497 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1498 retry record, even if constant. There may be multiple transports with
1499 different interface settings and the retry behaviour needs to be kept
1502 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1504 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1506 JH/10 Bug 840: fix log_defer_output option of pipe transport
1508 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1509 hosts_randomize. This matters a lot when combined with mua_wrapper.
1511 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1512 ${quote_pgsql:<string>} operator.
1514 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1516 JH/14 Change header file name for freeradius-client. Relevant if compiling
1517 with Radius support; from the Gentoo tree and checked under Fedora.
1519 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1521 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1522 option transport_filter, meaning no filtering.
1524 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1526 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1527 defaults to "*" (all hosts). The variable is now available when not built
1528 with TLS, default unset, mainly to enable keeping the testsuite sane.
1529 If a server certificate is not supplied (via tls_certificate) an error is
1530 logged, and clients will find TLS connections fail on startup. Presumably
1531 they will retry in-clear.
1532 Packagers of Exim are strongly encouraged to create a server certificate
1533 at installation time.
1535 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1536 with the $config_file variable.
1538 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1539 in transport context, after the attempt, and per-recipient. The latter type
1540 is per host attempted. The event data is the error message, and the errno
1541 information encodes the lookup type (A vs. MX) used for the (first) host,
1542 and the trailing two digits of the smtp 4xx response.
1544 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1545 to write to mainlog (or rejectlog, paniclog) in the window between file
1546 creation and permissions/ownership being changed. Particularly affects
1547 installations where exicyclog is run as root, rather than exim user;
1548 result is that the running daemon panics and dies.
1550 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1552 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1553 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1554 "pri" and "weight". Note that the previous implicit priority given by the
1555 list order is no longer honoured.
1557 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1558 for DKIM processing.
1560 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1561 by defining SUPPORT_SOCKS.
1563 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1564 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1565 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1566 variables are renamed to proxy_{local,external}_{address,port}.
1568 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1569 for a build by defining SUPPORT_I18N
1571 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1572 of the query string, and make ${quote_redis:} do that quoting.
1574 JH/29 Move Events support from Experimental to mainline, enabled by default
1575 and removable for a build by defining DISABLE_EVENT.
1577 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1579 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1580 cached by the daemon.
1582 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1583 by defining LOOKUP_REDIS. The libhiredis library is required.
1585 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1586 keys are given for lookup.
1588 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1589 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1590 only supported when built with TLS support. The PolarSSL SHA routines
1591 are still used when the TLS library is too old for convenient support.
1593 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1594 openssl_options), for security. OpenSSL forces this from version 1.1.0
1595 server-side so match that on older versions.
1597 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
1598 allocation for $value could be released as the expansion processing
1599 concluded, but leaving the global pointer active for it.
1601 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
1602 and to use the domains and local_parts ACL conditions.
1604 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
1605 incorrectly not doubled on cutthrough transmission, hence seen as a
1606 body-termination at the receiving system - resulting in truncated mails.
1607 Commonly the sender saw a TCP-level error, and retransmitted the message
1608 via the normal store-and-forward channel. This could result in duplicates
1609 received - but deduplicating mailstores were liable to retain only the
1610 initial truncated version.
1612 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
1614 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
1616 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
1617 we're in there, support oversigning also; bug 1309.
1619 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
1621 HS/04 Add support for keep_environment and add_environment options.
1623 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
1624 either intentional arithmetic overflow during PRNG, or testing config-
1627 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
1628 delivery resulted in actual delivery. Cancel cutthrough before DATA
1631 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
1632 rejects a recipient: pass the reject to the originator.
1634 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
1635 Many were false-positives and ignorable, but it's worth fixing the
1638 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
1639 for the new environment-manipulation done at startup. Move the routines
1640 from being local to tls.c to being global via the os.c file.
1642 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
1643 an extract embedded as result-arg for a map, the first arg for extract
1644 is unavailable so we cannot tell if this is a numbered or keyed
1645 extraction. Accept either.
1651 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
1654 JH/02 The smtp transport option "multi_domain" is now expanded.
1656 JH/03 The smtp transport now requests PRDR by default, if the server offers
1659 JH/04 Certificate name checking on server certificates, when exim is a client,
1660 is now done by default. The transport option tls_verify_cert_hostnames
1661 can be used to disable this per-host. The build option
1662 EXPERIMENTAL_CERTNAMES is withdrawn.
1664 JH/05 The value of the tls_verify_certificates smtp transport and main options
1665 default to the word "system" to access the system default CA bundle.
1666 For GnuTLS, only version 3.0.20 or later.
1668 JH/06 Verification of the server certificate for a TLS connection is now tried
1669 (but not required) by default. The verification status is now logged by
1670 default, for both outbound TLS and client-certificate supplying inbound
1673 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
1676 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
1677 Status Notification (bounce) messages are now MIME format per RFC 3464.
1678 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
1679 under the control of the dsn_advertise_hosts option, and routers may
1680 have a dsn_lasthop option.
1682 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
1683 default, modifiable by a malware= option. The list separator for
1684 the options can now be changed in the usual way. Bug 68.
1686 JH/10 The smtp_receive_timeout main option is now expanded before use.
1688 JH/11 The incoming_interface log option now also enables logging of the
1689 local interface on delivery outgoing connections.
1691 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
1692 if the interface and destination host and port all match.
1694 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
1697 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
1698 Patch from Andrew Lewis.
1700 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
1701 now supports optional time-restrictions, weighting, and priority
1702 modifiers per server. Patch originally by <rommer@active.by>.
1704 JH/16 The spamd_address main option now supports a mixed list of local
1705 and remote servers. Remote servers can be IPv6 addresses, and
1706 specify a port-range.
1708 JH/17 Bug 68: The spamd_address main option now supports an optional
1709 timeout value per server.
1711 JH/18 Bug 1581: Router and transport options headers_add/remove can
1712 now have the list separator specified.
1714 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
1717 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
1720 JH/21 Support for the A6 type of dns record is withdrawn.
1722 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
1723 rather than the verbs used.
1725 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
1726 from 255 to 1024 chars.
1728 JH/24 Verification callouts now attempt to use TLS by default.
1730 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
1731 are generic router options now. The defaults didn't change.
1733 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1734 Original patch from Alexander Shikoff, worked over by JH.
1736 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1737 files if ZCAT_COMMAND is not executable.
1739 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1741 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1743 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1744 Normally benign, it bites when the pair was led to by a CNAME;
1745 modern usage is to not canonicalize the domain to a CNAME target
1746 (and we were inconsistent anyway for A-only vs AAAA+A).
1748 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1750 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1751 when evaluating $sender_host_dnssec.
1753 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1754 $sender_helo_dnssec variable.
1756 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1758 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1760 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1762 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1763 documented as working, but never had. Support all but $spam_report.
1765 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1766 added for tls authenticator.
1768 HS/03 Add perl_taintmode main config option
1774 TL/01 When running the test suite, the README says that variables such as
1775 no_msglog_check are global and can be placed anywhere in a specific
1776 test's script, however it was observed that placement needed to be near
1777 the beginning for it to behave that way. Changed the runtest perl
1778 script to read through the entire script once to detect and set these
1779 variables, reset to the beginning of the script, and then run through
1780 the script parsing/test process like normal.
1782 TL/02 The BSD's have an arc4random API. One of the functions to induce
1783 adding randomness was arc4random_stir(), but it has been removed in
1784 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1785 function when detected.
1787 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1788 cause callback expansion.
1790 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1791 syntax errors in an expansion can be treated as a string instead of
1792 logging or causing an error, due to the internal use of bool_lax
1793 instead of bool when processing it.
1795 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1796 server certificates when making smtp deliveries.
1798 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1800 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1802 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1804 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1805 Merged patch from Sebastian Wiedenroth.
1807 JH/05 Fix results-pipe from transport process. Several recipients, combined
1808 with certificate use, exposed issues where response data items split
1809 over buffer boundaries were not parsed properly. This eventually
1810 resulted in duplicates being sent. This issue only became common enough
1811 to notice due to the introduction of connection certificate information,
1812 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1814 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1815 size buffer was used, resulting in syntax errors when an expansion
1818 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1819 version 3.3.6 or later.
1821 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1822 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1823 both become "event_action", the variables become $event_name, $event_data
1824 and $event_defer_errno. There is a new variable $verify_mode, usable in
1825 routers, transports and related events. The tls:cert event is now also
1826 raised for inbound connections, if the main configuration event_action
1829 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1830 early OCSP support, but no stapling (appears to be less than 1.0.0).
1832 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1833 server certificate names available under the smtp transport option
1834 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1837 JH/10 Time-related extraction expansions from certificates now use the main
1838 option "timezone" setting for output formatting, and are consistent
1839 between OpenSSL and GnuTLS compilations. Bug 1541.
1841 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1842 encoded parameter in the incoming message. Bug 1558.
1844 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1845 include certificate info, eximon was claiming there were spoolfile
1848 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1850 JH/14 Log delivery-related information more consistently, using the sequence
1851 "H=<name> [<ip>]" wherever possible.
1853 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1854 are problematic for Debian distribution, omit them from the release
1857 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1859 JH/16 Fix string representation of time values on 64bit time_t architectures.
1862 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1863 output list separator was used.
1868 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1869 checkers that were complaining about end of non-void function with no
1872 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1873 This was a regression introduced in 4.83 by another bugfix.
1875 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1877 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1878 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1884 TF/01 Correctly close the server side of TLS when forking for delivery.
1886 When a message was received over SMTP with TLS, Exim failed to clear up
1887 the incoming connection properly after forking off the child process to
1888 deliver the message. In some situations the subsequent outgoing
1889 delivery connection happened to have the same fd number as the incoming
1890 connection previously had. Exim would try to use TLS and fail, logging
1891 a "Bad file descriptor" error.
1893 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1894 utilities have not been installed.
1896 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1897 temporary space as the ACL may create new global variables.
1899 TL/01 LDAP support uses per connection or global context settings, depending
1900 upon the detected version of the libraries at build time.
1902 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1903 to extract and use the src ip:port in logging and expansions as if it
1904 were a direct connection from the outside internet. PPv2 support was
1905 updated based on HAProxy spec change in May 2014.
1907 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1909 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1910 Properly escape header and check for NULL return.
1912 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1915 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1917 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1918 characters in header names, implemented as a verify condition.
1919 Contributed by Michael Fischer v. Mollard.
1921 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1922 results permerror and temperror. Previous values are deprecated but
1923 still accepted. In a future release, err_perm and err_temp will be
1924 completely removed, which will be a backward incompatibility if the
1925 ACL tests for either of these two old results. Patch contributed by
1926 user bes-internal on the mailing list.
1928 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1930 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1931 selectors, in both main and reject logs.
1933 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1936 JH/07 Add malware type "sock" for talking to simple daemon.
1938 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1940 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1941 routers/transports under cutthrough routing.
1943 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1944 numbers. Touch up "bool" conditional to keep the same definition.
1946 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1948 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1950 JH/12 Expand items in router/transport headers_add or headers_remove lists
1951 individually rather than the list as a whole. Bug 1452.
1953 Required for reasonable handling of multiple headers_ options when
1954 they may be empty; requires that headers_remove items with embedded
1955 colons must have them doubled (or the list-separator changed).
1957 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1958 view the policy declared in the DMARC record. Currently, $dmarc_status
1959 is a combined value of both the record presence and the result of the
1962 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1964 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1965 dnslookup router and the smtp transport (applying to the forward
1968 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1969 of ldap servers used for a specific lookup. Patch provided by Heiko
1972 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1973 New variable $lookup_dnssec_authenticated for observability.
1975 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1976 Patch submitted by Lars Timman.
1978 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1980 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1981 Requires trusted mode and valid format message id, aborts otherwise.
1982 Patch contributed by Heiko Schlichting.
1984 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1985 certextract with support for various fields. Bug 1358.
1987 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1988 is requested by default, modifiable by smtp transport option
1991 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1992 operate on certificate variables to give certificate fingerprints
1993 Also new ${sha256:cert_variable}.
1995 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1997 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1998 Christian Aistleitner.
2000 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2002 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2003 file. Patch from Wolfgang Breyha.
2005 JH/25 Expand the coverage of the delivery $host and $host_address to
2006 client authenticators run in verify callout. Bug 1476.
2008 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2009 align with daemon_smtp_ports. Bug 72.
2011 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2012 support and error reporting did not work properly.
2014 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2015 and is readable. Patch from Andrew Colin Kissa.
2017 TL/14 Enhance documentation of ${run expansion and how it parses the
2018 commandline after expansion, particularly in the case when an
2019 unquoted variable expansion results in an empty value.
2021 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2023 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2024 by renaming away from it. Addresses GH issue 15, reported by
2027 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2029 TL/15 SECURITY: prevent double expansion in math comparison functions
2030 (can expand unsanitized data). Not remotely exploitable.
2037 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2039 PP/02 Make -n do something, by making it not do something.
2040 When combined with -bP, the name of an option is not output.
2042 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2045 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2046 $sender_host_name and config options to manage this, and basic check
2049 PP/05 DSCP support for outbound connections and control modifier for inbound.
2051 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2052 (Only plugin which currently uses this is kerberos4, which nobody should
2053 be using, but we should make it available and other future plugins might
2054 conceivably use it, even though it would break NAT; stuff *should* be
2055 using channel bindings instead).
2057 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2058 name; added for Sendmail compatibility; requires admin caller.
2059 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2060 just ignore it); requires trusted caller.
2061 Also parse but ignore: -Ac -Am -X<logfile>
2064 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2066 TL/02 Add +smtp_confirmation as a default logging option.
2068 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2069 Patch by Magnus Holmgren from 2007-02-20.
2071 TL/04 Bugzilla 1281 - Spec typo.
2072 Bugzilla 1283 - Spec typo.
2073 Bugzilla 1290 - Spec grammar fixes.
2075 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2077 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2079 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2080 mailing list by by Dmitry Isaikin.
2082 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2084 JH/02 Support "G" suffix to numbers in ${if comparisons.
2086 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2088 NM/01 Bugzilla 1197 - Spec typo
2089 Bugzilla 1196 - Spec examples corrections
2091 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2093 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2094 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2097 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2098 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2100 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2101 "acl {{name}{arg}...}", and optional args on acl condition
2104 JH/05 Permit multiple router/transport headers_add/remove lines.
2106 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2108 JH/07 Avoid using a waiting database for a single-message-only transport.
2109 Performance patch from Paul Fisher. Bugzilla 1262.
2111 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2114 JH/09 Add $headers_added variable, with content from use of ACL modifier
2115 add_header (but not yet added to the message). Bugzilla 199.
2117 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2118 Pulled from Bugzilla 817 by Wolfgang Breyha.
2120 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2122 (nb: this is the same fix as in Exim 4.80.1)
2124 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2127 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2130 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2132 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2133 advertise SMTP AUTH mechanism to us, instead of a generic
2134 protocol violation error. Also, make Exim more robust to bad
2135 data from the Dovecot auth socket.
2137 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2139 When a queue runner is handling a message, Exim first routes the
2140 recipient addresses, during which it prunes them based on the retry
2141 hints database. After that it attempts to deliver the message to
2142 any remaining recipients. It then updates the hints database using
2145 So if a recipient address works intermittently, it can get repeatedly
2146 deferred at routing time. The retry hints record remains fresh so the
2147 address never reaches the final cutoff time.
2149 This is a fairly common occurrence when a user is bumping up against
2150 their storage quota. Exim had some logic in its local delivery code
2151 to deal with this. However it did not apply to per-recipient defers
2152 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2154 This change adds a proper retry rule check during routing so that the
2155 final cutoff time is checked against the message's age. We only do
2156 this check if there is an address retry record and there is not a
2157 domain retry record; this implies that previous attempts to handle
2158 the address had the retry_use_local_parts option turned on. We use
2159 this as an approximation for the destination being like a local
2160 delivery, as in LMTP.
2162 I suspect this new check makes the old local delivery cutoff check
2163 redundant, but I have not verified this so I left the code in place.
2165 TF/02 Correct gecos expansion when From: is a prefix of the username.
2167 Test 0254 submits a message to Exim with the header
2171 When I ran the test suite under the user fanf2, Exim expanded
2172 the header to contain my full name, whereas it should have added
2173 a Resent-Sender: header. It erroneously treats any prefix of the
2174 username as equal to the username.
2176 This change corrects that bug.
2178 GF/01 DCC debug and logging tidyup
2179 Error conditions log to paniclog rather than rejectlog.
2180 Debug lines prefixed by "DCC: " to remove any ambiguity.
2182 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2184 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2185 Bug spotted by Jeremy Harris; was flawed since initial commit.
2186 Would have resulted in OCSP responses post-SNI triggering an Exim
2187 NULL dereference and crash.
2189 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2191 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2192 Bug detection, analysis and fix by Samuel Thibault.
2193 Bugzilla 1331, Debian bug #698092.
2195 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2197 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2198 Server implementation by Todd Lyons, client by JH.
2199 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2200 config variable "prdr_enable" controls whether the server
2201 advertises the facility. If the client requests PRDR a new
2202 acl_data_smtp_prdr ACL is called once for each recipient, after
2203 the body content is received and before the acl_smtp_data ACL.
2204 The client is controlled by both of: a hosts_try_prdr option
2205 on the smtp transport, and the server advertisement.
2206 Default client logging of deliveries and rejections involving
2207 PRDR are flagged with the string "PRDR".
2209 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2210 fclose(). Diagnosis by Todd Lyons.
2212 PP/17 Update configure.default to handle IPv6 localhost better.
2213 Patch by Alain Williams (plus minor tweaks).
2216 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2217 This is now consistent with GnuTLS, and is now documented: the
2218 previous undocumented portable approach to treating the option as
2219 unset was to force an expansion failure. That still works, and
2220 an empty string is now equivalent.
2222 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2223 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2224 not performing validation itself.
2226 PP/20 Added force_command boolean option to pipe transport.
2227 Patch from Nick Koston, of cPanel Inc.
2229 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2232 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2234 PP/21 Fix eximon continuous updating with timestamped log-files.
2235 Broken in a format-string cleanup in 4.80, missed when I repaired the
2236 other false fix of the same issue.
2237 Report and fix from Heiko Schlichting.
2240 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2241 Report from Prashanth Katuri.
2243 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2244 It's SecureTransport, so affects any MacOS clients which use the
2245 system-integrated TLS libraries, including email clients.
2247 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2248 using a MIME ACL for non-SMTP local injection.
2249 Report and assistance in diagnosis by Warren Baker.
2251 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2253 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2255 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2256 last id that failed so it may be referenced in subsequent ACL's.
2258 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2261 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2262 ldap library initialization, allowing self-signed CA's to be
2263 used. Also properly sets require_cert option later in code by
2264 using NULL (global ldap config) instead of ldap handle (per
2265 session). Bug diagnosis and testing by alxgomz.
2267 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2268 the src/util/ subdirectory.
2270 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2271 renamed to Transport Post Delivery Action by Jeremy Harris, as
2274 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2275 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2276 redis_servers = needs to be configured which will be used by the redis
2277 lookup. Patch from Warren Baker, of The Packet Hub.
2279 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2281 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2282 hostname or reverse DNS when processing a host list. Used suggestions
2283 from multiple comments on this bug.
2285 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2287 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2288 Missed a few lines, added it to make the runtest require no keyboard
2291 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2292 contains upper case chars. Make router use caseful_local_part.
2294 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2295 support when GnuTLS has been built with p11-kit.
2301 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2303 This, or similar/improved, will also be change PP/11 of 4.82.
2309 PP/01 Handle short writes when writing local log-files.
2310 In practice, only affects FreeBSD (8 onwards).
2311 Bugzilla 1053, with thanks to Dmitry Isaikin.
2313 NM/01 Bugzilla 949 - Documentation tweak
2315 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2318 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2320 PP/02 Implemented gsasl authenticator.
2322 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2324 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2325 `pkg-config foo` for cflags/libs.
2327 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2328 with rest of GSASL and with heimdal_gssapi.
2330 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2331 `pkg-config foo` for cflags/libs for the TLS implementation.
2333 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2334 properties get this fed in as external SSF. A number of robustness
2335 and debugging improvements to the cyrus_sasl authenticator.
2337 PP/08 cyrus_sasl server now expands the server_realm option.
2339 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2340 Patch by Jeremy Harris.
2342 PP/10 Added dbmjz lookup type.
2344 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2346 PP/12 MAIL args handles TAB as well as SP, for better interop with
2347 non-compliant senders.
2348 Analysis and variant patch by Todd Lyons.
2350 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2351 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2352 Patch from Dirk Mueller <dmueller@suse.com>
2354 PP/13 tls_peerdn now print-escaped for spool files.
2355 Observed some $tls_peerdn in wild which contained \n, which resulted
2356 in spool file corruption.
2358 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2359 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2360 or write after TLS renegotiation, which otherwise led to messages
2363 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2364 as a tracking header (ie: a signed header comes before the signature).
2365 Patch from Wolfgang Breyha.
2367 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2368 comma-sep list; embedded commas doubled.
2370 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2372 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2374 Report and patch from Dmitry Banschikov.
2376 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2377 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2378 failures. We appear to now support TLS1.1+ with Exim.
2380 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2381 lets Exim select keys and certificates based upon TLS SNI from client.
2382 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2383 before an outbound SMTP session. New log_selector, +tls_sni.
2385 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2386 NULL dereference. Report and patch from Alun Jones.
2388 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2389 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2390 Not seeing resolver debug output on NetBSD, but suspect this is a
2391 resolver implementation change.
2393 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2394 Left warnings. Added "eximon gdb" invocation mode.
2396 PP/21 Defaulting "accept_8bitmime" to true, not false.
2398 PP/22 Added -bw for inetd wait mode support.
2400 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2401 locate the relevant includes and libraries. Made this the default.
2403 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2404 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2406 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2407 This may cause build issues on older platforms.
2409 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2410 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2411 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2412 Added SNI support via GnuTLS too.
2413 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2415 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2417 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2418 Applied second patch from Janne, implementing suggestion to default
2419 multiple-strings-in-record handling to match SPF spec.
2421 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2423 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2424 read-only, out of scope).
2425 Patch from Wolfgang Breyha, report from Stuart Northfield.
2427 PP/29 Fix three issues highlighted by clang analyser static analysis.
2428 Only crash-plausible issue would require the Cambridge-specific
2429 iplookup router and a misconfiguration.
2430 Report from Marcin Mirosław.
2432 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2434 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2435 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2436 As part of this, removing so much warning spew let me fix some minor
2437 real issues in debug logging.
2439 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2440 assignment on my part. Fixed.
2442 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2443 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2444 Janne Snabb (who went above and beyond: thank you).
2446 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2447 string otherwise requires a connection and a bunch more work and it's
2448 relatively easy to get wrong. Should also expose TLS library linkage
2451 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2452 64-bit ${eval} (JH/03).
2454 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2455 GNU libc to support some of the 64-bit stuff, should not lead to
2456 conflicts. Defined before os.h is pulled in, so if a given platform
2457 needs to override this, it can.
2459 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2460 protection layer was required, which is not implemented.
2461 Bugzilla 1254, patch from Wolfgang Breyha.
2463 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2464 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2465 tls_dhparam take prime identifiers. Also unbreak combination of
2466 OpenSSL+DH_params+TLSSNI.
2468 PP/39 Disable SSLv2 by default in OpenSSL support.
2474 PP/01 Solaris build fix for Oracle's LDAP libraries.
2475 Bugzilla 1109, patch from Stephen Usher.
2477 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2479 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2482 TF/02 Fix a couple more cases where we did not log the error message
2483 when unlink() failed. See also change 4.74-TF/03.
2485 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2486 lock up or crash if it happened to be inside a call to libc when it
2487 got a SIGUSR1 from exiwhat.
2489 The SIGUSR1 handler appends the current process status to the process
2490 log which is later printed by exiwhat. It used to use the general
2491 purpose logging code to do this, but several functions it calls are
2492 not safe for signals.
2494 The new output code in the SIGUSR1 handler is specific to the process
2495 log, and simple enough that it's easy to inspect for signal safety.
2496 Removing some special cases also simplifies the general logging code.
2497 Removing the spurious timestamps from the process log simplifies
2500 TF/04 Improved ratelimit ACL condition.
2502 The /noupdate option has been deprecated in favour of /readonly which
2503 has clearer semantics. The /leaky, /strict, and /readonly update modes
2504 are mutually exclusive. The update mode is no longer included in the
2505 database key; it just determines when the database is updated. (This
2506 means that when you upgrade Exim will forget old rate measurements.)
2508 Exim now checks that the per_* options are used with an update mode that
2509 makes sense for the current ACL. For example, when Exim is processing a
2510 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2511 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2512 must specify per_mail/readonly. If you omit the update mode it defaults to
2513 /leaky where that makes sense (as before) or /readonly where required.
2515 The /noupdate option is now undocumented but still supported for
2516 backwards compatibility. It is equivalent to /readonly except that in
2517 ACLs where /readonly is required you may specify /leaky/noupdate or
2518 /strict/noupdate which are treated the same as /readonly.
2520 A useful new feature is the /count= option. This is a generalization
2521 of the per_byte option, so that you can measure the throughput of other
2522 aggregate values. For example, the per_byte option is now equivalent
2523 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2525 The per_rcpt option has been generalized using the /count= mechanism
2526 (though it's more complicated than the per_byte equivalence). When it is
2527 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2528 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2529 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2530 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2531 Note that using per_rcpt with a non-readonly update mode in more than
2532 one ACL will cause the recipients to be double-counted. (The per_mail
2533 and per_byte options don't have this problem.)
2535 The handling of very low rates has changed slightly. If the computed rate
2536 is less than the event's count (usually one) then this event is the first
2537 after a long gap. In this case the rate is set to the same as this event's
2538 count, so that the first message of a spam run is counted properly.
2540 The major new feature is a mechanism for counting the rate of unique
2541 events. The new per_addr option counts the number of different
2542 recipients that someone has sent messages to in the last time period. It
2543 behaves like per_rcpt if all the recipient addresses are different, but
2544 duplicate recipient addresses do not increase the measured rate. Like
2545 the /count= option this is a general mechanism, so the per_addr option
2546 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2547 example, measure the rate that a client uses different sender addresses
2548 with the options per_mail/unique=$sender_address. There are further
2549 details in the main documentation.
2551 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2553 TF/06 Removed a few PCRE remnants.
2555 TF/07 Automatically extract Exim's version number from tags in the git
2556 repository when doing development or release builds.
2558 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2559 Bugzilla 879. Patch from Paul Fisher.
2561 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2562 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2565 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2567 PP/05 Variable $av_failed, true if the AV scanner deferred.
2568 Bugzilla 1078. Patch from John Horne.
2570 PP/06 Stop make process more reliably on build failure.
2571 Bugzilla 1087. Patch from Heiko Schlittermann.
2573 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2574 Bugzilla 1089. Patch from Heiko Schlittermann.
2576 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2577 Bugzilla 1131. Patch from Holger Weiß.
2579 PP/09 Handle IPv6 addresses with SPF.
2580 Bugzilla 860. Patch from Wolfgang Breyha.
2582 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2584 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2587 PP/11 match_* no longer expand right-hand-side by default.
2588 New compile-time build option, EXPAND_LISTMATCH_RHS.
2589 New expansion conditions, "inlist", "inlisti".
2591 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2593 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2595 PP/14 fix log_write() format string regression from TF/03.
2596 Bugzilla 1152. Patch from Dmitry Isaikin.
2602 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
2604 PP/02 Harmonised TLS library version reporting; only show if debugging.
2605 Layout now matches that introduced for other libraries in 4.74 PP/03.
2607 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
2609 PP/04 New "dns_use_edns0" global option.
2611 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
2614 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
2615 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
2617 TK/01 Updated PolarSSL code to 0.14.2.
2618 Bugzilla 1097. Patch from Andreas Metzler.
2620 PP/07 Catch divide-by-zero in ${eval:...}.
2621 Fixes bugzilla 1102.
2623 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
2626 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
2627 format-string attack -- SECURITY: remote arbitrary code execution.
2629 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
2630 time unintentionally subject to list matching rules, letting the header
2631 cause arbitrary Exim lookups (of items which can occur in lists, *not*
2632 arbitrary string expansion). This allowed for information disclosure.
2634 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
2635 INT_MIN/-1 -- value coerced to INT_MAX.
2641 NM/01 Workaround for PCRE version dependency in version reporting
2644 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
2645 This fixes portability to compilers other than gcc, notably
2646 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
2648 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
2649 makefiles for portability to HP-UX and POSIX correctness.
2651 PP/01 Permit LOOKUP_foo enabling on the make command-line.
2652 Also via indented variable definition in the Makefile.
2653 (Debugging by Oliver Heesakkers).
2655 PP/02 Restore caching of spamd results with expanded spamd_address.
2656 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
2658 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
2659 Improves build reliability. Fix from: Frank Elsner
2661 NM/02 Fix wide character breakage in the rfc2047 coding
2662 Fixes bug 1064. Patch from Andrey N. Oktyabrski
2664 NM/03 Allow underscore in dnslist lookups
2665 Fixes bug 1026. Patch from Graeme Fowler
2667 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
2668 Code patches from Adam Ciarcinski of NetBSD.
2670 NM/04 Fixed exiqgrep to cope with mailq missing size issue
2673 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
2674 is logged, to avoid truncation. Patch from John Horne.
2676 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
2677 Patch from Jakob Hirsch.
2679 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
2680 SQL string expansion failure details.
2681 Patch from Andrey Oktyabrski.
2683 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
2684 Patch from Simon Arlott.
2686 PP/09 New lookups functionality failed to compile on old gcc which rejects
2687 extern declarations in function scope.
2688 Patch from Oliver Fleischmann
2690 PP/10 Use sig_atomic_t for flags set from signal handlers.
2691 Check getgroups() return and improve debugging.
2692 Fixed developed for diagnosis in bug 927 (which turned out to be
2695 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
2696 Patch from Mark Zealey.
2698 PP/12 Bugzilla 1056: Improved spamd server selection.
2699 Patch from Mark Zealey.
2701 PP/13 Bugzilla 1086: Deal with maildir quota file races.
2702 Based on patch from Heiko Schlittermann.
2704 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
2705 Patch from Uwe Doering, sign-off by Michael Haardt.
2707 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
2708 variable declaration deep within a block. Bug and patch from
2711 PP/15 lookups-Makefile IRIX compatibility coercion.
2713 PP/16 Make DISABLE_DKIM build knob functional.
2715 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
2716 Patch by Simon Arlott
2718 TF/03 Fix valgrind.h portability to C89 compilers that do not support
2719 variable argument macros. Our copy now differs from upstream.
2725 TF/01 Failure to get a lock on a hints database can have serious
2726 consequences so log it to the panic log.
2728 TF/02 Log LMTP confirmation messages in the same way as SMTP,
2729 controlled using the smtp_confirmation log selector.
2731 TF/03 Include the error message when we fail to unlink a spool file.
2733 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2734 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2735 for maintaining out-of-tree patches for some time.
2737 PP/01 Bugzilla 139: Documentation and portability issues.
2738 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2739 Handle per-OS dynamic-module compilation flags.
2741 PP/02 Let /dev/null have normal permissions.
2742 The 4.73 fixes were a little too stringent and complained about the
2743 permissions on /dev/null. Exempt it from some checks.
2744 Reported by Andreas M. Kirchwitz.
2746 PP/03 Report version information for many libraries, including
2747 Exim version information for dynamically loaded libraries. Created
2748 version.h, now support a version extension string for distributors
2749 who patch heavily. Dynamic module ABI change.
2751 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2752 privilege escalation vulnerability whereby the Exim run-time user
2753 can cause root to append content of the attacker's choosing to
2756 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2759 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2760 If dropping privileges for untrusted macros, we disabled normal logging
2761 on the basis that it would fail; for the Exim run-time user, this is not
2762 the case, and it resulted in successful deliveries going unlogged.
2763 Fixed. Reported by Andreas Metzler.
2769 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2770 only prepend for the Resent-* case. Fixes regression introduced in
2771 Exim 4.70 by NM/22 for Bugzilla 607.
2773 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2774 increasing numbers of administrators be bitten by this.
2776 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2778 PP/03 Bugzilla 994: added openssl_options main configuration option.
2780 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2782 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2784 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2786 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2787 without a peer certificate, leading to a segfault because of an
2788 assumption that peers always have certificates. Be a little more
2789 paranoid. Problem reported by Martin Tscholak.
2791 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2792 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2793 NB: ClamAV planning to remove STREAM in "middle of 2010".
2794 CL also introduces -bmalware, various -d+acl logging additions and
2795 more caution in buffer sizes.
2797 PP/09 Implemented reverse_ip expansion operator.
2799 PP/10 Bugzilla 937: provide a "debug" ACL control.
2801 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2803 PP/12 Bugzilla 973: Implement --version.
2805 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2807 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2809 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2811 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2812 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2813 ignore trailing whitespace.
2815 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2817 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2820 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2821 Notification from Dr Andrew Aitchison.
2823 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2824 ExtendedDetectionInfo response format.
2825 Notification from John Horne.
2827 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2830 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2831 XSL and documented dependency on system catalogs, with examples of how
2834 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2837 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2838 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2839 configuration file which is writeable by the Exim user or group.
2841 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2842 of configuration files to cover files specified with the -C option if
2843 they are going to be used with root privileges, not just the default
2846 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2847 option (effectively making it always true).
2849 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2850 files to be used while preserving root privileges.
2852 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2853 that rogue child processes cannot use them.
2855 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2856 run-time user, instead of root.
2858 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2859 Exim run-time user without dropping privileges.
2861 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2862 result string, instead of calling string_vformat() twice with the same
2865 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2866 for other users. Others should always drop root privileges if they use
2867 -C on the command line, even for a whitelisted configure file.
2869 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2871 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2877 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2878 $data_path, and $header_path variables; fixed documentation bugs and
2881 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2882 exipick to access non-standard spools, including the "frozen" queue
2885 NM/01 Bugzilla 965: Support mysql stored procedures.
2886 Patch from Alain Williams
2888 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2890 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2891 Patch from Andreas Metzler
2893 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2894 Patch from Kirill Miazine
2896 NM/05 Bugzilla 671: Added umask to procmail example.
2898 JJ/03 installed exipick 20100323.0, fixing doc bug
2900 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2901 directory. Notification and patch from Dan Rosenberg.
2903 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2905 TK/02 Improve log output when DKIM signing operation fails.
2907 MH/01 Treat the transport option dkim_domain as a colon separated
2908 list, not as a single string, and sign the message with each element,
2909 omitting multiple occurences of the same signer.
2911 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2912 Bugzilla 985, 986. Patch by Simon Arlott
2914 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2915 Patch by Simon Arlott
2917 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2918 MBX locking. Notification from Dan Rosenberg.
2924 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2926 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2928 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2930 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2932 NM/04 Bugzilla 915: Flag broken perl installation during build.
2938 TK/01 Added patch by Johannes Berg that expands the main option
2939 "spamd_address" if it starts with a dollar sign.
2941 TK/02 Write list of recipients to X-Envelope-Sender header when building
2942 the mbox-format spool file for content scanning (suggested by Jakob
2945 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2946 (http://www.dcc-servers.net/) support via dccifd. Activated by
2947 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2949 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2950 by Mark Daniel Reidel <mr@df.eu>.
2952 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2953 When building exim an external PCRE library is now needed -
2954 PCRE is a system library on the majority of modern systems.
2955 See entry on PCRE_LIBS in EDITME file.
2957 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2958 conversation. Added nologin parameter to request.
2959 Patch contributed by Kirill Miazine.
2961 TF/01 Do not log submission mode rewrites if they do not change the address.
2963 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2965 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2966 log files in place. Contributed by Roberto Lima.
2968 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2970 TF/03 Bugzilla 615: When checking the local_parts router precondition
2971 after a local_part_suffix or local_part_prefix option, Exim now
2972 does not use the address's named list lookup cache, since this
2973 contains cached lookups for the whole local part.
2975 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2976 Robert Millan. Documentation is in experimental-spec.txt.
2978 TF/04 Bugzilla 668: Fix parallel build (make -j).
2980 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2982 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2983 Patch provided by Jan Srzednicki.
2985 TF/05 Leading white space used to be stripped from $spam_report which
2986 wrecked the formatting. Now it is preserved.
2988 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2989 that they are available at delivery time.
2991 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2993 TF/08 TLS error reporting now respects the incoming_interface and
2994 incoming_port log selectors.
2996 TF/09 Produce a more useful error message if an SMTP transport's hosts
2997 setting expands to an empty string.
2999 NM/06 Bugzilla 744: EXPN did not work under TLS.
3000 Patch provided by Phil Pennock.
3002 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3003 Patch provided by Richard Godbee.
3005 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3006 acl_smtp_notquit, added index entry.
3008 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3009 Patch provided by Eugene Bujak.
3011 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3012 accept(). Patch provided by Maxim Dounin.
3014 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3015 Patch provided by Phil Pennock.
3017 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3019 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3020 Patch provided by Brad "anomie" Jorsch.
3022 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3023 Patch provided by Dean Brooks.
3025 TK/05 Add native DKIM support (does not depend on external libraries).
3027 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3028 Patch provided by Graeme Fowler.
3030 NM/16 Bugzilla 851: Documentation example syntax fix.
3032 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3034 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3037 NM/19 Bugzilla 745: TLS version reporting.
3038 Patch provided by Phil Pennock.
3040 NM/20 Bugzilla 167: bool: condition support.
3041 Patch provided by Phil Pennock.
3043 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3044 clients. Patch provided by Phil Pennock.
3046 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3047 Patch provided by Brad "anomie" Jorsch.
3049 NM/23 Bugzilla 687: Fix misparses in eximstats.
3050 Patch provided by Heiko Schlittermann.
3052 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3053 Patch provided by Heiko Schlittermann.
3055 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3056 plus update to original patch.
3058 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3060 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3061 Patch provided by David Brownlee.
3063 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3065 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3067 NM/30 Bugzilla 888: TLS documentation bugfixes.
3069 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3071 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3072 Unlike the original bugzilla I have changed all shell scripts in src tree.
3074 NM/33 Bugzilla 898: Transport filter timeout fix.
3075 Patch by Todd Rinaldo.
3077 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3078 Patch by Serge Demonchaux.
3080 NM/35 Bugzilla 39: Base64 decode bug fixes.
3081 Patch by Jakob Hirsch.
3083 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3085 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3087 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3089 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3095 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3096 ALT-N's libdkim that I have put here:
3097 http://duncanthrax.net/exim-experimental/
3099 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3100 were called 'true' and it seems that C99 defines that as a reserved
3101 keyword to be used with 'bool' variable types. That means you could
3102 not include C99-style headers which use bools without triggering
3103 build errors in sieve.c.
3105 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3106 as mailq or other aliases. Changed the --help handling significantly
3107 to do whats expected. exim_usage() emits usage/help information.
3109 SC/01 Added the -bylocaldomain option to eximstats.
3111 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3113 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3115 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3121 PH/01 Another patch from the Sieve maintainer.
3123 PH/02 When an IPv6 address is converted to a string for single-key lookup
3124 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3125 dots are used instead of colons so that keys in lsearch files need not
3126 contain colons. This was done some time before quoting was made available
3127 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3128 (notated using the quote facility) so as to distinguish them from IPv4
3129 keys. This meant that lookups for IP addresses in host lists did not work
3130 for iplsearch lookups.
3132 This has been fixed by arranging for IPv6 addresses to be expressed with
3133 colons if the lookup type is iplsearch. This is not incompatible, because
3134 previously such lookups could never work.
3136 The situation is now rather anomalous, since one *can* have colons in
3137 ordinary lsearch keys. However, making the change in all cases is
3138 incompatible and would probably break a number of configurations.
3140 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3143 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3144 conversion specification without a maximum field width, thereby enabling
3145 a rogue spamd server to cause a buffer overflow. While nobody in their
3146 right mind would setup Exim to query an untrusted spamd server, an
3147 attacker that gains access to a server running spamd could potentially
3148 exploit this vulnerability to run arbitrary code as the Exim user.
3150 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3151 $primary_hostname instead of what libspf2 thinks the hosts name is.
3153 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3154 a directory entry by the name of the lookup key. Previously, if a
3155 symlink pointed to a non-existing file or a file in a directory that
3156 Exim lacked permissions to read, a lookup for a key matching that
3157 symlink would fail. Now it is enough that a matching directory entry
3158 exists, symlink or not. (Bugzilla 503.)
3160 PH/03 The body_linecount and body_zerocount variables are now exported in the
3163 PH/04 Added the $dnslist_matched variable.
3165 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3166 This means they are set thereafter only if the connection becomes
3169 PH/06 Added the client_condition to authenticators so that some can be skipped
3170 by clients under certain conditions.
3172 PH/07 The error message for a badly-placed control=no_multiline_responses left
3173 "_responses" off the end of the name.
3175 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3177 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3178 (without spaces) instead of just copying the configuration text.
3180 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3182 PH/11 Added $max_received_linelength.
3184 PH/12 Added +ignore_defer and +include_defer to host lists.
3186 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3187 way in which PCRE > 7.0 is built.
3189 PH/14 Implemented queue_only_load_latch.
3191 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3192 MAIL command. The effect was to mangle the value on 64-bit systems.
3194 PH/16 Another patch from the Sieve maintainer.
3196 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3198 PH/18 If a system quota error occurred while trying to create the file for
3199 a maildir delivery, the message "Mailbox is full" was not appended to the
3200 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3201 only to a quota excession during the actual writing of the file.
3203 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3204 characters?) which causes problems in log lines. The DN values are now
3205 passed through string_printing() before being added to log lines.
3207 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3208 and InterBase are left for another time.)
3210 PH/21 Added message_body_newlines option.
3212 PH/22 Guard against possible overflow in moan_check_errorcopy().
3214 PH/23 POSIX allows open() to be a macro; guard against that.
3216 PH/24 If the recipient of an error message contained an @ in the local part
3217 (suitably quoted, of course), incorrect values were put in $domain and
3218 $local_part during the evaluation of errors_copy.
3224 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3225 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3228 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3229 issue a MAIL command.
3231 PH/02 In an ACL statement such as
3233 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3235 if a client was not listed at all, or was listed with a value other than
3236 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3237 the condition was not true (as it should be), so access was not denied.
3238 The bug was that the ! inversion was incorrectly passed on to the second
3239 item. This has been fixed.
3241 PH/03 Added additional dnslists conditions == and =& which are different from
3242 = and & when the dns lookup returns more than one IP address.
3244 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3245 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3247 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3248 FSYNC, which compiles an option called disable_fsync that allows for
3249 bypassing fsync(). The documentation is heavily laced with warnings.
3251 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3253 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3254 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3255 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3256 including adding "make clean"; (3) Added -fPIC when compiling the test
3257 dynamically loaded module, to get rid of a warning.
3259 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3260 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3261 = 0s. The bug is otherwise harmless.
3263 PH/07 There was a bug in the dovecot authenticator such that the value of
3264 $auth1 could be overwritten, and so not correctly preserved, after a
3265 successful authentication. This usually meant that the value preserved by
3266 the server_setid option was incorrect.
3268 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3270 PH/09 Installed PCRE release 7.0.
3272 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3273 run for batched SMTP input. It is now run at the start of every message
3274 in the batch. While fixing this I discovered that the process information
3275 (output by running exiwhat) was not always getting set for -bs and -bS
3276 input. This is fixed, and it now also says "batched" for BSMTP.
3278 PH/11 Added control=no_pipelining.
3280 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3281 patch, slightly modified), and move the expansion of helo_data till after
3282 the connection is made in the smtp transport (so it can use these
3285 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3287 PH/14 Added log_selector = +pid.
3289 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3291 PH/16 Add ${if forany and ${if forall.
3293 PH/17 Added dsn_from option to vary the From: line in DSNs.
3295 PH/18 Flush SMTP output before performing a callout, unless control =
3296 no_callout_flush is set.
3298 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3299 was true (the default) a successful delivery failed to delete the retry
3300 item, thus causing premature timeout of the address. The bug is now
3303 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3305 PH/21 Long custom messages for fakedefer and fakereject are now split up
3306 into multiline responses in the same way that messages for "deny" and
3307 other ACL rejections are.
3309 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3310 with slight modification.
3312 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3313 draft, changing the syntax and factoring some duplicate code".
3315 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3316 for deliveries of the second and subsequent messages over the same SMTP
3319 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3320 ${reduce, with only minor "tidies".
3322 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3324 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3325 expansion side effects.
3327 PH/27 When a message times out after an over-quota error from an Exim-imposed
3328 quota, the bounce message says "mailbox is full". This message was not
3329 being given when it was a system quota that was exceeded. It now should
3332 MH/03 Made $recipients available in local_scan(). local_scan() already has
3333 better access to the recipient list through recipients_list[], but
3334 $recipients can be useful in postmaster-provided expansion strings.
3336 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3337 in the case of a MAIL command with additional options following the
3338 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3339 were accidentally chopped off.
3341 PH/29 SMTP synchronization checks are implemented when a command is read -
3342 there is a check that no more input is waiting when there shouldn't be
3343 any. However, for some commands, a delay in an ACL can mean that it is
3344 some time before the response is written. In this time, more input might
3345 arrive, invalidly. So now there are extra checks after an ACL has run for
3346 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3347 pipelining has not been advertised.
3349 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3351 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3352 returned in the SMTP response when smtp_return_error_details was false.
3353 This has been fixed.
3355 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3356 instead of the C I/O that was originally supplied, because problems were
3357 reported on Solaris.
3359 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3360 Exim which did not show up earlier: it was assuming that a call to
3361 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3362 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3363 was a macro that became an assignment, so it seemed to work. This has
3364 changed to a proper function call with a void return, hence the compile
3365 error. Exim's code has been fixed.
3367 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3370 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3372 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3374 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3375 because it turns out to be too big or there is a timeout) there is no
3376 "Completed" line in the log. When some messages of this type were
3377 selected by exigrep, they were listed as "not completed". Others were
3378 picked up by some special patterns. I have improved the selection
3379 criteria to be more general.
3381 PH/38 The host_find_failed option in the manualroute router can now be set
3382 to "ignore", to completely ignore a host whose IP address cannot be
3383 found. If all hosts are ignored, the behaviour is controlled by the new
3384 host_all_ignored option.
3386 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3387 homing or because of multiple MX records with /mx) generated more than
3388 one IP address, and the following item turned out to be the local host,
3389 all the secondary addresses of the first item were incorrectly removed
3390 from the list, along with the local host and any following hosts (which
3391 is what is supposed to happen).
3393 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3394 whoever called Exim into the -H file. In the case of the daemon it was
3395 behaving confusingly. When first started, it used values for whoever
3396 started the daemon, but after a SIGHUP it used the Exim user (because it
3397 calls itself on a restart). I have changed the code so that it now always
3400 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3401 message are rejected with the same error (e.g. no authentication or bad
3402 sender address), and a DATA command is nevertheless sent (as can happen
3403 with PIPELINING or a stupid MUA), the error message that was given to the
3404 RCPT commands is included in the rejection of the DATA command. This is
3405 intended to be helpful for MUAs that show only the final error to their
3408 PH/42 Another patch from the Sieve maintainer.
3410 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3411 Eximstats - Fixed some broken HTML links and added missing column headers
3413 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3414 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3416 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3418 PH/43 Yet another patch from the Sieve maintainer.
3420 PH/44 I found a way to check for a TCP/IP connection going away before sending
3421 the response to the final '.' that terminates a message, but only in the
3422 case where the client has not sent further data following the '.'
3423 (unfortunately, this is allowed). However, in many cases there won't be
3424 any further data because there won't be any more messages to send. A call
3425 to select() can be used: if it shows that the input is "ready", there is
3426 either input waiting, or the socket has been closed. An attempt to read
3427 the next input character can distinguish the two cases. Previously, Exim
3428 would have sent an OK response which the client would never have see.
3429 This could lead to message repetition. This fix should cure that, at
3430 least in a lot of common cases.
3432 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3433 advertised in response to EHLO.
3439 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3440 fixed by 4.65/MH/01 (is this a record?) are fixed:
3442 (i) An empty string was always treated as zero by the numeric comparison
3443 operators. This behaviour has been restored.
3445 (ii) It is documented that the numeric comparison operators always treat
3446 their arguments as decimal numbers. This was broken in that numbers
3447 starting with 0 were being interpreted as octal.
3449 While fixing these problems I realized that there was another issue that
3450 hadn't been noticed. Values of message_size_limit (both the global option
3451 and the transport option) were treated as octal if they started with 0.
3452 The documentation was vague. These values are now always treated as
3453 decimal, and I will make that clear in the documentation.
3459 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3460 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3463 MH/01 Don't check that the operands of numeric comparison operators are
3464 integers when their expansion is in "skipping" mode (fixes bug
3465 introduced by 4.64-PH/07).
3467 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3468 child addresses, Exim now panics and dies. Previously, because the count
3469 is held in a short int, deliveries were likely to be lost. As such a
3470 large number of recipients for a single message is ridiculous
3471 (performance will be very, very poor), I have chosen to impose a limit
3472 rather than extend the field.
3478 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3479 leftover -K file (the existence of which was triggered by #402).
3480 While we were at it, introduced process PID as part of the -K
3481 filename. This should rule out race conditions when creating
3484 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3485 processing considerably. Previous code took too long for large mails,
3486 triggering a timeout which in turn triggers #401.
3488 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3489 in the DK code in transports.c. sendfile() is not really portable,
3490 hence the _LINUX specificness.
3492 TF/01 In the add_headers option to the mail command in an Exim filter,
3493 there was a bug that Exim would claim a syntax error in any
3494 header after the first one which had an odd number of characters
3497 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3498 callout verification, Exim cached a "reject" for the entire domain. This
3499 is correct for most verifications, but it is not correct for a recipient
3500 verification with use_sender or use_postmaster set, because in that case
3501 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3502 case of MAIL FROM:<> rejection from other early rejections (e.g.
3503 rejection of HELO). When verifying a recipient using a non-null MAIL
3504 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3505 Whatever the result of the callout, the value of the domain cache is
3506 left unchanged (for any other kind of callout, getting as far as trying
3507 RCPT means that the domain itself is ok).
3509 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3512 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3513 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3514 session, the daemon ignores ECONNECT errors and logs others; it now
3515 ignores EPIPE as well.
3517 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3518 (quoted-printable decoding).
3520 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3521 later the small subsequent patch to fix an introduced bug.
3523 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3525 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3527 PH/08 An error is now given if message_size_limit is specified negative.
3529 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3530 to be given (somewhat) arbitrary names.
3532 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3535 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3536 miscellaneous code fixes
3538 PH/10 Added the log_reject_target ACL modifier to specify where to log
3541 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3542 hostname. This is wrong, because it relates to the incoming message (and
3543 probably the interface on which it is arriving) and not to the outgoing
3544 callout (which could be using a different interface). This has been
3545 changed to use the value of the helo_data option from the smtp transport
3546 instead - this is what is used when a message is actually being sent. If
3547 there is no remote transport (possible with a router that sets up host
3548 addresses), $smtp_active_hostname is used.
3550 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3551 tweaks were necessary in order to get it to work (see also 21 below):
3552 (a) The code assumed that strncpy() returns a negative number on buffer
3553 overflow, which isn't the case. Replaced with Exim's string_format()
3555 (b) There were several signed/unsigned issues. I just did the minimum
3556 hacking in of casts. There is scope for a larger refactoring.
3557 (c) The code used strcasecmp() which is not a standard C function.
3558 Replaced with Exim's strcmpic() function.
3559 (d) The code set only $1; it now sets $auth1 as well.
3560 (e) A simple test gave the error "authentication client didn't specify
3561 service in request". It would seem that Dovecot has changed its
3562 interface. Fortunately there's a specification; I followed it and
3563 changed what the client sends and it appears to be working now.
3565 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3568 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3569 address A is aliased to B and C, where B exists and C does not. Without
3570 -v the output is "A verified" because verification stops after a
3571 successful redirection if more than one address is generated. However,
3572 with -v the child addresses are also verified. Exim was outputting "A
3573 failed to verify" and then showing the successful verification for C,
3574 with its parentage. It now outputs "B failed to verify", showing B's
3575 parentage before showing the successful verification of C.
3577 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3578 look up a TXT record in a specific list after matching in a combined
3581 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3582 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3583 they consult the DNS. I had assumed they would set it the way they
3584 wanted; and indeed my experiments on Linux seem to show that in some
3585 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3586 To be on the safe side, however, I have now made the interface to
3587 host_find_byname() similar to host_find_bydns(), with an argument
3588 containing the DNS resolver options. The host_find_byname() function now
3589 sets these options at its start, just as host_find_bydns() does. The smtp
3590 transport options dns_qualify_single and dns_search_parents are passed to
3591 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3592 of host_find_byname() use the default settings of RES_DEFNAMES
3593 (qualify_single) but not RES_DNSRCH (search_parents).
3595 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
3596 spool_read_header() do less string testing, by means of a preliminary
3597 switch on the second character of optional "-foo" lines. (This is
3598 overdue, caused by the large number of possibilities that now exist.
3599 Originally there were few.) While I was there, I also converted the
3600 str(n)cmp tests so they don't re-test the leading "-" and the first
3601 character, in the hope this might squeeze out yet more improvement.
3603 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
3604 flag allowing group syntax was set by the header_syntax check but not
3605 turned off, possible causing trouble later; (2) The flag was not being
3606 set at all for the header_verify test, causing "group"-style headers to
3607 be rejected. I have now set it in this case, and also caused header_
3608 verify to ignore an empty address taken from a group. While doing this, I
3609 came across some other cases where the code for allowing group syntax
3610 while scanning a header line wasn't quite right (mostly, not resetting
3611 the flag correctly in the right place). These bugs could have caused
3612 trouble for malformed header lines. I hope it is now all correct.
3614 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
3615 with the "reply" argument non-NULL. The code, however (which originally
3616 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
3617 but it didn't always do it. This confused somebody who was copying the
3618 code for some other use. I have removed all the tests.
3620 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
3621 feature that was used to support insecure browsers during the U.S. crypto
3622 embargo. It requires special client support, and Exim is probably the
3623 only MTA that supported it -- and would never use it because real RSA is
3624 always available. This code has been removed, because it had the bad
3625 effect of slowing Exim down by computing (never used) parameters for the
3626 RSA_EXPORT functionality.
3628 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
3629 authenticator to fail if there's a tab character in the incoming data
3630 (there should never be unless someone is messing about, as it's supposed
3631 to be base64-encoded). Also added, on Timo's advice, the "secured" option
3632 if the connection is using TLS or if the remote IP is the same as the
3633 local IP, and the "valid-client-cert option" if a client certificate has
3636 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
3637 authenticators. This can be used for authorization after authentication
3638 succeeds. (In the case of plaintext, it servers for both authentication
3641 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
3642 if any retry times were supplied.
3644 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
3645 connection, where there is no client IP address to check. In this
3646 situation, the verify now always succeeds.
3648 PH/25 Applied John Jetmore's -Mset patch.
3650 PH/26 Added -bem to be like -Mset, but loading a message from a file.
3652 PH/27 In a string expansion for a processed (not raw) header when multiple
3653 headers of the same name were present, leading whitespace was being
3654 removed from all of them, but trailing whitespace was being removed only
3655 from the last one. Now trailing whitespace is removed from each header
3656 before concatenation. Completely empty headers in a concatenation (as
3657 before) are ignored.
3659 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
3660 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
3662 PH/29 [Removed. This was a change that I later backed out, and forgot to
3663 correct the ChangeLog entry (that I had efficiently created) before
3664 committing the later change.]
3666 PH/30 Exim was sometimes attempting to deliver messages that had suffered
3667 address errors (4xx response to RCPT) over the same connection as other
3668 messages routed to the same hosts. Such deliveries are always "forced",
3669 so retry times are not inspected. This resulted in far too many retries
3670 for the affected addresses. The effect occurred only when there were more
3671 hosts than the hosts_max_try setting in the smtp transport when it had
3672 the 4xx errors. Those hosts that it had tried were not added to the list
3673 of hosts for which the message was waiting, so if all were tried, there
3674 was no problem. Two fixes have been applied:
3676 (i) If there are any address or message errors in an SMTP delivery, none
3677 of the hosts (tried or untried) are now added to the list of hosts
3678 for which the message is waiting, so the message should not be a
3679 candidate for sending over the same connection that was used for a
3680 successful delivery of some other message. This seems entirely
3681 reasonable: after all the message is NOT "waiting for some host".
3682 This is so "obvious" that I'm not sure why it wasn't done
3683 previously. Hope I haven't missed anything, but it can't do any
3684 harm, as the worst effect is to miss an optimization.
3686 (ii) If, despite (i), such a delivery is accidentally attempted, the
3687 routing retry time is respected, so at least it doesn't keep
3688 hammering the server.
3690 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
3691 in ${readsocket because some servers need this prod.
3693 PH/32 Added some extra debug output when updating a wait-xxx database.
3695 PH/33 The hint "could be header name not terminated by colon", which has been
3696 given for certain expansion errors for a long time, was not being given
3697 for the ${if def:h_colon_omitted{... case.
3699 PH/34 The spec says: "With one important exception, whenever a domain list is
3700 being scanned, $domain contains the subject domain." There was at least
3701 one case where this was not true.
3703 PH/35 The error "getsockname() failed: connection reset by peer" was being
3704 written to the panic log as well as the main log, but it isn't really
3705 panic-worthy as it just means the connection died rather early on. I have
3706 removed the panic log writing for the ECONNRESET error when getsockname()
3709 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
3710 runs only) independently of the message's sender address. This meant
3711 that, if the 4xx error was in fact related to the sender, a different
3712 message to the same recipient with a different sender could confuse
3713 things. In particular, this can happen when sending to a greylisting
3714 server, but other circumstances could also provoke similar problems.
3715 I have changed the default so that the retry time for these errors is now
3716 based a combination of the sender and recipient addresses. This change
3717 can be overridden by setting address_retry_include_sender=false in the
3720 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
3721 remote server are returned as part of bounce messages. This was not
3722 happening for LMTP over a pipe (the lmtp transport), but now it is the
3723 same for both kinds of LMTP.
3725 PH/38 Despite being documented as not happening, Exim was rewriting addresses
3726 in header lines that were in fact CNAMEs. This is no longer the case.
3728 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
3729 and queue runs started by the daemon processed all messages. This has
3730 been fixed so that -R and -S can now usefully be given with -q<time>.
3732 PH/40 Import PCRE release 6.7 (fixes some bugs).
3734 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3736 PH/42 Give an error if -q is specified more than once.
3738 PH/43 Renamed the variables $interface_address and $interface_port as
3739 $received_ip_address and $received_port, to make it clear that these
3740 values apply to message reception, and not to the outgoing interface when
3741 a message is delivered. (The old names remain recognized, of course.)
3743 PH/44 There was no timeout on the connect() call when using a Unix domain
3744 socket in the ${readsocket expansion. There now is.
3746 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3747 be meaningful with "accept".
3749 SC/01 Eximstats V1.43
3750 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3752 SC/02 Eximstats V1.44
3753 Use a glob alias rather than an array ref in the generated
3754 parser. This improves both readability and performance.
3756 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3757 Collect SpamAssassin and rejection statistics.
3758 Don't display local sender or destination tables unless
3759 there is data to show.
3760 Added average volumes into the top table text output.
3762 SC/04 Eximstats V1.46
3763 Collect data on the number of addresses (recipients)
3764 as well as the number of messages.
3766 SC/05 Eximstats V1.47
3767 Added 'Message too big' to the list of mail rejection
3768 reasons (thanks to Marco Gaiarin).
3770 SC/06 Eximstats V1.48
3771 Mainlog lines which have GMT offsets and are too short to
3772 have a flag are now skipped.
3774 SC/07 Eximstats V1.49 (Alain Williams)
3775 Added the -emptyok flag.
3777 SC/08 Eximstats V1.50
3778 Fixes for obtaining the IP address from reject messages.
3780 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3781 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3782 whitespace changes from 4.64-PH/27
3784 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3787 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3788 are found, allow negative numbers in numeric criteria)
3790 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3792 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3793 to match changes made in 4.64-PH/43
3795 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3797 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3798 "domain" is actually an IP address, and force a failure. This locks out
3799 those revolvers/nameservers that support "A-for-A" lookups, in
3800 contravention of the specifications.
3802 PH/48 When a host name was looked up from an IP address, and the subsequent
3803 forward lookup of the name timed out, the host name was left in
3804 $sender_host_name, contrary to the specification.
3806 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3807 restricted to single-key lookups, Exim was not diagnosing an error if
3808 * or *@ was used with a query-style lookup.
3810 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3812 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3813 long ago, but noone interested enough thought of it. Let's just say that
3814 the "1.1" means that there are some new functions that weren't there at
3815 some point in the past.
3817 PH/51 Error processing for expansion failure of helo_data from an smtp
3818 transport during callout processing was broken.
3820 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3821 tested/used via the -bh/-bhc/-bs options.
3823 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3824 bug, fixed in subsequent PCRE releases).
3826 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3827 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3829 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3835 SC/01 Use a glob alias rather than an array ref in eximstats generated
3836 parser. This improves both readability and performance.
3838 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3839 Don't display local sender or destination tables in eximstats unless
3840 there is data to show.
3841 Added average volumes into the eximstats top table text output.
3843 SC/03 Collect data on the number of addresses (recipients) as well
3844 as the number of messages in eximstats.
3846 TF/01 Correct an error in the documentation for the redirect router. Exim
3847 does (usually) call initgroups() when daemonizing.
3849 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3850 with consistent privilege compared to when running as a daemon.
3852 TF/03 Note in the spec that $authenticated_id is not set for local
3853 submissions from trusted users.
3855 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3856 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3858 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3859 by adding some example configuration directives to the default
3860 configuration file. A little bit of work is required to uncomment the
3861 directives and define how usernames and passwords are checked, but
3862 there is now a framework to start from.
3864 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3865 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3866 without this. I don't know how relevant this is to other LDAP libraries.
3868 PH/02 Add the verb name to the "unknown ACL verb" error.
3870 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3872 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3874 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3875 directory not expanded when it should be if an expanded home directory
3876 was set for the address (which is overridden by the transport).
3878 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3881 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3882 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3883 because it is too late at that time, and has no effect.
3885 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3886 security issue with \' (bugzilla #107). I could not use the
3887 PQescapeStringConn() function, because it needs a PGconn value as one of
3890 PH/08 When testing addresses using -bt, indicate those final addresses that
3891 are duplicates that would not cause an additional delivery. At least one
3892 person was confused, thinking that -bt output corresponded to deliveries.
3893 (Suppressing duplicates isn't a good idea as you lose the information
3894 about possibly different redirections that led to the duplicates.)
3896 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3897 systems where poll() doesn't work, in particular OS X.
3899 PH/10 Added more information to debugging output for retry time not reached.
3901 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3902 operations in malware.c.
3904 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3907 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3908 syslog_duplication set false, log lines that would normally be written
3909 both the the main log and to the reject log were not written to syslog at
3912 PH/14 In the default configuration, change the use of "message" in ACL warn
3913 statements to "add_header".
3915 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3916 not followed by a command (e.g. "seen endif").
3918 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3919 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3922 PH/17 Added extra conditions to the default value of delay_warning_condition
3926 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3927 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3928 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3931 The Auto-Submitted: and various List- headers are standardised, whereas I
3932 don't think Precedence: ever was.
3934 PH/18 Refactored debugging code in route_finduser() to show more information,
3935 in particular, the error code if getpwnam() issues one.
3937 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3938 This is apparently needed in addition to the PH/07 change above to avoid
3939 any possible encoding problems.
3941 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3942 but not after initializing Perl.
3944 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3945 output them only if debugging. By default they are written stderr,
3946 apparently, which is not desirable.
3948 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3951 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3954 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3956 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3957 authenticated or an ident call has been made. Suppress the default
3958 values for $authenticated_id and $authenticated_sender (but permit -oMai
3959 and -oMas) when testing with -bh.
3961 PH/24 Re-jigged the order of the tests in the default configuration so that the
3962 tests for valid domains and recipients precede the DNS black list and CSA
3963 tests, on the grounds that those ones are more expensive.
3965 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3966 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3967 This bug exists in every version of Exim that I still have, right back to
3970 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3971 However, an attempt to turn on case-sensitivity in a regex key by
3972 including (?-i) didn't work because the subject string was already
3973 lowercased, and the effects were non-intuitive. It turns out that a
3974 one-line patch can be used to allow (?-i) to work as expected.
3980 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3981 other effects) broke the use of negated acl sub-conditions.
3983 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3986 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3987 "Deny" causes Exim to reject the incoming connection with a 554 error.
3988 Unfortunately, if there is a major crisis, such as a disk failure,
3989 tcp-wrappers gives "deny", whereas what one would like would be some
3990 kind of temporary error. A kludge has been added to help with this.
3991 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3992 554 error is used if errno is still zero or contains ENOENT (which occurs
3993 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3996 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3998 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3999 errors. Otherwise a message that provokes a temporary error (when other
4000 messages do not) can cause a whole host to time out.
4002 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4003 addresses were routed directly to files or pipes from a redirect router.
4004 File deliveries just didn't batch; pipe deliveries might have suffered
4007 PH/06 A failure to get a lock for a hints database would erroneously always say
4008 "Failed to get write lock", even when it was really a read lock.
4010 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4011 of 0600. This has been changed to use the value of the lockfile_mode
4012 option (which defaults to 0600).
4014 PH/08 Applied small patch from the Sieve maintainer.
4016 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4017 folder from quota calculations, a direct delivery into this folder messed
4018 up the contents of the maildirsize file. This was because the regex was
4019 used only to exclude .Trash (or whatever) when the size of the mailbox
4020 was calculated. There was no check that a delivery was happening into an
4021 excluded directory. This bug has been fixed by ignoring all quota
4022 processing for deliveries into excluded directories.
4024 PH/10 Added the maildirfolder_create_regex option to appendfile.
4030 PH/01 The code for finding all the local interface addresses on a FreeBSD
4031 system running IPv6 was broken. This may well have applied to all BSD
4032 systems, as well as to others that have similar system calls. The broken
4033 code found IPv4 interfaces correctly, but gave incorrect values for the
4034 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4035 that it would not match correctly against @[] and not recognize the IPv6
4038 PH/02 The ipliteral router was not recognizing addresses of the form user@
4039 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4041 PH/03 Added disable_ipv6.
4043 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4044 decoded form, because it is most often used to construct To: headers
4045 lines in autoreplies, and the decoded form may well be syntactically
4046 invalid. However, $reply_address has leading white space removed, and all
4047 newlines turned into spaces so that the autoreply transport does not
4050 PH/05 If group was specified without a user on a router, and no group or user
4051 was specified on a transport, the group from the router was ignored.
4053 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4054 for visible compile-time settings that can be used to change these
4055 numbers, for those that want even more. Backwards compatibility with old
4056 spool files has been maintained. However, going back to a previous Exim
4057 release will lost any variables that are in spool files.
4059 PH/07 Two small changes when running in the test harness: increase delay when
4060 passing a TCP/IP connection to a new process, in case the original
4061 process has to generate a bounce, and remove special handling of
4062 127.0.0.2 (sic), which is no longer necessary.
4064 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4065 be the same on different OS.
4067 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4070 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4071 whether --show-vars was specified or not
4073 JJ/02 exipick: Added support for new ACL variable spool format introduced
4076 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4077 syntactically invalid From: or Reply-to: line, and a filter used this to
4078 generate an autoreply, and therefore failed to obtain an address for the
4079 autoreply, Exim could try to deliver to a non-existent relative file
4080 name, causing unrelated and misleading errors. What now happens is that
4081 it logs this as a hard delivery error, but does not attempt to create a
4084 PH/11 The exinext utility has a -C option for testing purposes, but although
4085 the given file was scanned by exinext itself; it wasn't being passed on
4086 when Exim was called.
4088 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4089 an end-of-file indication when reading a command response.
4091 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4092 compiled. In many other places in Exim, IPv6 addresses are always
4093 recognized, so I have changed this. It also means that IPv4 domain
4094 literals of the form [IPV4:n.n.n.n] are now always recognized.
4096 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4097 used if the router is not running as root, for example, when verifying at
4098 ACL time, or when using -bh. The debugging output from this situation was
4099 non-existent - all you got was a failure to exec. I have made two
4102 (a) Failures to set uid/gid, the current directory, or a process leader
4103 in a subprocess such as that created by queryprogram now generate
4104 suitable debugging output when -d is set.
4106 (b) The queryprogram router detects when it is not running as root,
4107 outputs suitable debugging information if -d is set, and then runs
4108 the subprocess without attempting to change uid/gid.
4110 PH/15 Minor change to Makefile for building test_host (undocumented testing
4113 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4114 additional section of a DNS packet that returns MX or SRV records.
4115 Instead, it always explicitly searches for A/AAAA records. This avoids
4116 major problems that occur when a DNS server includes only records of one
4117 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4118 fixed another bug: if SRV records were looked up and the corresponding
4119 address records were *not* found in the additional section, the port
4120 values from the SRV records were lost.
4122 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4123 using the correct key (the original address) when searching the retry
4124 rules in order to find which one to use for generating the retry hint.
4126 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4127 adding the default one. Similarly, if it contains a Reply-To: header, the
4128 errors_reply_to option, if set, is not used.
4130 PH/19 When calculating a retry time, Exim used to measure the "time since
4131 failure" by looking at the "first failed" field in the retry record. Now
4132 it does not use this if it is later than than the arrival time of the
4133 message. Instead it uses the arrival time. This makes for better
4134 behaviour in cases where some deliveries succeed, thus re-setting the
4135 "first failed" field. An example is a quota failure for a huge message
4136 when small messages continue to be delivered. Without this change, the
4137 "time since failure" will always be short, possible causing more frequent
4138 delivery attempts for the huge message than are intended.
4139 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4141 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4142 $1, $2, $3) because the numerical variables can be reset during some
4143 expansion items (e.g. "match"), thereby losing the authentication data.
4145 PH/21 Make -bV show the size of off_t variables so that the test suite can
4146 decide whether to run tests for quotas > 2G.
4148 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4149 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4150 filecount value is greater than 2G or if a quota value is greater than 2G
4151 on a system where the size of off_t is not greater than 4, a panic error
4154 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4155 never match. The debug and -bh output now contains an explicit error
4156 message indicating a malformed IPv4 address or mask.
4158 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4159 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4160 PH/23 above applies.
4162 PH/25 Do not write to syslog when running in the test harness. The only
4163 occasion when this arises is a failure to open the main or panic logs
4164 (for which there is an explicit test).
4166 PH/26 Added the /no_tell option to "control=freeze".
4168 PH/27 If a host name lookup failed very early in a connection, for example, if
4169 the IP address matched host_lookup and the reverse lookup yielded a name
4170 that did not have a forward lookup, an error message of the form "no IP
4171 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4172 could be logged. Now it outputs the IP address instead of "NULL".
4174 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4175 allows the sender and the authenticated sender to be set when
4176 submitting a message from within Exim. Since child_open_exim() is
4177 documented for local_scan(), the new function should be too.
4179 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4180 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4181 results in an empty string is now treated as unset.
4183 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4185 PH/31 Added sender_verify_fail logging option.
4187 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4188 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4189 tidied the source and removed it altogether.
4191 PH/33 When a queue run was abandoned because the load average was too high, a
4192 log line was always written; now it is written only if the queue_run log
4193 selector is set. In addition, the log line for abandonment now contains
4194 information about the queue run such as the pid. This is always present
4195 in "start" and "stop" lines but was omitted from the "abandon" line.
4197 PH/34 Omit spaces between a header name and the colon in the error message that
4198 is given when verify = headers_syntax fails (if there are lots of them,
4199 the message gets confusing).
4201 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4202 names, as there are now some PTR records that contain slashes. This check
4203 is only to protect against broken name servers that fall over on strange
4204 characters, so the fact that it applies to all lookups doesn't matter.
4206 PH/36 Now that the new test suite is complete, we can remove some of the
4207 special code in Exim that was needed for the old test suite. For example,
4208 sorting DNS records because real resolvers return them in an arbitrary
4209 order. The new test suite's fake resolver always returns records in the
4212 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4213 bounces) except when queue_only is set, to avoid logging races between
4214 the different processes.
4216 PH/38 Panic-die if .include specifies a non-absolute path.
4218 PH/39 A tweak to the "H" retry rule from its user.
4220 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4221 a label. They prevented compilation on older perls.
4223 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4224 a warning to be raised on newish perls.
4226 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4227 on queue. Changes to match documented behaviour of showing count of
4228 messages matching specified criteria.
4230 PH/40 Changed the default ident timeout from 30s to 5s.
4232 PH/41 Added support for the use of login_cap features, on those BSD systems
4233 that have them, for controlling the resources used by pipe deliveries.
4235 PH/42 The content-scanning code uses fopen() to create files in which to put
4236 message data. Previously it was not paying any attention to the mode of
4237 the files. Exim runs with umask(0) because the rest of the code creates
4238 files with open(), and sets the required mode explicitly. Thus, these
4239 files were ending up world-writeable. This was not a big issue, because,
4240 being within the spool directory, they were not world-accessible. I have
4241 created a function called modefopen, which takes an additional mode
4242 argument. It sets umask(777), creates the file, chmods it to the required
4243 mode, then resets the umask. All the relevant calls to fopen() in the
4244 content scanning code have been changed to use this function.
4246 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4247 to 24 hours. This avoids potential overflow problems when processing G
4248 and H retry rules. I suspect nobody ever tinkers with this value.
4250 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4252 PH/45 When the plaintext authenticator is running as a client, the server's
4253 challenges are checked to ensure they are valid base64 strings. By
4254 default, the authentication attempt is cancelled if an invalid string is
4255 received. Setting client_ignore_invalid_base64 true ignores these errors.
4256 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4257 they are received. Thus, the responses can be made to depend on the
4258 challenges. If an invalid string is ignored, an empty string is placed in
4261 PH/46 Messages that are created by the autoreply transport now contains a
4262 References: header, in accordance with RFCs 2822 and 3834.
4264 PH/47 Added authenticated_sender_force to the smtp transport.
4266 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4268 PH/49 Installed latest patch from the Sieve maintainer.
4270 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4271 was also set, the appendfile transport was unnecessarily scanning a
4272 directory of message files (e.g. for maildir delivery) to find the count
4273 of files (along with the size), even though it did not need this
4274 information. It now does the scan only if it needs to find either the
4275 size of the count of files.
4277 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4279 PH/52 Two bugs concerned with error handling when the smtp transport is
4282 (i) Exim was not creating retry information for temporary errors given
4283 for individual recipients after the DATA command when the smtp transport
4284 was used in LMTP mode. This meant that they could be retried too
4285 frequently, and not timed out correctly.
4287 (ii) Exim was setting the flag that allows error details to be returned
4288 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4289 recipients that were returned after the DATA command.
4291 PH/53 This is related to PH/52, but is more general: for any failing address,
4292 when detailed error information was permitted to be returned to the
4293 sender, but the error was temporary, then after the final timeout, only
4294 "retry timeout exceeded" was returned. Now it returns the full error as
4295 well as "retry timeout exceeded".
4297 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4298 do this, and (what is worse) MTAs that accept it.
4300 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4301 will now be deprecated.
4303 PH/56 New os.c-cygwin from the Cygwin maintainer.
4305 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4306 formats (previously only available in exim formats via -bpr, -bpru,
4307 and -bpra. Now also available in native and exiqgrep formats)
4309 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4310 with very large, slow to parse queues
4312 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4314 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4316 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4317 responses to authentication challenges, though it was showing the
4318 challenges; (ii) I've removed the CR characters from the debug output for
4321 PH/58 Allow for the insertion of a newline as well as a space when a string
4322 is turned into more than one encoded-word during RFC 2047 encoding. The
4323 Sieve code now uses this.
4325 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4326 data_4xx, lost_connection, tls_required.
4328 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4329 message was being sent as an SMTP response.
4331 PH/61 Add -l and -k options to exicyclog.
4333 PH/62 When verifying, if an address was redirected to one new address, so that
4334 verification continued, and the new address failed or deferred after
4335 having set something in $address_data, the value of $address_data was not
4336 passed back to the ACL. This was different to the case when no
4337 redirection occurred. The value is now passed back in both cases.
4339 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4340 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4341 use login_cap.h, so on its own it isn't the distinguishing feature. The
4342 new name refers directly to the setclassresources() function.
4344 PH/65 Added configuration files for NetBSD3.
4346 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4348 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4349 is preferred over IPv4.
4351 PH/68 The bounce_return_message and bounce_return_body options were not being
4352 honoured for bounces generated during the reception of non-SMTP messages.
4353 In particular, this applied to messages rejected by the ACL. This bug has
4354 been fixed. However, if bounce_return_message is true and bounce_return_
4355 body is false, the headers that are returned for a non-SMTP message
4356 include only those that have been read before the error was detected.
4357 (In the case of an ACL rejection, they have all been read.)
4359 PH/69 The HTML version of the specification is now built in a directory called
4360 spec_html instead of spec.html, because the latter looks like a path with
4361 a MIME-type, and this confuses some software.
4363 PH/70 Catch two compiler warnings in sieve.c.
4365 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4366 function verify_get_ident() calls ip_connect() to connect a socket, but
4367 if the "connect()" function timed out, ip_connect() used to close the
4368 socket. However, verify_get_ident() also closes the socket later, and in
4369 between Exim writes to the log, which may get opened at this point. When
4370 the socket was closed in ip_connect(), the log could get the same file
4371 descriptor number as the socket. This naturally causes chaos. The fix is
4372 not to close the socket in ip_connect(); the socket should be closed by
4373 the function that creates it. There was only one place in the code where
4374 this was missing, in the iplookup router, which I don't think anybody now
4375 uses, but I've fixed it anyway.
4377 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4378 well as to direct DNS lookups. Otherwise the handling of names in host
4379 lists is inconsistent and therefore confusing.
4385 PH/01 Two changes to the default runtime configuration:
4387 (1) Move the checks for relay_from_hosts and authenticated clients from
4388 after to before the (commented out) DNS black list checks.
4390 (2) Add control=submission to the relay_from_hosts and authenticated
4391 clients checks, on the grounds that messages accepted by these
4392 statements are most likely to be submissions.
4394 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4396 (1) Generate an error if the third argument for the ${prvs expansion is
4399 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4402 (3) Reset the variables that are obtained from the first argument of
4403 ${prvscheck and used in the second argument before leaving the code,
4404 because their memory is reclaimed, so using them afterwards may do
4407 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4408 one (it's much easier than Tom thought :-).
4410 (5) Because of (4), we can now allow for the use of $prvscheck_result
4411 inside the third argument.
4413 PH/03 For some reason, the default setting of PATH when running a command from
4414 a pipe transport was just "/usr/bin". I have changed it to
4417 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4418 anything to be listed in the output from -bV.
4420 PH/05 When a filter generated an autoreply, the entire To: header line was
4421 quoted in the delivery log line, like this:
4423 => >A.N.Other <ano@some.domain> <original@ddress> ...
4425 This has been changed so that it extracts the operative address. There
4426 may be more than one such address. If so, they are comma-separated, like
4429 => >ano@some.domain,ona@other.domain <original@ddress> ...
4431 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4432 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4433 not being looked up in the rDNS to get a host name, Exim was showing the
4434 IP address twice in Received: lines, even though the IP addresses were
4435 identical. For example:
4437 Received: from [1.2.3.4] (helo=[1.2.3.4])
4439 However, if the real host name was known, it was omitting the HELO data
4440 if it matched the actual IP address. This has been tidied up so that it
4441 doesn't show the same IP address twice.
4443 PH/07 When both +timestamp and +memory debugging was on, the value given by
4444 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4445 called by the debug printing, thereby overwriting the timestamp buffer.
4446 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4448 PH/08 When the original message was included in an autoreply transport, it
4449 always said "this is a copy of the message, including all the headers",
4450 even if body_only or headers_only was set. It now gives an appropriate
4453 PH/09 Applied a patch from the Sieve maintainer which:
4455 o fixes some comments
4456 o adds the (disabled) notify extension core
4457 o adds some debug output for the result of if/elsif tests
4458 o points to the current vacation draft in the documentation
4459 and documents the missing references header update
4463 o fixes a bug in processing the envelope test (when testing
4464 multiple envelope elements, the last element determined the
4467 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4468 Electronic Mail") by including:
4470 Auto-submitted: auto-generated
4472 in the messages that it generates (bounce messages and others, such as
4473 warnings). In the case of bounce messages for non-SMTP messages, there was
4474 also a typo: it was using "Auto_submitted" (underscore instead of
4475 hyphen). Since every message generated by Exim is necessarily in response
4476 to another message, thes have all been changed to:
4478 Auto-Submitted: auto-replied
4480 in accordance with these statements in the RFC:
4482 The auto-replied keyword:
4484 - SHOULD be used on messages sent in direct response to another
4485 message by an automatic process,
4487 - MUST NOT be used on manually-generated messages,
4489 - MAY be used on Delivery Status Notifications (DSNs) and Message
4490 Disposition Notifications (MDNs),
4492 - MUST NOT be used on messages generated by automatic or periodic
4493 processes, except for messages which are automatic responses to
4496 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4497 to the default Received: header definition.
4499 PH/12 Added log selector acl_warn_skipped (default on).
4501 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4502 variables because (a) they are in the wrong storage pool and (b) even if
4503 they were copied, it wouldn't work properly because of the caching.
4505 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4506 checking when decoding. Apparently there are clients that generate
4507 overlong encoded strings. Why am I not surprised?
4509 PH/15 If the first argument of "${if match_address" was not empty, but did not
4510 contain an "@" character, Exim crashed. Now it writes a panic log message
4511 and treats the condition as false.
4513 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4515 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4516 extension "envelope-auth". The code is finished and in agreement with
4517 other implementations, but there is no documentation so far and in fact,
4518 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4519 not changing the active code.
4521 Print executed "if" and "elsif" statements when debugging is used. This
4522 helps a great deal to understand what a filter does.
4524 Document more things not specified clearly in RFC3028. I had all this
4525 sorted out, when out of a sudden new issues came to my mind. Oops."
4527 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4530 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4531 canonical form (as documented). However, after a host name lookup from
4532 the IP address, check_host() was doing a simple string comparison with
4533 addresses acquired from the DNS when checking that the found name did
4534 have the original IP as one of its addresses. Since any found IPv6
4535 addresses are likely to be in abbreviated form, the comparison could
4536 fail. Luckily, there already exists a function for doing the comparison
4537 by converting both addresses to binary, so now that is used instead of
4538 the text comparison.
4540 PH/20 There was another similar case to PH/19, when a complete host name was
4541 given in a host list; looking up its IP address could give an abbreviated
4542 form, whereas the current host's name might or might not be abbreviated.
4543 The same fix has been applied.
4549 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4550 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4553 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4554 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4556 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4558 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4559 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4560 Also, for "Auto-Submitted", treat anything other than "no" as
4561 non-personal, in accordance with RFC 3834. (Previously it treated
4562 anything starting "auto-" as non-personal.)
4564 TF/01 The control=submission/name=... option had a problem with syntax
4565 errors if the name included a slash character. The /name= option
4566 now slurps the rest of the string, so it can include any characters
4567 but it must come last in the list of options (after /sender_retain
4570 PH/05 Some modifications to the interface to the fake nameserver for the new
4578 TK/01 Added the "success_on_redirect" address verification option. See
4579 NewStuff for rationale and an example.
4581 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4583 PH/02 Patch to exigrep to allow it to work on syslog lines.
4585 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4586 fread() to skip over the body file's header line, because in Cygwin the
4587 header line is locked and is inaccessible.
4589 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4590 co-exist for some time) to make it clear that it is the Exim ID that is
4591 referenced, not the Message-ID: header line.
4593 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4594 string_format() function, because snprintf() does not exist on all
4597 PH/06 The use of forbid_filter_existstest now also locks out the use of the
4598 ${stat: expansion item.
4600 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
4601 protocol synchronization error", to keep the pedants happy.
4603 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
4604 well as for IRIX systems, when gcc is being used. See the host.c source
4607 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
4609 PH/10 Named domain lists were not working if used in a queue_smtp_domains
4612 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
4613 transport and to the smtp transport in LMTP mode.
4615 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
4617 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
4618 run a filter in a subprocess. This could lead to confusion in subsequent
4619 lookups in the parent process. There should also be a search_tidyup() at
4620 the end of the subprocess.
4622 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
4623 only if the host matched helo_try_verify_hosts, which caused the
4624 verification to occur when the EHLO/HELO command was issued. The ACL just
4625 tested the remembered result. Now, if a previous verification attempt has
4626 not happened, "verify = helo" does it there and then.
4628 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
4630 TK/03 Fix log output including CR from clamd.
4632 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
4633 exist provoked a memory error which could cause a segfault.
4635 PH/15 Installed PCRE 6.2
4637 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
4639 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
4640 of the problem. Specifically, suggested +O2 rather than +O1 for the
4643 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
4645 PH/20 If a delivery was routed to a non-standard port by means of an SRV
4646 record, the port was not correctly logged when the outgoing_port log
4647 selector was set (it logged the transort's default port).
4649 PH/21 Added support for host-specific ports to manualroute, queryprogram,
4650 fallback_hosts, and "hosts" in the smtp transport.
4652 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
4653 host errors such as "Connection refused".
4655 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
4656 authentication with radiusclient 0.4.9:
4658 - Error returned from rc_read_config was caught wrongly
4659 - Username/password not passed on to radius server due to wrong length.
4661 The presumption is that some radiusclient API changes for 4.51/PH/17
4662 were not taken care of correctly. The code is still untested by me (my
4663 Linux distribution still has 0.3.2 of radiusclient), but it was
4664 contributed by a Radius user.
4666 PH/24 When doing a callout, the value of $domain wasn't set correctly when
4667 expanding the "port" option of the smtp transport.
4669 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
4670 while reading a MIME header. Thanks to Tom Hughes for a patch.
4672 PH/24 Include config.h inside local_scan.h so that configuration settings are
4675 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
4676 that in an ACL for RCPT (for example), you can examine exactly what was
4679 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
4680 commands, but it was not correctly comparing the address with the actual
4681 client host address. Thus, it would show the EHLO address in Received:
4682 header lines when this was not necessary.
4684 PH/27 Added the % operator to ${eval:}.
4686 PH/28 Exim tries to create and chdir to its spool directory when it starts;
4687 it should be ignoring failures (because with -C, for example, it has lost
4688 privilege). It wasn't ignoring creation failures other than "already
4691 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
4694 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
4695 because an input line was too long, either on its own, or by virtue of
4696 too many continuations, the temporary file was not being removed, and the
4697 return code was incorrect.
4699 PH/31 Missing "BOOL" in function definition in filtertest.c.
4701 PH/32 Applied Sieve patches from the maintainer.
4703 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
4705 PH/33 Added "verify = not_blind".
4707 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
4708 Local/Makefile (with some defaults set). These are used in built scripts
4709 such as exicyclog, but they have never been used in the exim_install
4710 script (though there are many overriding facilities there). I have
4711 arranged that the exim_install script now takes note of these two
4714 PH/35 Installed configuration files for Dragonfly.
4716 PH/36 When a locally submitted message by a trusted user did not contain a
4717 From: header, and the sender address was obtained from -f or from an SMTP
4718 MAIL command, and the trusted user did not use -F to supply a sender
4719 name, $originator_name was incorrectly used when constructing a From:
4720 header. Furthermore, $originator_name was used for submission mode
4721 messages from external hosts without From: headers in a similar way,
4722 which is clearly wrong.
4724 PH/37 Added control=suppress_local_fixups.
4726 PH/38 When log_selector = +received_sender was set, and the addition of the
4727 sender made the log line's construction buffer exactly full, or one byte
4728 less than full, an overflow happened when the terminating "\n" was
4731 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
4732 when the result of a list match is failure because a DNS lookup failed.
4734 PH/40 RM_COMMAND is now used in the building process.
4736 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4737 the "build-* directories that it finds.
4739 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4740 address in a domain literal was a prefix of an interface address.
4742 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4743 when verifying a sender address, unless rewrite_headers is false.
4745 PH/44 Wrote a long comment about why errors_to addresses are verified as
4746 recipients, not senders.
4748 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4749 the ratelimit ACL was added.
4751 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4753 PH/46 Added extra information about PostgreSQL errors to the error string.
4755 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4756 avoiding the need to install special zones in a real server. This is
4757 backwards compatible; if it can't find the fake resolver, it drops back.
4758 Thus, both old and new test suites can be run.
4760 TF/02 Added util/ratelimit.pl
4762 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4763 clock is set back in time.
4765 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4766 Candler <B.Candler@pobox.com>.
4768 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4769 OK for addresses that are the result of redirections.
4771 PH/48 A number of further additions for the benefit of the new test suite,
4772 including a fake gethostbyname() that interfaces to the fake DNS resolver
4775 TF/06 The fix for widen_domains has also been applied to qualify_single and
4776 search_parents which are the other dnslookup options that can cause
4779 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4782 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4784 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4785 service" if it encountered an MX record with an empty target hostname.
4786 The message is now "an MX or SRV record indicated no SMTP service".
4788 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4789 if the DNS of the sending site is misconfigured. This is quite a
4790 common situation. This change restores the behaviour of treating a
4791 helo verification defer as a failure.
4793 PH/51 If self=fail was set on a router, the bounce message did not include the
4794 actual error message.
4800 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4802 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4803 command itself ended in error, the underlying message about the transport
4804 filter timeout was being overwritten with the pipe command error. Now the
4805 underlying error message should be appended to the second error message.
4807 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4809 PH/02 Reduce the amount of output that "make" produces by default. Full output
4810 can still be requested.
4812 PH/03 The warning log line about a condition test deferring for a "warn" verb
4813 was being output only once per connection, rather than after each
4814 occurrence (because it was using the same function as for successful
4815 "warn" verbs). This seems wrong, so I have changed it.
4817 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4818 it should not have, which might have caused a crash in the right
4819 circumstances, but probably never did.
4821 PH/04 Installed a modified version of Tony Finch's patch to make submission
4822 mode fix the return path as well as the Sender: header line, and to
4823 add a /name= option so that you can make the user's friendly name appear
4826 TF/03 Added the control = fakedefer ACL modifier.
4828 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4829 Mark Lowes for thorough testing.
4831 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4833 TK/03 Merged latest SRS patch from Miles Wilton.
4835 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4836 with the definition in sysexits.h (which is #included earlier).
4837 Fortunately, Exim does not actually use EX_OK. The code used to try to
4838 preserve the sysexits.h value, by assuming that macro definitions were
4839 scanned for macro replacements. I have been disabused of this notion,
4840 so now the code just undefines EX_OK before #including unistd.h.
4842 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4843 in the smtp transport. When a block could not be written in a single
4844 write() function, the timeout was being re-applied to each part-write.
4845 This seems wrong - if the receiver was accepting one byte at a time it
4846 would take for ever. The timeout is now adjusted when this happens. It
4847 doesn't have to be particularly precise.
4849 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4850 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4852 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4853 without a domain if the check to <postmaster@domain> fails.
4855 SC/01 Eximstats: added -xls and the ability to specify output files
4856 (patch written by Frank Heydlauf).
4858 SC/02 Eximstats: use FileHandles for outputting results.
4860 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4862 SC/04 Eximstats: fixed display of large numbers with -nvr option
4864 SC/05 Eximstats: fixed merging of reports with empty tables.
4866 SC/06 Eximstats: added the -include_original_destination flag
4868 SC/07 Eximstats: removed tabs and trailing whitespace.
4870 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4872 TK/06 MBOX spool code: Add real "From " MBOX separator line
4873 so the .eml file is really in mbox format (even though
4874 most programs do not really care). Patch from Alex Miller.
4876 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4877 The latter is generated from $received_to and is only set if the
4878 message has one envelope recipient. SA can use these headers,
4879 obviously out-of-the-box. Patch from Alex Miller.
4881 PH/08 The ${def test on a variable was returning false if the variable's
4882 value was "0", contrary to what the specification has always said!
4883 The result should be true unless the variable is empty.
4885 PH/09 The syntax error of a character other than { following "${if
4886 def:variable_name" (after optional whitespace) was not being diagnosed.
4887 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4888 accidental colon was present, for example, could give incorrect results.
4890 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4891 result is used (not including appendfile, where other changes are about
4894 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4895 This involved changing a lot of size variables from int to off_t. It
4896 should work with maildirs and everything.
4898 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4899 spamd dying while we are connected to it.
4901 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4904 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4905 function for BDB 4.3.
4907 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4909 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4910 the sources for PCRE have been re-arranged and more files are now
4913 PH/15 The code I had for printing potentially long long variables in PH/11
4914 above was not the best (it lost precision). The length of off_t variables
4915 is now inspected at build time, and an appropriate printing format (%ld
4916 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4917 to be "long long int" or "long int". This is needed for the internal
4918 formatting function string_vformat().
4920 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4921 the configuration file to be ":syslog", then the script "guesses" where
4922 the logs files are, rather than using the compiled in default. In our
4923 case the guess is not the same as the compiled default, so the script
4924 suddenly stopped working when I started to use syslog. The patch checks
4925 to see if log_file_path is "". If so, it attempts to read it from exim
4926 with no configuration file to get the compiled in version, before it
4927 falls back to the previous guessing code."
4929 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4930 implementing BATV in an Exim configuration. See NewStuff for the gory
4933 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4934 Makefile that are specific to HP-UX.
4936 PH/18 If the "use_postmaster" option was set for a recipient callout together
4937 with the "random" option, the postmaster address was used as the MAIL
4938 FROM address for the random test, but not for the subsequent recipient
4939 test. It is now used for both.
4941 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4942 patch removes a few documentation additions to RFC 3028, because the
4943 latest draft now contains them. It adds the new en;ascii-case comparator
4944 and a new error check for 8bit text in MIME parts. Comparator and
4945 require names are now matched exactly. I enabled the subaddress
4946 extension, but it is not well tested yet (read: it works for me)."
4948 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4949 rework some of the code of TK/09 above to avoid the hardwired use of
4950 "%lld" and "long long". Replaced the call to snprintf() with a call to
4953 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4954 records point to non-existent hosts", "retry timeout exceeded", and
4955 "retry time not reached for any host after a long failure period".
4957 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4958 experimental DomainKeys support:
4960 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4961 (2) On an error such as an illegally used "control", the wrong name for
4962 the control was given.
4964 These problems did NOT occur unless DomainKeys support was compiled.
4966 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4968 PH/24 Added ${if match_ip condition.
4970 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4971 databases so that it will be absolutely obvious if a crash occurs in the
4972 DB library. This is a regular occurrence (often caused by mis-matched
4975 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4976 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4977 picked up on a user's system that detects such things. There doesn't seem
4978 to be a gcc warning option for this - only an attribute that has to be
4979 put on the function's prototype. It seems that in Fedora Core 4 they have
4980 set this on a number of new functions. No doubt there will be more in due
4983 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4984 specify a transport. However, if an address that was verified by such a
4985 router was the subject of a callout, Exim crashed because it tried to
4986 read the rcpt_include_affixes from the non-existent transport. Now it
4987 just assumes that the setting of that option is false. This bug was
4988 introduced by 4.51/PH/31.
4990 PH/28 Changed -d+all to exclude +memory, because that information is very
4991 rarely of interest, but it makes the output a lot bigger. People tend to
4992 do -d+all out of habit.
4994 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4995 code in os-type was giving problems when libc.so lives in lib64, like on
4998 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4999 aren't the modern standard, and it seems that some systems' include files
5000 don't always have them. Exim was already checking for some of the newer
5001 ones like T_AAAA, and defining it itself. I've added checks for all the
5002 record types that Exim uses.
5004 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5005 not automatically generating a new one, as it is supposed to. This
5006 prevented TLS from working. If the file did exist, but contained invalid
5007 data, a new version was generated, as expected. It was only the case of a
5008 non-existent file that was broken.
5010 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5011 with a change in libdomainkeys > 0.64.
5013 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5014 from DNS. If the selector record carries the flag, it now has
5015 precedence over the domain-wide flag.
5017 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5019 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5020 the use of a port name that isn't defined in /etc/services) occurred, the
5021 message was deferred as in a normal delivery, and thus remained on the
5022 spool, instead of being failed because of the mua_wrapper setting. This
5023 is now fixed, and I tidied up some of the mua_wrapper messages at the
5026 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5027 the messages in a hash of arrays rather than using individual hashes.
5028 This is a bit cleaner and results in dramatic memory savings, albeit
5029 at a slight CPU cost.
5031 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5032 as requested by Marc Sherman.
5034 SC/10 Eximstats: added histograms for user specified patterns as requested
5037 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5039 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5040 fopen() in the content-scanning modules that did not already have it.
5046 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5047 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5049 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5051 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5053 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5054 to test Sieve filters that use "vacation".
5056 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5057 that changes the way the GnuTLS parameters are stored in the cache file.
5058 The new format can be generated externally. For backward compatibility,
5059 if the data in the cache doesn't make sense, Exim assumes it has read an
5060 old-format file, and it generates new data and writes a new file. This
5061 means that you can't go back to an older release without removing the
5064 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5065 work if there are any delivery delays because "one_time" forces the
5066 parent to be marked "delivered", so its unseen clone is never tried
5067 again. For this reason, Exim now forbids the simultaneous setting of
5070 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5071 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5072 entry if you want to know the details. The fix, however, neglected to
5073 consider the case when local delivery batching is involved. The test for
5074 "previously delivered" was not happening when checking to see if an
5075 address could be batched with a previous (undelivered) one; under
5076 certain circumstances this could lead to multiple deliveries to the same
5079 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5080 in its include files, and this causes problems building Exim.
5082 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5083 header_syntax) but Exim was just ignoring anything given after a slash.
5084 In particular, this caused confusion with an attempt to use "verify =
5085 reverse_host_lookup/defer_ok". An error is now given when options are
5086 supplied for verify items that do not have them. (Maybe reverse_host_
5087 lookup should have a defer_ok option, but that's a different point.)
5089 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5090 defined by RFC 821) to 2048, because there were problems with some AUTH
5091 commands, and RFC 1869 says the size should be increased for extended
5092 SMTP commands that take arguments.
5094 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5097 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5098 "unknown" error; now it says that the functionality isn't in the binary.
5100 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5101 an address' error message when a string expansion fails (syntax or
5102 whatever). Otherwise the password may appear in the log. Following change
5103 PH/42 below, there is no longer a chance of it appearing in a bounce
5106 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5108 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5109 of its addresses was ever tried. (Bugzilla bug #2.)
5111 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5112 the result incorrectly in the debug output. (It correctly added a newline
5113 to what was transported.)
5115 TF/01 Added $received_time.
5117 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5118 commented out examples of how to interface to a virus scanner and to
5119 SpamAssassin. Also added commented examples of av_scanner and
5120 spamd_address settings.
5122 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5123 and controls are allowed in which ACLs. There were a couple of minor
5124 errors. Some of the entries in the conditions table (which is a table of
5125 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5126 negation of where the condition IS allowed.
5128 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5130 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5131 header file does not have a version number, so I've had to invent a new
5132 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5133 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5134 radiusclient), but it was contributed by a Radius user.
5136 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5137 files or directories, for OpenSSL.
5139 PH/19 When an Exim process that is running as root has to create an Exim log
5140 file, it does so in a subprocess that runs as exim:exim so as to get the
5141 ownership right at creation (otherwise, other Exim processes might see
5142 the file with the wrong ownership). There was no test for failure of this
5143 fork() call, which would lead to the process getting stuck as it waited
5144 for a non-existent subprocess. Forks do occasionally fail when resources
5145 run out. I reviewed all the other calls to fork(); they all seem to check
5148 PH/20 When checking for unexpected SMTP input at connect time (before writing
5149 the banner), Exim was not dealing correctly with a non-positive return
5150 from the read() function. If the client had disconnected by this time,
5151 the result was a log entry for a synchronization error with an empty
5152 string after "input=" when read() returned zero. If read() returned -1
5153 (an event I could not check), uninitialized data bytes were printed.
5154 There were reports of junk text (parts of files, etc) appearing after
5157 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5159 PH/22 Added support for macro redefinition, and (re)definition in between
5160 driver and ACL definitions.
5162 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5163 forgetting to use the resulting value; it was using the unexpanded value.
5165 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5166 hadn't been configured. The fix is from Juergen Kreileder, who
5167 understands it better than I do:
5169 "Here's what I see happening with three configured cyrus_sasl
5170 authenticators configured (plain, login, cram-md5):
5172 On startup auth_cyrus_sasl_init() gets called for each of these.
5173 This means three calls to sasl_listmech() without a specified mech_list.
5174 => SASL tests which mechs of all available mechs actually work
5175 => three warnings about OTP not working
5176 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5178 With the patch, sasl_listmech() also gets called three times. But now
5179 SASL's mech_list option is set to the server_mech specified in the the
5180 authenticator. Or in other words, the answer from sasl_listmech()
5181 gets limited to just the mech you're testing for (which is different
5183 => the return list contains just 'plain' or 'login', 'cram-md5' or
5184 nothing depending on the value of ob->server_mech.
5186 I've just tested the patch: Authentication still works fine,
5187 unavailable mechs specified in the exim configuration are still
5188 caught, and the auth.log warnings about OTP are gone."
5190 PH/25 When debugging is enabled, the contents of the command line are added
5191 to the debugging output, even when log_selector=+arguments is not
5194 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5195 answer is "GNU", and only if the return is "GNU/something" is the answer
5198 PH/27 $acl_verify_message is now set immediately after the failure of a
5199 verification in an ACL, and so is available in subsequent modifiers. In
5200 particular, the message can be preserved by coding like this:
5202 warn !verify = sender
5203 set acl_m0 = $acl_verify_message
5205 Previously, $acl_verify_message was set only while expanding "message"
5206 and "log_message" when a very denied access.
5208 PH/28 Modified OS/os.c-Linux with
5210 -#ifndef OS_LOAD_AVERAGE
5211 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5213 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5214 nomenclature these days.)
5216 PH/29 Installed patch from the Sieve maintainer that adds the options
5217 sieve_useraddress and sieve_subaddress to the redirect router.
5219 PH/30 In these circumstances:
5220 . Two addresses routed to the same list of hosts;
5221 . First host does not offer TLS;
5222 . First host accepts first address;
5223 . First host gives temporary error to second address;
5224 . Second host offers TLS and a TLS session is established;
5225 . Second host accepts second address.
5226 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5227 and peerdn, if requested) that were in fact used only for the second
5230 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5231 attention to any local part prefix or suffix that was matched by the
5232 router that accepted the address. It now behaves in the same way as it
5233 does for delivery: the affixes are removed from the local part unless
5234 rcpt_include_affixes is set on the transport.
5236 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5237 timeout during the DATA phase of an incoming message.
5239 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5240 applied a patch sanctioned by the Sieve maintainer.
5242 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5243 the uid or gid is negative. A case of a negative gid caused this to be
5244 noticed. The fix allows for either to be negative.
5246 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5247 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5250 PH/36 The change PH/12 above was broken. Fixed it.
5252 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5253 the grounds that routing the same address twice would always produce the
5254 same answer. This might have been true once, but it is certainly no
5255 longer true now. Routing a child address may depend on the previous
5256 routing that produced that child. Some complicated redirection strategies
5257 went wrong when messages had multiple recipients, and made Exim's
5258 behaviour dependent on the order in which the addresses were given.
5260 I have moved the duplicate checking until after the routing is complete.
5261 Exim scans the addresses that are assigned to local and remote
5262 transports, and removes any duplicates. This means that more work will be
5263 done, as duplicates will always all be routed, but duplicates are
5264 presumably rare, so I don't expect this is of any significance.
5266 For deliveries to pipes, files, and autoreplies, the duplicate checking
5267 still happens during the routing process, since they are not going to be
5270 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5271 It corrects a timeout issue with spamd. This is Ian's comment: "The
5272 background is that sometimes spamd either never reads data from a
5273 connection it has accepted, or it never writes response data. The exiscan
5274 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5275 blindly assumes that writes won't block so it may never time out."
5277 PH/39 Allow G after quota size as well as K and M.
5279 PH/40 The value set for $authenticated_id in an authenticator may not contain
5280 binary zeroes or newlines because the value is written to log lines and
5281 to spool files. There was no check on this. Now the value is run through
5282 the string_printing() function so that such characters are converted to
5283 printable escape sequences.
5285 PH/41 $message_linecount is a new variable that contains the total number of
5286 lines in the message. Compare $body_linecount, which is the count for the
5289 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5290 bounce and delay warning messages, except in certain special cases, which
5293 (a) An SMTP error message from a remote host;
5294 (b) A message specified in a :fail: redirection;
5295 (c) A message specified in a "fail" command in a system filter;
5296 (d) A message specified in a FAIL return from the queryprogram router;
5297 (e) A message specified by the cannot_route_message router option.
5299 In these cases only, Exim does include the error details in bounce and
5300 warning messages. There are also a few cases where bland messages such
5301 as "unrouteable address" or "local delivery error" are given.
5303 PH/43 $value is now also set for the "else" part of a ${run expansion.
5305 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5306 being worked on, but at least Exim now implements the latest version to
5309 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5310 process to complete was treated as a delivery failure, a timeout while
5311 writing the message to the pipe was logged, but erroneously treated as a
5312 successful delivery. Such timeouts include transport filter timeouts. For
5313 consistency with the overall process timeout, these timeouts are now
5314 treated as errors, giving rise to delivery failures by default. However,
5315 there is now a new Boolean option for the pipe transport called
5316 timeout_defer, which, if set TRUE, converts the failures into defers for
5317 both kinds of timeout. A transport filter timeout is now identified in
5320 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5321 systems where "make" and "gmake" are different, calling "gmake" at top
5322 level broke things. I've arranged for the value of $(MAKE) to be passed
5323 from the Makefile to this script so that it can call the same version of
5327 A note about Exim versions 4.44 and 4.50
5328 ----------------------------------------
5330 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5331 changes of various kinds. As a consequence, a big documentation update was
5332 needed. This delayed the release for rather longer than seemed good, especially
5333 in the light of a couple of (minor) security issues. Therefore, the changes
5334 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5335 release. So 4.44 and 4.50 are in effect two different branches that both start
5338 I have left the 4.50 change log unchanged; it contains all the changes since
5339 4.43. The change log for 4.44 is below; many of its items are identical to
5340 those for 4.50. This seems to be the most sensible way to preserve the
5341 historical information.
5347 1. Minor wording change to the doc/README.SIEVE file.
5349 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5350 computation of the current number of files was incorrect.
5352 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5353 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5356 4. Give more explanation in the error message when the command for a transport
5357 filter fails to execute.
5359 5. There are several places where Exim runs a non-Exim command in a
5360 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5361 was being done only for the command run by the queryprogram router. It is
5362 now done for all such subprocesses. The other cases are: ${run, transport
5363 filters, and the commands run by the lmtp and pipe transports.
5365 6. Added CONFIGURE_GROUP build-time option.
5367 7. Some older OS have a limit of 256 on the maximum number of file
5368 descriptors. Exim was using setrlimit() to set 1000 as a large value
5369 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5370 systems. I've change it so that if it can't get 1000, it tries for 256.
5372 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5373 was an oversight, and furthermore, ever since the addition of extra
5374 controls (e.g. 4.43/32), the checks on when to allow different forms of
5375 "control" were broken. There should now be diagnostics for all cases when a
5376 control that does not make sense is encountered.
5378 9. Added the /retain_sender option to "control=submission".
5380 10. $recipients is now available in the predata ACL (oversight).
5382 11. Tidy the search cache before the fork to do a delivery from a message
5383 received from the command line. Otherwise the child will trigger a lookup
5384 failure and thereby defer the delivery if it tries to use (for example) a
5385 cached ldap connection that the parent has called unbind on.
5387 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5388 of $address_data from the recipient verification was clobbered by the
5389 sender verification.
5391 13. The value of address_data from a sender verification is now available in
5392 $sender_address_data in subsequent conditions in the ACL statement.
5394 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5396 15. Added a new option "connect=<time>" to callout options, to set a different
5399 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5400 was its contents. (It was OK if the option was not defined at all.)
5402 17. A "Completed" log line is now written for messages that are removed from
5403 the spool by the -Mrm option.
5405 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5406 information about exactly what failed.
5408 19. Added -dd to debug only the daemon process.
5410 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5411 handles timeouts, both on the server side and network timeouts. Renamed the
5412 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5414 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5415 It is now set to "smtps".
5417 22. $host_address is now set to the target address during the checking of
5418 ignore_target_hosts.
5420 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5421 being passed; this would have caused $sender_host_name to have been used if
5422 matching the list had actually called for a host name (not very likely,
5423 since this list is usually IP addresses). A host name is now passed as
5426 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5427 code that specifies a non-restarting handler (typically sigaction() in
5428 modern systems) in an attempt to fix a rare and obscure crash bug.
5430 25. Narrowed the window for a race in the daemon that could cause it to ignore
5431 SIGCHLD signals. This is not a major problem, because they are used only to
5432 wake it up if nothing else does.
5434 26. A malformed maildirsize file could cause Exim to calculate negative values
5435 for the mailbox size or file count. Odd effects could occur as a result.
5436 The maildirsize information is now recalculated if the size or filecount
5439 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5440 support for a long time. Removed HAVE_SYS_VFS_H.
5442 28. Installed the latest version of exipick from John Jetmore.
5444 29. In an address list, if the pattern was not a regular expression, an empty
5445 subject address (from a bounce message) matched only if the pattern was an
5446 empty string. Non-empty patterns were not even tested. This was the wrong
5447 because it is perfectly reasonable to use an empty address as part of a
5448 database query. An empty address is now tested by patterns that are
5449 lookups. However, all the other forms of pattern expect the subject to
5450 contain a local part and a domain, and therefore, for them, an empty
5451 address still always fails if the pattern is not itself empty.
5453 30. Exim went into a mad DNS loop when attempting to do a callout where the
5454 host was specified on an smtp transport, and looking it up yielded more
5455 than one IP address.
5457 31. Re-factored the code for checking spool and log partition space into a
5458 function that finds that data and another that does the check. The former
5459 is then used to implement four new variables: $spool_space, $log_space,
5460 $spool_inodes, and $log_inodes.
5462 32. The RFC2047 encoding function was originally intended for short strings
5463 such as real names; it was not keeping to the 75-character limit for
5464 encoded words that the RFC imposes. It now respects the limit, and
5465 generates multiple encoded words if necessary. To be on the safe side, I
5466 have increased the buffer size for the ${rfc2047: expansion operator from
5469 33. It is now permitted to omit both strings after an "if" condition; if the
5470 condition is true, the result is "true". As before, when the second string
5471 is omitted, a false condition yields an empty string. This makes it less
5472 cumbersome to write custom ACL and router conditions.
5474 34. Failure to deliver a bounce message always caused it to be frozen, even if
5475 there was an errors_to setting on the router. The errors_to setting is now
5478 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5479 canonical form (fully expanded) before being placed in
5480 $sender_host_address.
5482 36. The table in the code that translates DNS record types into text (T_A to
5483 "A" for instance) was missing entries for NS and CNAME. It is just possible
5484 that this could have caused confusion if both these types were looked up
5485 for the same domain, because the text type is used as part of Exim's
5486 per-process caching. But the chance of anyone hitting this buglet seems
5489 37. The dnsdb lookup has been extended in a number of ways.
5491 (1) There is a new type, "zns", which walks up the domain tree until it
5492 finds some nameserver records. It should be used with care.
5494 (2) There is a new type, "mxh", which is like "mx" except that it returns
5495 just the host names, not the priorities.
5497 (3) It is now possible to give a list of domains (or IP addresses) to be
5498 looked up. The behaviour when one of the lookups defers can be
5499 controlled by a keyword.
5501 (4) It is now possible to specify the separator character for use when
5502 multiple records are returned.
5504 38. The dnslists ACL condition has been extended: it is now possible to supply
5505 a list of IP addresses and/or domains to be looked up in a particular DNS
5508 39. Added log_selector=+queue_time_overall.
5510 40. When running the queue in the test harness, wait just a tad after forking a
5511 delivery process, to get repeatability of debugging output.
5513 41. Include certificate and key file names in error message when GnuTLS fails
5514 to set them up, because the GnuTLS error message doesn't include the name
5515 of the failing file when there is a problem reading it.
5517 42. Allow both -bf and -bF in the same test run.
5519 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5521 44. The "Exiscan patch" is now merged into the mainline Exim source.
5523 45. Sometimes the final signoff response after QUIT could fail to get
5524 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5525 < 0 before doing a fflush(). This bug looks as though it goes back to the
5526 introduction of TLS in release 3.20, but "sometimes" must have been rare
5527 because the tests only now provoked it.
5529 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5530 (this can affect the format of dates).
5532 47. exim_tidydb, when checking for the continued existence of a message for
5533 which it has found a message-specific retry record, was not finding
5534 messages that were in split spool directories. Consequently, it was
5535 deleting retry records that should have stayed in existence.
5537 48. Steve fixed some bugs in eximstats.
5539 49. The SPA authentication driver was not abandoning authentication and moving
5540 on to the next authenticator when an expansion was forced to fail,
5541 contradicting the general specification for all authenticators. Instead it
5542 was generating a temporary error. It now behaves as specified.
5544 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5545 (the order specifies the preference for clients). The order is now AES256,
5546 AES128, 3DES, ARCFOUR128.
5548 51. Small patch to Sieve code - explicitly set From: when generating an
5551 52. Exim crashed if a remote delivery caused a very long error message to be
5552 recorded - for instance if somebody sent an entire SpamAssassin report back
5553 as a large number of 550 error lines. This bug was coincidentally fixed by
5554 increasing the size of one of Exim's internal buffers (big_buffer) that
5555 happened as part of the Exiscan merge. However, to be on the safe side, I
5556 have made the code more robust (and fixed the comments that describe what
5559 53. Now that there can be additional text after "Completed" in log lines (if
5560 the queue_time_overall log selector is set), a one-byte patch to exigrep
5561 was needed to allow it to recognize "Completed" as not the last thing in
5564 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5565 patch that reportedly fixes this has been added. I am not expert enough to
5566 create a test for it. This is what the patch creator wrote:
5568 "I found a little strange behaviour of ldap code when working with
5569 Windows 2003 AD Domain, where users was placed in more than one
5570 Organization Units. When I tried to give exim partial DN, the exit code
5571 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5572 But simultaneously result of request was absolutely normal ldap result,
5573 so I produce this patch..."
5575 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5576 REFERENCE, so I have modified the code to exclude the patch when that macro
5579 55. Some experimental protocols are using DNS PTR records for new purposes. The
5580 keys for these records are domain names, not reversed IP addresses. The
5581 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5582 leaves it alone. Component reversal etc. now happens only for IP addresses.
5585 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5587 57. Double the size of the debug message buffer (to 2048) so that more of very
5588 long debug lines gets shown.
5590 58. The exicyclog utility now does better if the number of log files to keep
5591 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5593 59. Two changes related to the smtp_active_hostname option:
5595 (1) $smtp_active_hostname is now available as a variable.
5596 (2) The default for smtp_banner uses $smtp_active_hostname instead
5597 of $primary_hostname.
5599 60. The host_aton() function is supposed to be passed a string that is known
5600 to be a valid IP address. However, in the case of IPv6 addresses, it was
5601 not checking this. This is a hostage to fortune. Exim now panics and dies
5602 if the condition is not met. A case was found where this could be provoked
5603 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5604 components; fortuitously, this particular loophole had already been fixed
5605 by change 4.50/55 above.
5607 If there are any other similar loopholes, the new check in host_aton()
5608 itself should stop them being exploited. The report I received stated that
5609 data on the command line could provoke the exploit when Exim was running as
5610 exim, but did not say which command line option was involved. All I could
5611 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5612 running as the user.
5615 61. There was a buffer overflow vulnerability in the SPA authentication code
5616 (which came originally from the Samba project). I have added a test to the
5617 spa_base64_to_bits() function which I hope fixes it.
5620 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
5621 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
5623 63. The daemon start-up calls getloadavg() while still root for those OS that
5624 need the first call to be done as root, but it missed one case: when
5625 deliver_queue_load_max is set with deliver_drop_privilege. This is
5626 necessary for the benefit of the queue runner, because there is no re-exec
5627 when deliver_drop_privilege is set.
5629 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
5630 This has been fixed.
5632 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5633 was in use, was not putting the data itself into the right store pool;
5634 consequently, it could be overwritten for a subsequent message in the same
5635 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5638 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
5640 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
5641 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
5642 the function were treating the return as a boolean value, which happened to
5643 work because 0=false and not-0=true, but is not correct code.
5645 68. The host_aton() function was not handling scoped IPv6 addresses (those
5646 with, for example, "%eth0" on the end) correctly.
5648 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
5649 negated items (that is, ~something) in unsigned ints. Some compilers
5650 apparently mutter when there is no cast.
5652 70. If an address verification called from an ACL failed, and did not produce a
5653 user-specific message (i.e. there was only a "system" message), nothing was
5654 put in $acl_verify_message. In this situation, it now puts the system
5657 71. Change 4.23/11 added synchronization checking at the start of an SMTP
5658 session; change 4.31/43 added the unwanted input to the log line - except
5659 that it did not do this in the start of session case. It now does.
5661 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
5662 This is wrong and can cause the other end to generate a synchronization
5663 error if it is another Exim or anything else that does the synchronization
5664 check. A QUIT command is no longer sent after a timeout.
5666 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
5667 during host lookups.
5669 74. The defer_ok option of callout verification was not working if it was used
5670 when verifying addresses in header lines, that is, for this case:
5672 verify = header_sender/callout=defer_ok
5674 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
5675 those file descriptors could be used for SMTP connections. If anything
5676 wrote to stderr (the example that came up was "warn" in embedded Perl), it
5677 could be sent to the SMTP client, causing chaos. The daemon now opens
5678 stdin, stdout, and stderr to /dev/null when it puts itself into the
5681 76. Arrange for output from Perl's "warn" command to be written to Exim's main
5682 log by default. The user can override this with suitable Perl magic.
5684 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
5685 the log message when discard triggers, was not working for the DATA ACL or
5686 for the non-SMTP ACL.
5688 78. Error message wording change in sieve.c.
5690 79. If smtp_accept_max_per_host was set, the number of connections could be
5691 restricted to fewer than expected, because the daemon was trying to set up
5692 a new connection before checking whether the processes handling previous
5693 connections had finished. The check for completed processes is now done
5694 earlier. On busy systems, this bug wouldn't be noticed because something
5695 else would have woken the daemon, and it would have reaped the completed
5698 80. If a message was submitted locally by a user whose login name contained one
5699 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
5700 It caused a spool format error. I have fixed the spool reading code. A
5701 related problem was that the "from" clause in the Received: line became
5702 illegal because of the space(s). It is now covered by ${quote_local_part.
5704 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
5707 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
5709 83. Patch from Sieve maintainer to fix unterminated string problem in
5710 "vacation" handling.
5712 84. Some minor changes to the Linux configuration files to help with other
5713 OS variants using glibc.
5715 85. One more patch for Sieve to update vacation handling to latest spec.
5718 ----------------------------------------------------
5719 See the note above about the 4.44 and 4.50 releases.
5720 ----------------------------------------------------
5726 1. Change 4.43/35 introduced a bug that caused file counts to be
5727 incorrectly computed when quota_filecount was set in an appendfile
5730 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5731 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5734 3. Give more explanation in the error message when the command for a transport
5735 filter fails to execute.
5737 4. There are several places where Exim runs a non-Exim command in a
5738 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5739 was being done only for the command run by the queryprogram router. It is
5740 now done for all such subprocesses. The other cases are: ${run, transport
5741 filters, and the commands run by the lmtp and pipe transports.
5743 5. Some older OS have a limit of 256 on the maximum number of file
5744 descriptors. Exim was using setrlimit() to set 1000 as a large value
5745 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5746 systems. I've change it so that if it can't get 1000, it tries for 256.
5748 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5749 was an oversight, and furthermore, ever since the addition of extra
5750 controls (e.g. 4.43/32), the checks on when to allow different forms of
5751 "control" were broken. There should now be diagnostics for all cases when a
5752 control that does not make sense is encountered.
5754 7. $recipients is now available in the predata ACL (oversight).
5756 8. Tidy the search cache before the fork to do a delivery from a message
5757 received from the command line. Otherwise the child will trigger a lookup
5758 failure and thereby defer the delivery if it tries to use (for example) a
5759 cached ldap connection that the parent has called unbind on.
5761 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5762 of $address_data from the recipient verification was clobbered by the
5763 sender verification.
5765 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5766 was its contents. (It was OK if the option was not defined at all.)
5768 11. A "Completed" log line is now written for messages that are removed from
5769 the spool by the -Mrm option.
5771 12. $host_address is now set to the target address during the checking of
5772 ignore_target_hosts.
5774 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5775 being passed; this would have caused $sender_host_name to have been used if
5776 matching the list had actually called for a host name (not very likely,
5777 since this list is usually IP addresses). A host name is now passed as
5780 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5781 code that specifies a non-restarting handler (typically sigaction() in
5782 modern systems) in an attempt to fix a rare and obscure crash bug.
5784 15. Narrowed the window for a race in the daemon that could cause it to ignore
5785 SIGCHLD signals. This is not a major problem, because they are used only to
5786 wake it up if nothing else does.
5788 16. A malformed maildirsize file could cause Exim to calculate negative values
5789 for the mailbox size or file count. Odd effects could occur as a result.
5790 The maildirsize information is now recalculated if the size or filecount
5793 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5794 support for a long time. Removed HAVE_SYS_VFS_H.
5796 18. Updated exipick to current release from John Jetmore.
5798 19. Allow an empty sender to be matched against a lookup in an address list.
5799 Previously the only cases considered were a regular expression, or an
5802 20. Exim went into a mad DNS lookup loop when doing a callout where the
5803 host was specified on the transport, if the DNS lookup yielded more than
5806 21. The RFC2047 encoding function was originally intended for short strings
5807 such as real names; it was not keeping to the 75-character limit for
5808 encoded words that the RFC imposes. It now respects the limit, and
5809 generates multiple encoded words if necessary. To be on the safe side, I
5810 have increased the buffer size for the ${rfc2047: expansion operator from
5813 22. Failure to deliver a bounce message always caused it to be frozen, even if
5814 there was an errors_to setting on the router. The errors_to setting is now
5817 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5818 canonical form (fully expanded) before being placed in
5819 $sender_host_address.
5821 24. Updated eximstats to version 1.33
5823 25. Include certificate and key file names in error message when GnuTLS fails
5824 to set them up, because the GnuTLS error message doesn't include the name
5825 of the failing file when there is a problem reading it.
5827 26. Expand error message when OpenSSL has problems setting up cert/key files.
5830 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5831 (this can affect the format of dates).
5833 28. exim_tidydb, when checking for the continued existence of a message for
5834 which it has found a message-specific retry record, was not finding
5835 messages that were in split spool directories. Consequently, it was
5836 deleting retry records that should have stayed in existence.
5838 29. eximstats updated to version 1.35
5839 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5840 1.35 - bugfix such that pie charts by volume are generated correctly
5842 30. The SPA authentication driver was not abandoning authentication and moving
5843 on to the next authenticator when an expansion was forced to fail,
5844 contradicting the general specification for all authenticators. Instead it
5845 was generating a temporary error. It now behaves as specified.
5847 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5848 (the order specifies the preference for clients). The order is now AES256,
5849 AES128, 3DES, ARCFOUR128.
5851 31. Small patch to Sieve code - explicitly set From: when generating an
5854 32. Exim crashed if a remote delivery caused a very long error message to be
5855 recorded - for instance if somebody sent an entire SpamAssassin report back
5856 as a large number of 550 error lines. This bug was coincidentally fixed by
5857 increasing the size of one of Exim's internal buffers (big_buffer) that
5858 happened as part of the Exiscan merge. However, to be on the safe side, I
5859 have made the code more robust (and fixed the comments that describe what
5862 33. Some experimental protocols are using DNS PTR records for new purposes. The
5863 keys for these records are domain names, not reversed IP addresses. The
5864 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5865 leaves it alone. Component reversal etc. now happens only for IP addresses.
5868 34. The host_aton() function is supposed to be passed a string that is known
5869 to be a valid IP address. However, in the case of IPv6 addresses, it was
5870 not checking this. This is a hostage to fortune. Exim now panics and dies
5871 if the condition is not met. A case was found where this could be provoked
5872 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5873 components; fortuitously, this particular loophole had already been fixed
5874 by change 4.50/55 or 4.44/33 above.
5876 If there are any other similar loopholes, the new check in host_aton()
5877 itself should stop them being exploited. The report I received stated that
5878 data on the command line could provoke the exploit when Exim was running as
5879 exim, but did not say which command line option was involved. All I could
5880 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5881 running as the user.
5884 35. There was a buffer overflow vulnerability in the SPA authentication code
5885 (which came originally from the Samba project). I have added a test to the
5886 spa_base64_to_bits() function which I hope fixes it.
5889 36. The daemon start-up calls getloadavg() while still root for those OS that
5890 need the first call to be done as root, but it missed one case: when
5891 deliver_queue_load_max is set with deliver_drop_privilege. This is
5892 necessary for the benefit of the queue runner, because there is no re-exec
5893 when deliver_drop_privilege is set.
5895 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5896 was in use, was not putting the data itself into the right store pool;
5897 consequently, it could be overwritten for a subsequent message in the same
5898 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5901 38. Sometimes the final signoff response after QUIT could fail to get
5902 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5903 < 0 before doing a fflush(). This bug looks as though it goes back to the
5904 introduction of TLS in release 3.20, but "sometimes" must have been rare
5905 because the tests only now provoked it.
5911 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5912 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5913 Now it returns an integer. A number of calls were still expecting a T/F
5914 return. Fortuitously, in all cases, the tests worked in OK situations,
5915 which is the norm. However, things would have gone wrong on any write
5916 failures on the smtp file descriptor. This function is used when sending
5917 messages over SMTP and also when doing verify callouts.
5919 2. When Exim is called to do synchronous delivery of a locally submitted
5920 message (the -odf or -odi options), it no longer closes stderr before doing
5923 3. Implemented the mua_wrapper option.
5925 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5927 5. Implemented the functions header_remove(), header_testname(),
5928 header_add_at_position(), and receive_remove_recipient(), and exported them
5931 6. If an ACL "warn" statement specified the addition of headers, Exim already
5932 inserted X-ACL-Warn: at the start if there was no header name. However, it
5933 was not making this test for the second and subsequent header lines if
5934 there were newlines in the string. This meant that an invalid header could
5935 be inserted if Exim was badly configured.
5937 7. Allow an ACL "warn" statement to add header lines at the start or after all
5938 the Received: headers, as well as at the end.
5940 8. Added the rcpt_4xx retry error code.
5942 9. Added postmaster_mailfrom=xxx to callout verification option.
5944 10. Added mailfrom=xxxx to the callout verification option, for verify=
5947 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5948 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5950 12. Inserted some casts to stop certain compilers warning when using pointer
5951 differences as field lengths or precisions in printf-type calls (mostly
5952 affecting debugging statements).
5954 13. Added optional readline() support for -be (dynamically loaded).
5956 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5957 same clock tick as a message's arrival, so that its received time was the
5958 same as the "first fail" time on the retry record, and that message
5959 remained on the queue past the ultimate address timeout, every queue runner
5960 would try a delivery (because it was past the ultimate address timeout) but
5961 after another failure, the ultimate address timeout, which should have then
5962 bounced the address, did not kick in. This was a "< instead of <=" error;
5963 in most cases the first failure would have been in the next clock tick
5964 after the received time, and all would be well.
5966 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5967 being recognized when the domain list was tested by the match_domain
5968 condition in an expansion string.
5970 16. Added the ${str2b64: operator.
5972 17. Exim was always calling setrlimit() to set a large limit for the number of
5973 processes, without checking whether the existing limit was already
5974 adequate. (It did check for the limit on file descriptors.) Furthermore,
5975 errors from getrlimit() and setrlimit() were being ignored. Now they are
5976 logged to the main and panic logs, but Exim does carry on, to try to do its
5977 job under whatever limits there are.
5979 18. Imported PCRE 5.0.
5981 19. Trivial typo in log message " temporarily refused connection" (the leading
5984 20. If the log selector return_path_on_delivery was set and an address was
5985 redirected to /dev/null, the delivery process crashed because it assumed
5986 that a return path would always be set for a "successful" delivery. In this
5987 case, the whole delivery is bypassed as an optimization, and therefore no
5990 21. Internal re-arrangement: the function for sending a challenge and reading
5991 a response while authentication was assuming a zero-terminated challenge
5992 string. It's now changed to take a pointer and a length, to allow for
5993 binary data in such strings.
5995 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5997 23. Exim was not respecting finduser_retries when seeking the login of the
5998 uid under which it was called; it was always trying 10 times. (The default
5999 setting of finduser_retries is zero.) Also, it was sleeping after the final
6000 failure, which is pointless.
6002 24. Implemented tls_on_connect_ports.
6004 25. Implemented acl_smtp_predata.
6006 26. If the domain in control=submission is set empty, Exim assumes that the
6007 authenticated id is a complete email address when it generates From: or
6008 Sender: header lines.
6010 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6011 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6012 chown and chgrp in /bin and hostname in /usr/bin.
6014 28. Exim was keeping the "process log" file open after each use, just as it
6015 does for the main log. This opens the possibility of it remaining open for
6016 long periods when the USR1 signal hits a daemon. Occasional processlog
6017 errors were reported, that could have been caused by this. Anyway, it seems
6018 much more sensible not to leave this file open at all, so that is what now
6021 29. The long-running daemon process does not normally write to the log once it
6022 has entered its main loop, and it closes the log before doing so. This is
6023 so that log files can straightforwardly be renamed and moved. However,
6024 there are a couple of unusual error situations where the daemon does write
6025 log entries, and I had neglected to close the log afterwards.
6027 30. The text of an SMTP error response that was received during a remote
6028 delivery was being truncated at 512 bytes. This is too short for some of
6029 the long messages that one sometimes sees. I've increased the limit to
6032 31. It is now possible to make retry rules that apply only when a message has a
6033 specific sender, in particular, an empty sender.
6035 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6036 it possible to be selective about when SMTP synchronization is enforced.
6038 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6040 32. Implemented hosts_connection_nolog.
6042 33. Added an ACL for QUIT.
6044 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6047 35. Added mailbox_size and mailbox_filecount to appendfile.
6049 36. Added control = no_multiline_responses to ACLs.
6051 37. There was a bug in the logic of the code that waits for the clock to tick
6052 in the case where the clock went backwards by a substantial amount such
6053 that the microsecond fraction of "now" was more than the microsecond
6054 fraction of "then" (but the whole seconds number was less).
6056 38. Added support for the libradius Radius client library this is found on
6057 FreeBSD (previously only the radiusclient library was supported).
6063 1. When certain lookups returned multiple values in the form name=value, the
6064 quoting of the values was not always being done properly. Specifically:
6065 (a) If the value started with a double quote, but contained no whitespace,
6067 (b) If the value contained whitespace other than a space character (i.e.
6068 tabs or newlines or carriage returns) it was not quoted.
6069 This fix has been applied to the mysql and pgsql lookups by writing a
6070 separate quoting function and calling it from the lookup code. The fix
6071 should probably also be applied to nisplus, ibase and oracle lookups, but
6072 since I cannot test any of those, I have not disturbed their existing code.
6074 2. A hit in the callout cache for a specific address caused a log line with no
6075 reason for rejecting RCPT. Now it says "Previous (cached) callout
6076 verification failure".
6078 3. There was an off-by-one bug in the queryprogram router. An over-long
6079 return line was truncated at 256 instead of 255 characters, thereby
6080 overflowing its buffer with the terminating zero. As well as fixing this, I
6081 have increased the buffer size to 1024 (and made a note to document this).
6083 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6084 when Exim is waiting for an SMTP response from a remote server, Exim
6085 restarts its select() call on the socket, thereby resetting its timeout.
6086 This is not a problem when such interrupts are rare. Somebody set up a cron
6087 job to run exiwhat every 2 minutes, which is less than the normal select()
6088 timeout (5 or 10 minutes). This meant that the select() timeout never
6089 kicked in because it was always reset. I have fixed this by comparing the
6090 time when an interrupt arrives with the time at the start of the first call
6091 to select(). If more time than the timeout has elapsed, the interrupt is
6092 treated as a timeout.
6094 5. Some internal re-factoring in preparation for the addition of Sieve
6095 extensions (by MH). In particular, the "personal" test is moved to a
6096 separate function, and given an option for scanning Cc: and Bcc: (which is
6097 not set for Exim filters).
6099 6. When Exim created an email address using the login of the caller as the
6100 local part (e.g. when creating a From: or Sender: header line), it was not
6101 quoting the local part when it contained special characters such as @.
6103 7. Installed new OpenBSD configuration files.
6105 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6106 try to make them clearer.
6108 9. Callout options, other than the timeout value, were being ignored when
6109 verifying sender addresses in header lines. For example, when using
6111 verify = header_sender/callout=no_cache
6113 the cache was (incorrectly) being used.
6115 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6116 only the Cygwin environment.
6118 11. When return_path_on_delivery was set as a log selector, if different remote
6119 addresses in the same message used different return paths and parallel
6120 remote delivery occurred, the wrong values would sometimes be logged.
6121 (Whenever a remote delivery process finished, the return path value from
6122 the most recently started remote delivery process was logged.)
6124 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6125 lines when AUTH and/or TLS are in use. This is the "received protocol"
6126 field. Exim used to use "asmtp" for authenticated SMTP, without any
6127 indication (in the protocol name) for TLS use. Now it follows the RFC and
6128 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6129 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6130 names appear in log lines as well as in Received: header lines.
6132 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6133 extensions, and comparison tests, and to fix some bugs.
6135 14. Changes to the "personal" filter test:
6137 (1) The test was buggy in that it was just doing the equivalent of
6138 "contains" tests on header lines. For example, if a user's address was
6139 anne@some.where, the "personal" test would incorrectly be true for
6141 To: susanne@some.where
6143 This test is now done by extracting each address from the header in turn,
6144 and checking the entire address. Other tests that are part of "personal"
6145 are now done using regular expressions (for example, to check local parts
6146 of addresses in From: header lines).
6148 (2) The list of non-personal local parts in From: addresses has been
6149 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6150 taken from the Sieve specification recommendations.
6152 (3) If the message contains any header line starting with "List-" it is
6153 treated as non-personal.
6155 (4) The test for "circular" in the Subject: header line has been removed
6156 because it now seems ill-conceived.
6158 15. Minor typos in src/EDITME comments corrected.
6160 16. Installed latest exipick from John Jetmore.
6162 17. If headers_add on a router specified a text string that was too long for
6163 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6164 of string_sprintf() is now avoided.
6166 18. $message_body_size was not set (it was always zero) when running the DATA
6167 ACL and the local_scan() function.
6169 19. For the "mail" command in an Exim filter, no default was being set for
6170 the once_repeat time, causing a random time value to be used if "once" was
6171 specified. (If the value happened to be <= 0, no repeat happened.) The
6172 default is now 0s, meaning "never repeat". The "vacation" command was OK
6173 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6174 (I found it when inspecting the code).
6176 20. There is now an overall timeout for performing a callout verification. It
6177 defaults to 4 times the callout timeout, which applies to individual SMTP
6178 commands during the callout. The overall timeout applies when there is more
6179 than one host that can be tried. The timeout is checked before trying the
6180 next host. This prevents very long delays if there are a large number of
6181 hosts and all are timing out (e.g. when the network connections are timing
6182 out). The value of the overall timeout can be changed by specifying an
6183 additional sub-option for "callout", called "maxwait". For example:
6185 verify = sender/callout=5s,maxwait=20s
6187 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6188 to the end before writing, but this should make it even safer).
6190 22. Exim was forgetting that it had advertised PIPELINING for the second and
6191 subsequent messages on an SMTP connection. It was also not resetting its
6192 memory on STARTTLS and an internal HELO.
6194 23. When Exim logs an SMTP synchronization error within a session, it now
6195 records whether PIPELINING has been advertised or not.
6197 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6198 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6199 rather than long int.
6201 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6203 26. Added the never_mail option to autoreply.
6209 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6210 crash if the getsockname() call failed; this can happen if a connection is
6211 closed very soon after it is established. The problem was simply in the
6212 order in which certain operations were done, causing Exim to try to write
6213 to the SMTP stream before it had set up the file descriptor. The bug has
6214 been fixed by making things happen in the correct order.
6220 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6221 before the connection was closed, thus losing the rejection response.
6223 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6224 some early Solaris releases, but causes trouble in current releases where
6225 socklen_t is defined.
6227 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6230 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6233 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6234 that releasing the top of it at the end releases what was used for sub-
6235 expansions (unless the block got too big). However, discard this block if
6236 the first thing is a variable or header, so that we can use its block when
6237 it is dynamic (useful for very large $message_headers, for example).
6239 6. Lookups now cache *every* query, not just the most recent. A new, separate
6240 store pool is used for this. It can be recovered when all lookup caches are
6241 flushed. Lookups now release memory at the end of their result strings.
6242 This has involved some general refactoring of the lookup sources.
6244 7. Some code has been added to the store_xxx() functions to reduce the amount
6245 of flapping under certain conditions.
6247 8. log_incoming_interface used to affect only the <= reception log lines. Now
6248 it causes the local interface and port to be added to several more SMTP log
6249 lines, for example "SMTP connection from", and rejection lines.
6251 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6253 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6255 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6256 could be overwritten at the end of the current message (or the start of a
6257 new message if it was set in a HELO ACL). The value is now preserved for
6258 the duration of the SMTP connection.
6260 12. If a transport had a headers_rewrite setting, and a matching header line
6261 contained an unqualified address, that address was qualified, even if it
6262 did not match any rewriting rules. The underlying bug was that the values
6263 of the flags that permit the existence of unqualified sender and recipient
6264 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6265 non-local messages, and by -bnq for local messages) were not being
6266 preserved with the message after it was received.
6268 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6269 "next input=" as part of the text comprising the host identity instead of
6270 the correct text. The code was using the same buffer for two different
6271 strings. However, depending on which order the printing function evaluated
6272 its arguments, the bug did not always show up. Under Linux, for example, my
6273 test suite worked just fine.
6275 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6276 doesn't work with some older versions of Perl. It has been changed to "my",
6277 which in any case is probably the better facility to use.
6279 15. A really picky compiler found some instances of statements for creating
6280 error messages that either had too many or two few arguments for the format
6283 16. The size of the buffer for calls to the DNS resolver has been increased
6284 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6285 for addresses that have a lot of PTR records. This alleviates a problem; it
6286 does not fully solve it.
6288 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6289 buffer now truncates the list and logs the incident, which is the same
6290 action as happens when Exim is looking up a host name and its aliases.
6291 Previously in this situation something unpredictable would happen;
6292 sometimes it was "internal error: store_reset failed".
6294 18. If a server dropped the connection unexpectedly when an Exim client was
6295 using GnuTLS and trying to read a response, the client delivery process
6296 crashed while trying to generate an error log message.
6298 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6299 string, for example:
6301 warn message = H1: something\nH2: something
6303 the text was added as a single header line from Exim's point of view
6304 though it ended up OK in the delivered message. However, searching for the
6305 second and subsequent header lines using $h_h2: did not work. This has been
6306 fixed. Similarly, if a system filter added multiple headers in this way,
6307 the routers could not see them.
6309 20. Expanded the error message when iplsearch is called with an invalid key to
6310 suggest using net-iplsearch in a host list.
6312 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6313 ACLs are no longer actually imposed (and a message to that effect is
6316 22. If a "gecos" field in a passwd entry contained escaped characters, in
6317 particular, if it contained a \" sequence, Exim got it wrong when building
6318 a From: or a Sender: header from that name. A second bug also caused
6319 incorrect handling when an unquoted " was present following a character
6320 that needed quoting.
6322 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6323 was not being matched caselessly.
6325 24. Arranged for all hyphens in the exim.8 source to be escaped with
6328 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6329 properly. Recipient callout cache records were still being keyed to include
6330 the sender, even when use_sender was set false. This led to far more
6331 callouts that were necessary. The sender is no longer included in the key
6332 when use_sender is false.
6334 26. Added "control = submission" modifier to ACLs.
6336 27. Added the ${base62d: operator to decode base 62 numbers.
6338 28. dnsdb lookups can now access SRV records.
6340 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6341 the configuration file.
6343 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6344 (-v) mode. This makes the output for a verbose queue run more intelligible.
6346 31. Added a use_postmaster feature to recipient callouts.
6348 32. Added the $body_zerocount variable, containing the number of binary zero
6349 bytes in the message body.
6351 33. The time of last modification of the "new" subdirectory is now used as the
6352 "mailbox time last read" when there is a quota error for a maildir
6355 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6357 35. Added +ignore_unknown as a special item in host lists.
6359 36. Code for decoding IPv6 addresses in host lists is now included, even if
6360 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6361 address was recognized as an IP address, but was then not correctly decoded
6362 into binary, causing unexpected and incorrect effects when compared with
6369 1. Very minor rewording of debugging text in manualroute to say "list of
6370 hosts" instead of "hostlist".
6372 2. If verify=header_syntax was set, and a header line with an unqualified
6373 address (no domain) and a large number of spaces between the end of the
6374 name and the colon was received, the reception process suffered a buffer
6375 overflow, and (when I tested it) crashed. This was caused by some obsolete
6376 code that should have been removed. The fix is to remove it!
6378 3. When running in the test harness, delay a bit after writing a bounce
6379 message to get a bit more predictability in the log output.
6381 4. Added a call to search_tidyup() just before forking a reception process. In
6382 theory, someone could use a lookup in the expansion of smtp_accept_max_
6383 per_host which, without the tidyup, could leave open a database connection.
6385 5. Added the variables $recipient_data and $sender_data which get set from a
6386 lookup success in an ACL "recipients" or "senders" condition, or a router
6387 "senders" option, similar to $domain_data and $local_part_data.
6389 6. Moved the writing of debug_print from before to after the "senders" test
6392 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6393 problems for message scanning, either using a data ACL, or using
6394 local_scan() because the Received: header was not generated till after they
6395 were called (in order to set the time as the time of reception completion).
6396 I have revised the way this works. The header is now generated after the
6397 body is received, but before the ACL or local_scan() are called. After they
6398 are run, the timestamp in the header is updated.
6404 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6405 before starting a queue runner without re-exec. This happened only when
6406 deliver_drop_privilege was set or when the Exim user was set to root. The
6407 effect of the bug was that timeouts during subsequent deliveries caused
6408 crashes instead of being properly handled. The handler is now left at its
6409 default (and expected) setting.
6411 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6412 message, again when deliver_drop_privilege is set or Exim is run as root.
6413 The bug described in (1) was not present in this case, but the tidying up
6414 of the other signals was missing. I have made the two cases consistent.
6416 3. The ignore_target_hosts setting on a manualroute router was being ignored
6417 for hosts that were looked up using the /MX notation.
6419 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6422 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6423 operated on the sender address. After changing the $sender_address to <>
6424 for the sender address verify, Exim was re-instated it as the original
6425 (before rewriting) address, but remembering that it had rewritten it, so it
6426 wasn't rewriting it again. This bug also had the effect of breaking the
6427 sender address verification caching when the sender address was rewritten.
6429 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6430 This has been changed so that if the ip literal address matches
6431 ignore_target_hosts, the router declines.
6433 7. Added expansion conditions match_domain, match_address, and match_local_
6434 part (NOT match_host).
6436 8. The placeholder for the Received: header didn't have a length field set.
6438 9. Added code to Exim itself and to exim_lock to test for a specific race
6439 condition that could lead to file corruption when using MBX delivery. The
6440 issue is with the lockfile that is created in /tmp. If this file is removed
6441 after a process has opened it but before that process has acquired a lock,
6442 there is the potential for a second process to recreate the file and also
6443 acquire a lock. This could lead to two Exim processes writing to the file
6444 at the same time. The added code performs the same test as UW imapd; it
6445 checks after acquiring the lock that its file descriptor still refers to
6446 the same named file.
6448 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6449 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6452 11. Added the smtp_active_hostname option. If used, this will typically be made
6453 to depend on the incoming interface address. Because $interface_address is
6454 not set up until the daemon has forked a reception process, error responses
6455 that can happen earlier (such as "too many connections") no longer contain
6458 12. If an expansion in a condition on a "warn" statement fails because a lookup
6459 defers, the "warn" statement is abandoned, and the next ACL statement is
6460 processed. Previously this caused the whole ACL to be aborted.
6462 13. Added the iplsearch lookup type.
6464 14. Added ident_timeout as a log selector.
6466 15. Added tls_certificate_verified as a log selector.
6468 16. Added a global option tls_require_ciphers (compare the smtp transport
6469 option of the same name). This controls incoming TLS connections.
6471 17. I finally figured out how to make tls_require_ciphers do a similar thing
6472 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6473 before starting the TLS session.
6475 18. Tabs are now shown as \t in -bP output.
6477 19. If the log selector return_path_on_delivery was set, Exim crashed when
6478 bouncing a message because it had too many Received: header lines.
6480 20. If two routers both had headers_remove settings, and the first one included
6481 a superfluous trailing colon, the final name in the first list and the
6482 first name in the second list were incorrectly joined into one item (with a
6483 colon in the middle).
6489 1. Added -C and -D options to the exinext utility, mainly to make it easier
6490 to include in the automated testing, but these could be helpful when
6491 multiple configurations are in use.
6493 2. The exinext utility was not formatting the output nicely when there was
6494 an alternate port involved in the retry record key, nor when there was a
6495 message id as well (for retries that were specific to a specific message
6496 and a specific host). It was also confused by IPv6 addresses, because of
6497 the additional colons they contain. I have fixed the IPv4 problem, and
6498 patched it up to do a reasonable job for IPv6.
6500 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6501 delivery, the log line now contains "pipelined" if PIPELINING was used.
6503 4. An SMTP transport process used to panic and die if the bind() call to set
6504 an explicit outgoing interface failed. This has been changed; it is now
6505 treated in the same way as a connect() failure.
6507 5. A reference to $sender_host_name in the part of a conditional expansion
6508 that was being skipped was still causing a DNS lookup. This no longer
6511 6. The def: expansion condition was not recognizing references to header lines
6512 that used bh_ and bheader_.
6514 7. Added the _cache feature to named lists.
6516 8. The code for checking quota_filecount in the appendfile transport was
6517 allowing one more file than it should have been.
6519 9. For compatibility with Sendmail, the command line option
6527 and sets the incoming protocol and host name (for trusted callers). The
6528 host name and its colon can be omitted when only the protocol is to be set.
6529 Note the Exim already has two private options, -pd and -ps, that refer to
6530 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6531 "s", but I don't think that's a major issue.
6533 10. A number of refactoring changes to the code, none of which should affect
6536 (a) The number of logging options was getting close to filling up the
6537 32-bit word that was used as a bit map. I have split them into two classes:
6538 those that are passed in the argument to log_write(), and those that are
6539 only ever tested independently outside of that function. These are now in
6540 separate 32-bit words, so there is plenty of room for expansion again.
6541 There is no change in the user interface or the logging behaviour.
6543 (b) When building, for example, log lines, the code previously used a
6544 macro that called string_cat() twice, in order to add two strings. This is
6545 not really sufficiently general. Furthermore, there was one instance where
6546 it was actually wrong because one of the argument was used twice, and in
6547 one call a function was used. (As it happened, calling the function twice
6548 did not affect the overall behaviour.) The macro has been replaced by a
6549 function that can join an arbitrary number of extra strings onto a growing
6552 (c) The code for expansion conditions now uses a table and a binary chop
6553 instead of a serial search (which was left over from when there were very
6554 few conditions). Also, it now recognizes conditions like "pam" even when
6555 the relevant support is not compiled in: a suitably worded error message is
6556 given if an attempt is made to use such a condition.
6558 11. Added ${time_interval:xxxxx}.
6560 12. A bug was causing one of the ddress fields not to be passed back correctly
6561 from remote delivery subprocesses. The field in question was not being
6562 subsequently used, so this caused to problems in practice.
6564 13. Added new log selectors queue_time and deliver_time.
6566 14. Might have fixed a bug in maildirsizefile handling that threw up
6567 "unexpected character" debug warnings, and recalculated the data
6568 unnecessarily. In any case, I expanded the warning message to give more
6571 15. Added the message "Restricted characters in address" to the statements in
6572 the default ACL that block characters like @ and % in local parts.
6574 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6575 Three changes have been made:
6577 (a) There was a serious bug; a negative response to MAIL caused the whole
6578 recipient domain to be cached as invalid, thereby blocking all messages
6579 to all local parts at the same domain, from all senders. This bug has
6580 been fixed. The domain is no longer cached after a negative response to
6581 MAIL if the sender used is not empty.
6583 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6586 (c) A new callout option, "use_sender" has been added for people who want
6587 the modified behaviour.
6593 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
6596 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
6597 indeed breaks things for older releases.
6599 3. Added additional logging to the case where there is a problem reading data
6600 from a filter that is running in a subprocess using a pipe, in order to
6601 try to track down a specific problem.
6603 4. Testing facility fudge: when running in the test harness and attempting
6604 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
6605 getting "No route to host". Convert this to a timeout.
6607 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
6610 6. Some OS don't have socklen_t but use size_t instead. This affects the
6611 fifth argument of getsockopt() amongst other things. This is now
6612 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
6613 can be set for individual OS. I have set it for SunOS5, OSF1, and
6614 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
6615 some earlier ones do not.
6617 7. Change 4.30/15 was not doing the test caselessly.
6619 8. The standard form for an IPv6 address literal was being rejected by address
6620 parsing in, for example, MAIL and RCPT commands. An example of this kind of
6621 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6622 this, as well as the form without the "IPv6" on the front (but only when
6623 address literals are enabled, of course).
6625 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
6627 10. Exim crashed if a message with an empty sender address specified by -f
6628 encountered a router with an errors_to setting. This could be provoked only
6629 by a command such as
6633 where an empty string was supplied; "<>" did not hit this bug.
6635 11. Installed PCRE release 4.5.
6637 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
6638 remained set. It is now erased.
6640 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
6641 times from message ids (which are base 36 rather than the normal 62).
6643 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
6644 were being counted as actual protocol errors, and logged if the log
6645 selector +smtp_protocol_error was set. One cannot be perfect in this test,
6646 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
6647 and DATA following a set of rejected RCPTs do not count as protocol errors.
6648 In other words, Exim assumes they were pipelined, though this may not
6649 actually be the case. Of course, in all cases the client gets an
6650 appropriate error code.
6652 15. If a lookup fails in an ACL condition, a message about the failure may
6653 be available; it is used if testing the ACL cannot continue, because most
6654 such messages specify what the cause of the deferral is. However, some
6655 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
6656 that caused an old message to be retained and used if a later statement
6657 caused a defer, replacing the real cause of the deferral.
6659 16. If an IP address had so many PTR records that the DNS lookup buffer
6660 was not large enough to hold them, Exim could crash while trying to process
6661 the truncated data. It now detects and logs this case.
6663 17. Further to 4.21/58, another change has been made: if (and only if) the
6664 first line of a message (the first header line) ends with CRLF, a bare LF
6665 in a subsequent header line has a space inserted after it, so as not to
6666 terminate the header.
6668 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
6669 unnecessarily, used atoi() instead of strtol(), and didn't check the
6670 termination when getting file sizes from file names by regex.
6672 19. Completely re-implemented the support for maildirsize files, in the light
6673 of a number of problems with the previous contributed implementation
6674 (4.30/29). In particular:
6676 . If the quota is zero, the maildirsize file is maintained, but no quota is
6679 . If the maildir directory does not exist, it is created before any attempt
6680 to write a maildirsize file.
6682 . The quota value in the file is just a cache; if the quota is changed in
6683 the transport, the new value overrides.
6685 . A regular expression is available for excluding directories from the
6688 20. The autoreply transport checks the characters in options that define the
6689 message's headers; it allows continued headers, but it was checking with
6690 isspace() after an embedded newline instead of explicitly looking for a
6693 21. If all the "regular" hosts to which an address was routed had passed their
6694 expiry times, and had not reached their retry times, the address was
6695 bounced, even if fallback hosts were defined. Now Exim should go on to try
6698 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
6699 equivalent code in the SMTP transport. Some hosts send humungous responses
6700 to HELO/EHLO, more than 1024 it seems.
6702 23. Refactoring: code in filter.c used (void *) for "any old type" but this
6703 gives compiler warnings in some environments. I've now done it "properly",
6706 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
6707 (because of problems with the built-in one) was declared to return uschar *
6708 instead of char *, causing compiler failure.
6710 25. Fixed a file descriptor leak when processing alias/forward files.
6712 26. Fixed a minor format string issue in dbfn.c.
6714 27. Typo in exim.c: ("dmbnz" for "dbmnz").
6716 28. If a filter file refered to $h_xxx or $message_headers, and the headers
6717 contained RFC 2047 "words", Exim's memory could, under certain conditions,
6720 29. When a sender address is verified, it is cached, to save repeating the test
6721 when there is more than one recipient in a message. However, when the
6722 verification involves a callout, it is possible for different callout
6723 options to be set for different recipients. It is too complicated to keep
6724 track of this in the cache, so now Exim always runs a verification when a
6725 callout is required, relying on the callout cache for the optimization.
6726 The overhead is duplication of the address routing, but this should not be
6729 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
6730 to be verified with callout=postmaster, and the main callout worked but the
6731 postmaster check failed, the verification correctly failed. However, if a
6732 subsequent RCPT command asked for sender verification *without* the
6733 postmaster check, incorrect caching caused this verification also to fail,
6736 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6737 it was not caching the DNS options (qualify_single, search_parents) that
6738 were used when the lookup failed. A subsequent lookup with different
6739 options therefore always gave the same answer, though there were cases
6740 where it should not have. (Example: a "domains = !$mx_any" option on a
6741 dnslookup router: the "domains" option is always processed without any
6742 widening, but the router might have qualify_single set.) Now Exim uses the
6743 cached value only when the same options are set.
6745 32. Added John Jetmore's "exipick" utility to the distribution.
6747 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6748 than a timeout (e.g. a certificate is required, and is not provided), an
6749 Exim server now closes the connection immediately. Previously it waited for
6750 the client to close - but if the client is SSL, it seems that they each
6751 wait for each other, leading to a delay before one of them times out.
6753 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6754 maintained 0.8.x compatibility because I don't think many are using it, and
6755 it is clearly obsolete.
6757 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6760 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6761 client certificate was expired. A simple patch fixes this, though I don't
6762 understand the full logic of why the verify callback is called multiple
6765 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6766 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6767 which causes problems with some clients (such as the Certicom SSL Plus
6768 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6769 disables the coutermeasure allowing Eudora to connect."
6771 38. Exim was not checking that a write() to a log file succeeded. This could
6772 lead to Bad Things if a log got too big, in particular if it hit a file
6773 size limit. Exim now panics and dies if it cannot write to a log file, just
6774 as it does if it cannot open a log file.
6776 39. Modified OS/Makefile-Linux so that it now contains
6778 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6780 The two -D definitions ensure that Exim is compiled with large file
6781 support, which makes it possible to handle log files that are bigger than
6784 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6785 instance) a domain was checked against a named list that involved a lookup,
6786 causing $domain_data to be set, then another domain was checked against the
6787 same list, then the first domain was re-checked, the value of $domain_data
6788 after the final check could be wrong. In particular, if the second check
6789 failed, it could be set empty. This bug probably also applied to
6792 41. The strip_trailing_dot option was not being applied to the address given
6793 with the -f command-line option.
6795 42. The code for reading a message's header from the spool was incrementing
6796 $received_count, but never initializing it. This meant that the value was
6797 incorrect (doubled) while delivering a message in the same process in which
6798 it was received. In the most common configuration of Exim, this never
6799 happens - a fresh exec is done - but it can happen when
6800 deliver_drop_privilege is set.
6802 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6803 it now includes up to 150 characters of the unexpected data in the log
6806 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6807 and building data strings. The size of both of these buffers was 10 000
6808 bytes - far larger than anybody would *ever* want, thought I. Needless to
6809 say, somebody hit the limit. I have increased the maximum line length to
6810 20 000 and the maximum data length of concatenated lines to 100 000. I have
6811 also fixed two bugs, because there was no checking on these buffers. Tsk,
6812 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6813 buffer is too small.
6815 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6816 lsearch lookups. Now it does.
6818 46. When parsing a route_list item in a manualroute router, a fixed-length
6819 buffer was used for the list of hosts. I made this 1024 bytes long,
6820 thinking that nobody would ever have a list of hosts that long. Wrong.
6821 Somebody had a whole pile of complicated expansion conditions, and the
6822 string was silently truncated, leading to an expansion error. It turns out
6823 that it is easier to change to an unlimited length (owing to other changes
6824 that have happened since this code was originally written) than to build
6825 structure for giving a limitation error. The length of the item that
6826 expands into the list of hosts is now unlimited.
6828 47. The lsearch lookup could not handle data where the length of text line was
6829 more than 4095 characters. Such lines were truncated, leading to shortened
6830 data being returned. It should now handle lines of any length.
6832 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6833 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6836 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6837 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6838 no longer happens in comment lines. A list of replacements is now placed
6839 at the head of all of the source files, except those whose only change is
6840 to replace PERL_COMMAND in the very first #! line.
6842 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6843 messages on the queue, with a bottom-up merge sort, using code contributed
6844 by Michael Haardt. This should make operations like -bp somewhat faster on
6845 large queues. It won't affect queue runners, except when queue_run_in_order
6848 51. Installed eximstats 1.31 in the distribution.
6850 52. Added support for SRV lookups to the dnslookup router.
6852 53. If an ACL referred to $message_body or $message_body_end, the value was not
6853 reset for any messages that followed in the same SMTP session.
6855 54. The store-handling optimization for building very long strings was not
6856 differentiating between the different store pools. I don't think this
6857 actually made any difference in practice, but I've tidied it.
6859 55. While running the routers to verify a sender address, $sender_address
6860 was still set to the sender address. This is wrong, because when routing to
6861 send a bounce to the sender, it would be empty. Therefore, I have changed
6862 it so that, while verifying a sender address, $sender_address is set to <>.
6863 (There is no change to what happens when verifying a recipient address.)
6865 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6866 target A or AAAA records (if not already returned) without resetting the
6867 qualify_single or search_parents options of the DNS resolver. These are
6868 inappropriate in this case because the targets of MX and SRV records must
6869 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6870 target that, when qualified, matched something in the local domain. These
6871 two options are now turned off when doing these lookups.
6873 57. It seems that at least some releases of Reiserfs (which does not have the
6874 concept of a fixed number of inodes) returns zero and not -1 for the
6875 number of available inodes. This interacted badly with check_spool_inodes,
6876 which assumed that -1 was the "no such thing" setting. What I have done is
6877 to check that the total number of inodes is greater than zero before doing
6878 the test of how many are available.
6880 58. When a "warn" ACL statement has a log_message modifier, the message is
6881 remembered, and not repeated. This is to avoid a lot of repetition when a
6882 message has many recipients that cause the same warning to be written.
6883 However, Exim was preserving the list of already written lines for an
6884 entire SMTP session, which doesn't seem right. The memory is now reset if a
6885 new message is started.
6887 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6888 debugging output unless log_rewrite was also set.
6890 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6891 of (int)(handle) when we know that handle contains (void *)(-1).
6893 61. The Exim daemon panic-logs an error return when it closes the incoming
6894 connection. However "connection reset by peer" seems to be common, and
6895 isn't really an error worthy of noting specially, so that particular error
6898 62. When Exim is trying to find all the local interfaces, it used to panic and
6899 die if the ioctl to get the interface flags failed. However, it seems that
6900 on at least one OS (Solaris 9) it is possible to have an interface that is
6901 included in the list of interfaces, but for which you get a failure error
6902 for this call. This happens when the interface is not "plumbed" into a
6903 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6904 failure of the "get flags" call assumes that the interface is down.
6906 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6907 makes life easier for people who are doing arithmetic on fields extracted
6908 from dates, where you often get leading zeros that should not be
6909 interpreted as octal.
6911 64. Added qualify_domain to the redirect router, to override the global
6914 65. If a pathologically long header line contained very many addresses (the
6915 report of this problem mentioned 10 000) and each of them was rewritten,
6916 Exim could use up a very large amount of memory. (It kept on making new
6917 copies of the header line as it rewrote, and never released the old ones.)
6918 At the expense of a bit more processing, the header rewriting function has
6919 been changed so that it no longer eats memory in this way.
6921 66. The generation of the Received: header has been moved from the time that a
6922 message starts to be received, to the time that it finishes. The timestamp
6923 in the Received: header should now be very close to that of the <= log
6924 line. There are two side-effects of this change:
6926 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6927 logged header lines no longer include the local Received: line, because
6928 it has not yet been created. The same applies to a copy of the message
6929 that is returned to a non-SMTP sender when a message is rejected.
6931 (b) When a filter file is tested using -bf, no additional Received: header
6932 is added to the test message. After some thought, I decided that this
6935 This change does not affect the value of $received_for. It is still set
6936 after address rewriting, but before local_scan() is called.
6938 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6940 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6941 gave an unhelpful panic error message, and a defer error. I have managed to
6942 change this behaviour so that it now rejects any supplied certificate,
6943 which seems right, as the list of acceptable certificates is empty.
6945 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6946 gave an unhelpful defer error. I have not managed to make this reject any
6947 supplied certificates, but the error message it gives is "no certificate
6948 supplied", which is not helpful.
6950 70. exigrep's output now also includes lines that are not associated with any
6951 message, but which match the given pattern. Implemented by a patch from
6952 Martin Sluka, which also tidied up the Perl a bit.
6954 71. Recipient callout verification, like sender verification, was using <> in
6955 the MAIL FROM command. This isn't really the right thing, since the actual
6956 sender may affect whether the remote host accepts the recipient or not. I
6957 have changed it to use the actual sender in the callout; this means that
6958 the cache record is now keyed on a recipient/sender pair, not just the
6959 recipient address. There doesn't seem to be a real danger of callout loops,
6960 since a callout by the remote host to check the sender would use <>.
6961 [SEE ABOVE: changed after hitting problems.]
6963 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6964 temporary errors. However, in the case of such a code being given after
6965 the end of a data transmission (i.e. after ".") Exim was failing to write
6966 a retry record for the message. (Yes, there was some broken host that was
6967 actually sending 8xx at this point.)
6969 73. An unknown lookup type in a host list could cause Exim to panic-die when
6970 the list was checked. (An example that provoked this was putting <; in the
6971 middle of a list instead of at the start.) If this happened during a DATA
6972 ACL check, a -D file could be left lying around. This kind of configuration
6973 error no longer causes Exim to die; instead it causes a defer error. The
6974 incident is still logged to the main and panic logs.
6976 74. Buglet left over from Exim 3 conversion. The message "too many messages
6977 in one connection" was written to the rejectlog but not the mainlog, except
6978 when address rewriting (yes!) was being logged.
6980 75. Added write_rejectlog option.
6982 76. When a system filter was run not as root (that is, when system_filter_user
6983 was set), the values of the $n variables were not being returned to the
6984 main process; thus, they were not subsequently available in the $sn
6987 77. Added +return_path_on_delivery log selector.
6989 78. A connection timeout was being treated differently from recipients deferred
6990 when testing hosts_max_try with a message that was older than the host's
6991 retry timeout. (The host should not be counted, thus allowing all hosts to
6992 be tried at least once before bouncing.) This may have been the cause of an
6993 occasionally reported bug whereby a message would remain on the queue
6994 longer than the retry timeout, but would be bounced if a delivery was
6995 forced. I say "may" because I never totally pinned down the problem;
6996 setting up timeout/retry tests is difficult. See also the next item.
6998 79. The ultimate address timeout was not being applied to errors that involved
6999 a combination of host plus message (for example, a timeout on a MAIL
7000 command). When an address resolved to a number of possible hosts, and they
7001 were not all tried for each delivery (e.g. because of hosts_max_try), a
7002 message could remain on the queue longer than the retry timeout.
7004 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7007 81. Fixed an obscure SMTP outgoing bug which required at least the following
7008 conditions: (a) there was another message waiting for the same server;
7009 (b) the server returned 5xx to all RCPT commands in the first message so
7010 that the message was not completed; (c) the server dropped the connection
7011 or gave a negative response to the RSET that Exim sends to abort the
7012 transaction. The observed case was a dropped connection after DATA that had
7013 been sent in pipelining mode. That is, the server had advertised PIPELINING
7014 but was not implementing it correctly. The effect of the bug was incorrect
7015 behaviour, such as trying another host, and this could lead to a crash.
7021 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7022 and daemon.c were passed as pointers to ints; they should have been
7023 pointers to socklen_t variables (which are typically unsigned ints).
7025 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7028 3. Fixed a really odd bug that affected only the testing scheme; patching a
7029 certain fixed string in the binary changed the value of another string that
7030 happened to be identical to the end of the original first string.
7032 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7033 name", it returns that address as the IP address. On some operating
7034 systems (e.g. Solaris), it also passes back the IP address string as the
7035 "host name". However, on others (e.g. Linux), it passes back an empty
7036 string. Exim wasn't checking for this, and was changing the host name to an
7037 empty string, assuming it had been canonicalized.
7039 5. Although rare, it is permitted to have more than one PTR record for a given
7040 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7041 all the names associated with an address, because they do in Solaris.
7042 However, it seems that they do not in Linux for data that comes from the
7043 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7044 I found this out when I moved to a new Linux workstation and tried to run
7045 the Exim test suite.
7047 To get round this problem I have changed the code so that it now does its
7048 own call to the DNS to look up PTR records when searching for a host name.
7049 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7050 addresses that are only in /etc/hosts are still found.
7052 This behaviour is, however, controlled by an option called host_lookup_
7053 order, which defaults to "bydns:byaddr". If people want to use the other
7054 order, or indeed, just use one or the other means of lookup, they can
7055 specify it in this variable.
7057 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7058 some operating systems, this comes back from gethostbyaddr() as an empty
7059 string, and this is what Exim used to test for. However, it seems that in
7060 other systems, "." is yielded. Exim now tests for this case too.
7062 7. The values of check_spool_space and check_log_space are now held internally
7063 as a number of kilobytes instead of an absolute number of bytes. If a
7064 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7065 kilobyte. This means that much larger values can be stored.
7067 8. Exim monitor: an attempt to get the action menu when not actually pointing
7068 at a message produces an empty menu entitled "No message selected". This
7069 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7070 no entries in it ("Shell widget menu has zero width and/or height"). So I
7071 have added a single, blank menu entry in this case.
7073 9. Added ${quote_local_part.
7075 10. MIME decoding is now applied to the contents of Subject: header lines when
7078 11. Now that a reference to $sender_host_address automatically causes a reverse
7079 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7080 host lookup before query-style lookups in lists that might use this
7081 variable. This has therefore been abolished, and the "net-" prefix is no
7082 longer necessary for query-style lookups.
7084 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7085 have been a typo for LFLAGS, so it has been changed.
7087 13. The install script calls Exim with "-C /dev/null" in order to find the
7088 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7089 to be output. However, since Exim outputs its version number before the
7090 error, it didn't break the script. It just looked ugly. I fixed this by
7091 always allowing "-C /dev/null" if the caller is root.
7093 14. Ignore overlarge ACL variable number when reading spool file - insurance
7094 against a later release with more variables having written the file.
7096 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7097 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7098 this, as well as the form without the "IPv6" on the front.
7100 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7101 OS/Makefile-Darwin file.
7103 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7104 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7106 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7107 variable. However, some OS use ints for inodes. I've added cast to long int
7108 to get rid of the compiler warning.
7110 19. I had forgotten to lock out "/../" in configuration file names when
7111 ALT_CONFIG_PREFIX was set.
7113 20. Routers used for verification do not need to specify transports. However,
7114 if such a router generated a host list, and callout was configured, Exim
7115 crashed, because it could not find a port number from the (non-existent)
7116 transport. It now assumes port 25 in this circumstance.
7118 21. Added the -t option to exigrep.
7120 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7121 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7122 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7123 them is set, LOOKUP_LSEARCH is forced.
7125 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7127 24. Added sender and host information to the "rejected by local_scan()" log
7128 line; previously there was no indication of these.
7130 25. Added .include_if_exists.
7132 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7133 Linux. It turns out that not all file systems support this. Apparently some
7134 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7135 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7136 not supported on the file descriptor, is now ignored when Exim is trying to
7137 sync a directory. This applies only to Linux.
7139 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7141 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7142 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7145 29. Added support for maildirsize files from supplied patch (modified a bit).
7147 30. The use of :fail: followed by an empty string could lead Exim to respond to
7148 sender verification failures with (e.g.):
7150 550 Verification failed for <xxx>
7151 550 Sender verify failed
7153 where the first response line was missing the '-' that indicates it is not
7154 the final line of the response.
7156 31. The loop for finding the name of the user that called Exim had a hardwired
7157 limit of 10; it now uses the value of finduser_retries, which is used for
7158 all other user lookups.
7160 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7163 33. Exim was neglecting to zero errno before one call of strtol() when
7164 expanding a string and expecting an integer value. On some systems this
7165 resulted in spurious "integer overflow" errors. Also, it was casting the
7166 result into an int without checking.
7168 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7169 did not work. The code looks as if it has *never* worked, though it appears
7170 to have been documented since at least release 1.62. I have made it work.
7172 35. The "timeout_DNS" error in retry rules, also documented since at least
7173 1.62, also never worked. As it isn't clear exactly what this means, and
7174 clearly it isn't a major issue, I have abolished the feature by treating it
7175 as "timeout", and writing a warning to the main and panic logs.
7177 36. The display of retry rules for -brt wasn't always showing the error code
7180 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7181 timeout_connect_A, timeout_connect_MX.
7183 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7184 to the empty sender.
7186 39. The daemon was not analysing the content of -oX till after it had closed
7187 stderr and disconnected from the controlling terminal. This meant that any
7188 syntax errors were only noted on the panic log, and the return code from
7189 the command was 0. By re-arranging the code a little, I've made the
7190 decoding happen first, so such errors now appear on stderr, and the return
7191 code is 1. However, the actual setting up of the sockets still happens in
7192 the disconnected process, so errors there are still only recorded on the
7195 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7196 connections (as happens on some IP stacks) was logged at start up time as
7197 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7198 it from "IPv6 and IPv4", which means that two separate sockets are being
7201 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7202 says whether AF_INET or AF_INET6 was passed as an argument.
7204 42. Exiwhat output was messed up when time zones were included in log
7207 43. Exiwhat now gives more information about the daemon's listening ports,
7208 and whether -tls-on-connect was used.
7210 44. The "port" option of the smtp transport is now expanded.
7212 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7213 silently ignored. Now an error message is written to the main and panic
7216 46. There's a new ACL modifier called "logwrite" which writes to a log file
7217 as soon as it is encountered.
7219 47. Added $local_user_uid and $local_user_gid at routing time.
7221 48. Exim crashed when trying to verify a sender address that was being
7224 49. Exim was recognizing only a space character after ".include". It now also
7225 recognizes a tab character.
7227 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7228 extracting the relevant information from the specification. The man page no
7229 longer contains scrambled data for the -d option, and I've added a section
7230 at the front about calling Exim under different names.
7232 51. Added "extra_headers" argument to the "mail" command in filter files.
7234 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7237 53. Installed eximstats 1.29.
7239 54. Added transport_filter_timeout as a generic transport option.
7241 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7242 Cc: header lines. This was required by RFC 822, but it not required by RFC
7245 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7246 incoming messages that did not have them. Now it does so only if the
7247 message originates locally, that is, if there is no associated remote host
7248 address. When Resent- header lines are present, this applies to the Resent-
7249 lines rather than the non-Resent- lines.
7251 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7252 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7254 58. Messages for configuration errors now include the name of the main
7255 configuration file - useful now that there may be more than one file in a
7256 list (.included file names were always shown).
7258 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7259 for those rare installations that do not start the daemon as root or run it
7260 setuid root. I've cut out the call to initgroups() if the daemon is not
7263 60. The Exim user and group can now be bound into the binary as text strings
7264 that are looked up at the start of Exim's processing.
7266 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7268 62. Added $mailstore_basename variable.
7270 63. Installed patch to sieve.c from Michael Haardt.
7272 64. When Exim failed to open the panic log after failing to open the main log,
7273 the original message it was trying to log was written to stderr and debug
7274 output, but if they were not available (the usual case in production), it
7275 was lost. Now it is written to syslog before the two lines that record the
7276 failures to open the logs.
7278 65. Users' Exim filters run in subprocesses under the user's uid. It is
7279 possible for a "deliver" command or an alias in a "personal" command to
7280 provoke an address rewrite. If logging of address rewriting is configured,
7281 this fails because the process is not running as root or exim. There may be
7282 a better way of dealing with this, but for the moment (because 4.30 needs
7283 to be released), I have disabled address rewrite logging when running a
7284 filter in a non-root, non-exim process.
7290 1. The buildconfig auxiliary program wasn't quoting the value set for
7291 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7292 not defined. This bug was masked in 4.22 by the effect that was fixed in
7295 2. Some messages that were rejected after a message id was allocated were
7296 shown as "incomplete" by exigrep. It no longer does this for messages that
7297 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7299 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7300 have allow_domain_literals set, the ID did not get logged in the <= line.
7301 Domain literals are now always recognized in Message-ID: header lines.
7303 4. The first argument for a ${extract expansion item is the key name or field
7304 number. Leading and trailing spaces in this item were not being ignored,
7305 causing some misleading effects.
7307 5. When deliver_drop_privilege was set, single queue runner processes started
7308 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7309 same command in the process it spins off) were not dropping privilege.
7311 6. When the daemon running as "exim" started a queue runner, it always
7312 re-executed Exim in the spun-off process. This is a waste of effort when
7313 deliver_drop_privilege is set. The new process now just calls the
7314 queue-runner function directly.
7320 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7323 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7324 ignored. Though the use of -oP was forcing the writing of a pid file, it
7325 was always written to the default place.
7327 3. If the message "no IP address found for host xxxx" is generated during
7328 incoming verification, it is now followed by identification of the incoming
7329 connection (so you can more easily find what provoked it).
7331 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7333 5. Added some features to "harden" Exim a bit more against certain attacks:
7335 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7336 be put in Local/Makefile. This is like the never_users runtime option,
7337 but it cannot be overridden. The default setting is "root".
7339 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7340 prefix string with which any file named in a -C command line option
7343 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7344 is retained for -C and -D only if the caller of Exim is root. Without
7345 it, the exim user may also use -C and -D and retain privilege.
7347 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7348 command line option is disabled.
7350 6. Macro names set by the -D option must start with an upper case letter, just
7351 like macro names defined in the configuration file.
7353 7. Added "dereference=" facility to LDAP.
7355 8. Two instances of the typo "uknown" in the source files are fixed.
7357 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7358 the Configure-Makefile script screwed up while processing it.
7360 10. Incorporated PCRE 4.4.
7362 11. The SMTP synchronization check was not operating right at the start of an
7363 SMTP session. For example, it could not catch a HELO sent before the client
7364 waited for the greeting. There is now a check for outstanding input at the
7365 point when the greeting is written. Because of the duplex, asynchronous
7366 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7367 way, but not yet received, when the check is performed.
7369 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7370 on TCP/IP sockets, because this apparently causes some broken clients to
7373 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7374 unchanged) from the Cygwin maintainer.
7376 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7377 was supported instead of testing for mbx. Effectively a typo.
7379 15. The spa authenticator server code was not checking that the input it
7380 received was valid base64.
7382 16. The debug output line for the "set" modifier in ACLs was not showing the
7383 name of the variable that was being set.
7385 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7387 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7388 lookup on its own. Something else was needed to trigger the lookup. For
7389 example, a match in host_lookup or the need for a host name in a host list.
7390 Now, if $sender_host_name is referenced and the host name has not yet been
7391 looked up, a lookup is performed. If the lookup fails, the variable remains
7392 empty, and $host_lookup_failed is set to "1".
7394 19. Added "eqi" as a case-independent comparison operator.
7396 20. The saslauthd authentication condition could segfault if neither service
7397 nor realm was specified.
7399 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7400 error message that was logged was misleading, and incoming SMTP
7401 connections were dropped. The message is now more accurate, and temporary
7402 errors are given to SMTP connections.
7404 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7405 (except RSET) with a 421 error, until QUIT is received. However, it was
7406 failing to send a response to QUIT.
7408 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7409 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7410 ACL is now run after the helo_try_verify_hosts code.
7412 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7413 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7414 case-independent, so other case variants are also recognized). Apparently
7415 some systems use these upper case variants.
7417 25. If more than two messages were waiting for the same host, and a transport
7418 filter was specified for the transport, Exim sent two messages over the
7419 same TCP/IP connection, and then failed with "socket operation on non-
7420 socket" when it tried to send the third.
7422 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7424 27. The extern definition of crypt16() in expand.c was not being excluded when
7425 the OS had its own crypt16() function.
7427 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7428 as a preferred synonym for return_size_limit, both as an option and as an
7431 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7433 30. Changed the default configuration ACL to relax the local part checking rule
7434 for addresses that are not in any local domains. For these addresses,
7435 slashes and pipe symbols are allowed within local parts, but the sequence
7436 /../ is explicitly forbidden.
7438 31. SPA server authentication was not clearing the challenge buffer before
7441 32. log_message in a "warn" ACL statement was writing to the reject log as
7442 well as to the main log, which contradicts the documentation and doesn't
7443 seem right (because no rejection is happening). So I have stopped it.
7445 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7446 However, I am unable to do any testing of this.
7448 34. Fixed an infelicity in the appendfile transport. When checking directories
7449 for a mailbox, to see if any needed to be created, it was accidentally
7450 using path names with one or more superfluous leading slashes; tracing
7451 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7453 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7454 contents are added to the log line that is written for every discarded
7455 recipient. (Previously a log_message setting was ignored.)
7457 36. The ${quote: operator now quotes the string if it is empty.
7459 37. The install script runs exim in order to find its version number. If for
7460 some reason other than non-existence or emptiness, which it checks, it
7461 could not run './exim', it was installing it with an empty version number,
7462 i.e. as "exim-". This error state is now caught, and the installation is
7465 38. An argument was missing from the function that creates an error message
7466 when Exim fails to connect to the socket for saslauthd authentication.
7467 This could cause Exim to crash, or give a corrupted message.
7469 39. Added isip, isip4, and isip6 to ${if conditions.
7471 40. The ACL variables $acl_xx are now saved with the message, and can be
7472 accessed later in routers, transports, and filters.
7474 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7475 strings in the file are not string-expanded.
7477 42. If a MAIL command specified a SIZE value that was too large to fit into an
7478 int variable, the check against message_size_limit failed. Such values are
7479 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7480 day this will have to be increased, but I don't think I want to be around
7481 when emails are that large.
7488 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7489 iconv() is not standard in FreeBSD.
7491 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7492 IPv6 enabled. The observed symptom was a segmentation fault on return from
7493 the function os_common_find_running_interfaces() in src/os.c.
7495 3. In the check_special_case() function in daemon.c I had used "errno" as an
7496 argument name, which causes warnings on some systems. This was basically a
7497 typo, since it was named "eno" in the comments!
7499 4. The code that waits for the clock to tick (at a resolution of some fraction
7500 of a second) so as to ensure message-id uniqueness was always waiting for
7501 at least one whole tick, when it could have waited for less. [This is
7502 almost certainly not relevant at current processor speeds, where it is
7503 unlikely to ever wait at all. But we try to future-proof.]
7505 5. The function that sleeps for a time interval that includes fractions of a
7506 second contained a race. It did not block SIGALRM between setting the
7507 timer, and suspending (a couple of lines later). If the interval was short
7508 and the sigsuspend() was delayed until after it had expired, the suspension
7509 never ended. On busy systems this could lead to processes getting stuck for
7512 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7513 process, before it forks any delivery processes. The open lookup caching
7514 mechanism meant that the open file or database connection was passed into
7515 the delivery process. The problem was that delivery processes always tidy
7516 up cached lookup data. This could cause a problem for the next delivery
7517 process started by the queue runner, because the external queue runner
7518 process does not know about the closure. So the next delivery process
7519 still has data in the lookup cache. In the case of a file lookup, there was
7520 no problem because closing a file descriptor in a subprocess doesn't affect
7521 the parent. However, if the lookup was caching a connection to a database,
7522 the connection was closed, and the second delivery process was likely to
7523 see errors such as "PGSQL: query failed: server closed the connection
7524 unexpectedly". The problem has been fixed by closing all cached lookups
7525 in a queue runner before running a delivery process.
7527 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7528 seem to have the "const" qualifier which it has on other OS. I've
7531 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7532 *of the same type* (client or server) with the same public name that an
7533 error should be diagnosed.
7535 9. When Exim looked up a host name for an IP address, but failed to find the
7536 original IP address when looking up the host name (a safety check), it
7537 output the message "<ip address> does not match any IP for NULL", which was
7538 confusing, to say the least. The bug was that the host name should have
7539 appeared instead of "NULL".
7541 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7542 user that is built into the binary, and the -C or -D options is used, root
7543 privilege is dropped before the configuration file is read. In addition,
7544 logging is switched to stderr instead of the normal log files. If the
7545 configuration then re-defines the Exim user, the unprivileged environment
7546 is probably not what is expected, so Exim logs a panic warning message (but
7549 However, if deliver_drop_privilege is set, the unprivileged state may well
7550 be exactly what is intended, so the warning has been cut out in that case,
7551 and Exim is allowed to try to write to its normal log files.
7557 1. smtp_return_error_details was not giving details for temporary sender
7558 or receiver verification errors.
7560 2. Diagnose a configuration error if two authenticators have the same public
7563 3. Exim used not to create the message log file for a message until the first
7564 delivery attempt. This could be confusing when incoming messages were held
7565 for policy or load reasons. The message log file is now created at the time
7566 the message is received, and an initial "Received" line is written to it.
7568 4. The automatically generated man page for command line options had a minor
7569 bug that caused no ill effects; however, a more serious problem was that
7570 the procedure for building the man page automatically didn't always
7571 operate. Consequently, release 4.20 contains an out-of-date version. This
7572 shouldn't happen again.
7574 5. When building Exim with embedded Perl support, the script that builds the
7575 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7576 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7578 6. The freeze_tell option was not being used for messages that were frozen on
7579 arrival, either by an ACL or by local_scan().
7581 7. Added the smtp_incomplete_transaction log selector.
7583 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7584 was accepting AUTH without a new EHLO.
7586 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7587 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7590 10. Logging of TCP/IP connections (when configured) now happens in the main
7591 daemon process instead of the child process, so that the TCP/IP connection
7592 count is more accurate (but it can never be perfect).
7594 11. The use of "drop" in a nested ACL was not being handled correctly in the
7595 outer ACL. Now, if condition failure induced by the nested "drop" causes
7596 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
7597 or "require"), the connection is dropped.
7599 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
7600 that yield "discard" can now be used with an "accept" or a "discard" verb,
7601 but an error is generated for any others (because I can't see a useful way
7602 to define what should happen).
7604 13. When an ACL is read dynamically from a file (or anywhere else), the lines
7605 are now processed in the same way as lines in the Exim configuration file.
7606 In particular, continuation lines are supported.
7608 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
7610 15. Added -ti meaning -t -i.
7612 16. Check for letters, digits, hyphens, and dots in the names of dnslist
7613 domains, and warn by logging if others are found.
7615 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
7616 returned from GIFCONF when Exim is trying to find the list of interfaces on
7617 a host. The code in os.c has been modified to copy each ifreq to an aligned
7618 structure in all cases.
7620 Also, in some cases, the returned ifreq's were being copied to a 'struct
7621 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
7622 means the last couple of bytes of an IPv6 address could be chopped if the
7623 ifreq contained only a normal sockaddr (14 bytes storage).
7625 18. Named domain lists were not supported in the hosts_treat_as_local option.
7626 An entry such as +xxxx was not recognized, and was treated as a literal
7629 19. Ensure that header lines added by a DATA ACL are included in the reject log
7630 if the ACL subsequently rejects the message.
7632 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
7633 MD5 (which is deprecated).
7635 21. When testing a filter file using -bf, Exim was writing a message when it
7636 took the sender from a "From " line in the message, but it was not doing so
7637 when it took $return_path from a Return-Path: header line. It now does.
7639 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
7640 with a valid header line field name (a series of printing characters
7641 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
7643 23. Changed "disc" in the source to "disk" to conform to the documentation and
7644 the book and for uniformity.
7646 24. Ignore Sendmail's -Ooption=value command line item.
7648 25. When execve() failed while trying to run a command in a pipe transport,
7649 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
7650 could be confused with a return value of 69 from the command itself. This
7651 has been changed to 127, the value the shell returns if it is asked to run
7652 a non-existent command. The wording for the related log line suggests a
7653 non-existent command as the problem.
7655 26. If received_header_text expands to an empty string, do not add a Received:
7656 header line to the message. (Well, it adds a token one on the spool, but
7657 marks it "old" so that it doesn't get used or transmitted.)
7659 27. Installed eximstats 1.28 (addition of -nt option).
7661 28. There was no check for failure on the call to getsockname() in the daemon
7662 code. This can fail if there is a shortage of resources on the system, with
7663 ENOMEM, for example. A temporary error is now given on failure.
7665 29. Contrary to the C standard, it seems that in some environments, the
7666 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
7667 program. Exim now does this explicitly; it affects the formatting of
7668 timestamps using strftime().
7670 30. If exiqsumm was given junk data, it threw up some uninitialized variable
7671 complaints. I've now initialized all the variables, to avoid this.
7673 32. Header lines added by a system filter were not being "seen" during
7674 transport-time rewrites.
7676 33. The info_callback() function passed to OpenSSL is set up with type void
7677 (*)(SSL *, int, int), as described somewhere. However, when calling the
7678 function (actually a macro) that sets it up, the type void(*)() is
7679 expected. I've put in a cast to prevent warnings from picky compilers.
7681 34. If a DNS black list lookup found a CNAME record, but there were no A
7682 records associated with the domain it pointed at, Exim crashed.
7684 35. If a DNS black list lookup returned more than one A record, Exim ignored
7685 all but the first. It now scans all returned addresses if a particular IP
7686 value is being sought. In this situation, the contents of the
7687 $dnslist_value variable are a list of all the addresses, separated by a
7690 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
7691 to accept a host name and all its aliases if the forward lookup for any of
7692 them yielded the IP address of the incoming connection. Now it accepts only
7693 those names whose forward lookup yields the correct IP address. Any other
7694 names are discarded. This closes a loophole whereby a rogue DNS
7695 administrator could create reverse DNS records to break through a
7696 wildcarded host restriction in an ACL.
7698 37. If a user filter or a system filter that ran in a subprocess used any of
7699 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
7700 the wrong values were passed to the pipe command ($thisaddress had the
7701 value of $0, $0 had the value of $1, etc). This bug was introduced by
7702 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
7704 38. Improved the line breaking for long SMTP error messages from ACLs.
7705 Previously, if there was no break point between 40 and 75 characters, Exim
7706 left the rest of the message alone. Two changes have been made: (a) I've
7707 reduced the minimum length to 35 characters; (b) if it can't find a break
7708 point between 35 and 75 characters, it looks ahead and uses the first one
7709 that it finds. This may give the occasional overlong line, but at least the
7710 remaining text gets split now.
7712 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
7713 file descriptors might be low, and that setting 1000 would always raise it.
7714 It turns out that in some environments, the limit is already over 1000 and
7715 that lowering it causes trouble. So now Exim takes care not to decrease it.
7717 40. When delivering a message, the value of $return_path is set to $sender_
7718 address at the start of routing (routers may change the value). By an
7719 oversight, this default was not being set up when an address was tested by
7720 -bt or -bv, which affected the outcome if any router or filter referred to
7723 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
7724 log only when "message" or "log_message" are set. However, if one of the
7725 conditions was an address verification, or a call to a nested ACL, the
7726 messages generated by the underlying test were being passed through. This
7727 no longer happens. The underlying message is available in $acl_verify_
7728 message for both "message" and "log_message" expansions, so it can be
7729 passed through if needed.
7731 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
7732 new expansion $bh_ to give the encoded byte string without charset
7733 translation. Translation happens only if iconv() is available; HAVE_ICONV
7734 indicates this at build time. HEADERS_CHARSET gives the charset to
7735 translate to; headers_charset can change it in the configuration, and
7736 "headers charset" can change it in an individual filter file.
7738 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7739 that creates RFC 2047 encoded "words" labels them as that charset instead
7740 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7741 expansion operator; (ii) when Exim creates a From: line for a local
7742 message; (iii) when a header line is rewritten to include a "phrase" part.
7744 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7745 buggy, causing it to skip the first lines of messages whose message ID
7746 ended in 'D'. This would not have bitten before Exim release 4.14, because
7747 message IDs were unlikely to end in 'D' before then. The effect was to have
7748 incorrect size information for certain domains.
7750 45. #include "config.h" was missing at the start of the crypt16.c module. This
7751 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7754 46. If there was a timeout during a "random" callout check, Exim treated it as
7755 a failure of the random address, and carried on sending RSET and the real
7756 address. If the delay was just some slowness somewhere, the response to the
7757 original RCPT would be taken as a response to RSET and so on, causing
7758 mayhem of various kinds.
7760 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7761 when I implemented it. It didn't allow for the fact that some option values
7762 may legitimately be negative (e.g. size_addition), and it didn't even do
7763 the right test for positive values.
7765 48. Domain names in DNS records are case-independent. Exim always looks them up
7766 in lower case. Some resolvers return domain names in exactly the case they
7767 appear in the zone file, that is, they may contain uppercase letters. Not
7768 all resolvers do this - some return always lower case. Exim was treating a
7769 change of case by a resolver as a change of domain, similar to a widening
7770 of a domain abbreviation. This triggered its re-routing code and so it was
7771 trying to route what was effectively the same domain again. This normally
7772 caused routing to fail (because the router wouldn't handle the domain
7773 twice). Now Exim checks for this case specially, and just changes the
7774 casing of the domain that it ultimately uses when it transmits the message
7777 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7780 50. If a filter generated a file delivery with a non-absolute name (possible if
7781 no home directory exists for the router), the forbid_file option was not
7784 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7785 the existing equality matching.
7787 52. Exim was using ints instead of ino_t variables in some places where it was
7788 dealing with inode numbers.
7790 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7791 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7792 called TMPDIR, and if it finds it is different, it changes its value.
7794 54. The smtp_printf() function is now made available to local_scan() so
7795 additional output lines can be written before returning. There is also an
7796 smtp_fflush() function to enable the detection of a dropped connection.
7797 The variables smtp_input and smtp_batched_input are exported to
7800 55. Changed the default runtime configuration: the message "Unknown user"
7801 has been removed from the ACL, and instead placed on the localuser router,
7802 using the cannot_route_message feature. This means that any verification
7803 failures that generate their own messages won't get overridden. Similarly,
7804 the "Unrouteable address" message that was in the ACL for unverifiable
7805 relay addresses has also been removed.
7807 56. Added hosts_avoid_esmtp to the smtp transport.
7809 57. The exicyclog script was not checking for the esoteric option
7810 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7811 will work only if exicyclog is run under the appropriate euid.
7813 58. Following a discussion on the list, the rules by which Exim recognises line
7814 endings on incoming messages have been changed. The -dropcr and drop_cr
7815 options are now no-ops, retained only for backwards compatibility. The
7816 following line terminators are recognized: LF CRLF CR. However, special
7817 processing applies to CR:
7819 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7820 nor a local message in the state where . is a terminator.
7822 (ii) If a bare CR is encountered in a header line, an extra space is added
7823 after the line terminator so as not to end the header. The reasoning
7824 behind this is that bare CRs in header lines are most likely either
7825 to be mistakes, or people trying to play silly games.
7827 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7828 was being incorrectly given as 18 bytes larger than it should have been.
7829 This is a VOB (very old bug).
7831 60. This may never have affected anything current, but just in case it has:
7832 When the local host is found other than at the start of a list of hosts,
7833 the local host, those with the same MX, and any that follow, are discarded.
7834 When the list in question was part of a longer list of hosts, the following
7835 hosts (not currently being processed) were also being discarded. This no
7836 longer happens. I'm not sure if this situation could ever has previously
7839 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7842 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7843 header. This is something that is recommended in a new Internet Draft, and
7844 is something that is documented as being done by Sendmail. There are two
7845 possible values. For messages generated by the autoreply transport, Exim
7848 Auto-Submitted: auto-replied
7850 whereas for all other generated messages (e.g. bounces) it adds
7852 Auto-Submitted: auto-generated
7854 63. The "personal" condition in filters now includes a test for the
7855 Auto-Submitted: header. If it contains the string "auto-" the message it
7856 not considered personal.
7858 64. Added rcpt_include_affixes as a generic transport option.
7860 65. Added queue_only_override (default true).
7862 66. Added the syslog_duplication option.
7864 67. If what should have been the first header line of a message consisted of
7865 a space followed by a colon, Exim was mis-interpreting it as a header line.
7866 It isn't of course - it is syntactically invalid and should therefore be
7867 treated as the start of the message body. The misbehaviour could have
7868 caused a number of strange effects, including loss of data in subsequent
7869 header lines, and spool format errors.
7871 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7872 client host had authenticated. This control can now be exercised by an ACL
7873 for more flexibility.
7875 69. By default, callouts do not happen when testing with -bh. There is now a
7876 variant, -bhc, which does actually run the callout code, including
7877 consulting and updating the callout cache.
7879 70. Added support for saslauthd authentication, courtesy of Alexander
7882 71. If statvfs() failed on the spool or log directories while checking their
7883 size for availability, Exim confusingly gave the error "space shortage".
7884 Furthermore, in debugging mode it crashed with a floating point exception.
7885 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7886 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7887 serious problem, Exim now writes to the main and panic logs when this
7888 happens, with details of the failure. It then refuses to accept the
7889 incoming message, giving the message "spool directory problem" or "log
7890 directory problem" with a 421 code for SMTP messages.
7892 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7893 0, 1, and 2 exist, because some OS complain for execs without them (see
7894 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7895 descriptors. However, the code omitted to check that the open succeeded,
7896 causing mysterious errors if for some reason the permissions on /dev/null
7897 got screwed. Now Exim writes a message to the main and panic logs, and
7898 bombs out if it can't open /dev/null.
7900 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7901 interact so that it is all more flexible. It is supposed to remain
7902 backwards compatible. Also added extra_local_interfaces.
7904 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7905 to bomb out with an assertion failure - to the client this appears as a
7906 connection drop. This problem occurs in the part of the code that was taken
7907 from the Samba project. Fortunately, the assertion is in a very simple
7908 function, so I have fixed this by reproducing the function inline in the
7909 one place where it is called, and arranging for authentication to fail
7910 instead of killing the process with assert().
7912 75. The SPA client code was not working when the server requested OEM rather
7913 than Unicode encoding.
7915 76. Added code to make require_files with a specific uid setting more usable in
7916 the case where statting the file as root fails - usually a non-root-mounted
7917 NFS file system. When this happens and the failure is EACCES, Exim now
7918 forks a subprocess and does the per-uid checking as the relevant uid.
7920 77. Added process_log_path.
7922 78. If log_file_path was not explicitly set, a setting of check_log_space or
7923 check_log_inodes was ignored.
7925 79. If a space check for the spool or log partitions fails, the incident is now
7926 logged. Of course, in the latter case the data may get lost...
7928 80. Added the %p formatting code to string_format() so that it can be used to
7929 print addresses in debug_print(). Adjusted all the address printing in the
7930 debugging in store.c to use %p rather than %d.
7932 81. There was a concern that a line of code in smtp_in.c could overflow a
7933 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7934 initially expressed, the concern was not well-founded, because trailing
7935 spaces are removed early. However, if the trailing spaces were followed by
7936 a NULL, they did not get removed, so the overflow was possible. Two fixes
7939 (a) I re-wrote the offending code in a cleaner fashion.
7940 (b) If an incoming SMTP command contains a NULL character, it is rejected
7943 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7944 runs initgroups(), so that if the Exim user is in any additional groups,
7945 they will be used during message reception.
7951 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob