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.
8 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
9 SMTP connection" log lines.
11 JH/02 Option default value updates:
12 - queue_fast_ramp (main) true (was false)
13 - remote_max_parallel (main) 4 (was 2)
15 JH/03 Cache static regex pattern compilations, for use by ACLs.
17 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
18 Make the rewrite never match and keep the logging. Trust the
19 admin to be using verify=header-syntax (to actually reject the message).
21 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
22 (under Linux) we watch the dir containing the final file; previously
23 it would be the dir with the first symlink. We still do not monitor
26 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
27 dn_expand() is happy to pass them through.
29 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
30 the file with the proof had an unchanged name, the new proof(s) were
31 loaded on top of the old ones (and nover used; the old ones were stapled).
33 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
34 more than one message arrived in a single connection a reference from
35 the earlier message could be re-used. Often a sigsegv resulted.
36 These variables were introduced in Exim 4.87.
37 Debug help from Graeme Fowler.
39 JH/09 Fix ${filter } for conditions that modify $value. Previously the
40 modified version would be used in construction the result, and a memory
43 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
44 Find and fix by Jasen Betts.
46 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
47 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
48 the systemwide configuration to override the Exim config.
50 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
51 API changes in libopendmarc.
53 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
54 pid 1, in the normal "background daemon" mode, having to drop process-
55 group leadership also lost track of needing to create listener sockets.
57 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
58 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
60 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
61 a capture group which obtained no text (eg. "(abc)*" matching zero
62 occurrences) could cause a segfault if the corresponding $<n> was
65 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
66 included a close-brace character (eg. it itself used an expansion) an
69 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
70 starting TLS. Previously it was after, meaning that attackers on such
71 ports had to be screened using the host_reject_connection main config
72 option. The new sequence aligns better with the STARTTLS behaviour, and
73 permits defences against crypto-processing load attacks, even though it
74 is strictly an incompatible change.
75 Also, avoid sending any SMTP fail response for either the connect ACL
76 or host_reject_connection, for TLS-on-connect ports.
78 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
79 Previously this was not permitted, but it makes reasonable sense.
80 While there, restore a restriction on using it from a connect ACL; given
81 the change JH/16 it could only return false (and before 4.91 was not
88 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
89 after reception to before a subsequent reception. This should
90 mean slightly faster delivery, and also confirmation of reception
93 JH/02 Move from using the pcre library to pcre2. The former is no longer
94 being developed or supported (by the original developer).
96 JH/03 Constification work in the filters module required a major version
97 bump for the local-scan API. Specifically, the "headers_charset"
98 global which is visible via the API is now const and may therefore
99 not be modified by local-scan code.
101 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
102 sendfile() didi not account for the way the ClamAV driver code called it.
104 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
105 time check was being done for every character; replace that with one
108 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
109 sent was prefixed with a length byte.
111 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
112 RFC 5321. Previously Dovecot (at least) would log errors during
115 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
116 in 4.61 without a reason logged. Binaries will be bigger, which might
117 matter on diskspace-constrained systems, but debug is easier.
119 JH/09 Fix macro-definition during "-be" expansion testing. The move to
120 write-protected store for macros had not accounted for these runtime
121 additions; fix by removing this protection for "-be" mode.
123 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
124 handing out large-numbered file descriptors, violating the usual Unix
125 assumption (and required by Posix) that the lowest possible number will be
126 allocated by the kernel when a new one is needed. In the daemon, and any
127 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
128 are not useable for FD_SET() [and hence select()] and overwrite the stack.
129 Assorted crashes happen.
131 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
132 main-section options or in a connect ACL, the value from the first ever
133 connection was never replaced for subsequent connections. Found by
136 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
137 though only once PCRE2 was introduced: the memory accounting used under
138 debug offset allocations by an int, giving a hard trap in early startup.
139 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
141 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
142 with underbars is given. The write-protection of configuration introduced
143 in 4.95 trapped when normalisation was applied to an option not needing
146 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
148 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
149 failing to close the certificates directory, every hour or any time it
152 JH/16 Debugging initiated by an ACL control now continues through into routing
153 and transport processes. Previously debugging stopped any time Exim
154 re-execs, or for processing a queued message.
156 JH/17 The "expand" debug selector now gives more detail, specifically on the
157 result of expansion operators and items.
159 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
160 bad comparison between the option value and the name of the file to
161 be included was done, and a mismatch was wrongly identified.
162 4.88 to 4.95 are affected.
164 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
166 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
167 supplied as the db file-pair basename. Previously, if a directory
168 path was given, for example via the autoreply "once" option, the DB
169 file.pag and file.dir files would be created in that directory's
172 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
173 "taint" log_selector. These were previously deprecated.
175 JH/22 Fix static address-list lookups to properly return the matched item.
176 Previously only the domain part was returned.
178 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
179 the call into OpenSSL to send a TLS Close was being repeated; this
180 resulted in the library waiting for the peer's Close. If that was never
181 sent we waited forever. Fix by tracking send calls.
183 JH/24 The ${run} expansion item now expands its command string elements after
184 splitting. Previously it was before; the new ordering makes handling
185 zero-length arguments simpler. The old ordering can be obtained by
186 appending a new option "preexpand", after a comma, to the "run".
188 JH/25 Taint-check exec arguments for transport-initiated external processes.
189 Previously, tainted values could be used. This affects "pipe", "lmtp" and
190 "queryprogram" transport, transport-filter, and ETRN commands.
191 The ${run} expansion is also affected: in "preexpand" mode no part of
192 the command line may be tainted, in default mode the executable name
195 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
196 the the facility was not passed across execs, and only the first message
197 passed over a connection could use BDAT; any further ones using DATA.
199 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
200 uses $sending_ip_address and an interface is specified.
201 Previously any use of the local address in the EHLO name disabled
202 PIPECONNECT, the common case being to use the rDNS of it.
204 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
205 resumption. Previously verify failed because no certificate status is
206 passed on the wire for the restarted session. Fix by using the recorded
207 ocsp status of the stored session for the new connection.
209 JH/29 TLS resumption: the key for session lookup in the client now includes
210 more info that a server could potentially use in configuring a TLS
211 session, avoiding oferring mismatching sessions to such a server.
212 Previously only the server IP was used.
214 JH/30 Fix string_copyn() for limit greater than actual string length.
215 Previously the copied amount was the limit, which could result in a
216 overlapping memcpy for newly allocated destination soon after a
217 source string shorter than the limit. Found/investigated by KM.
219 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
220 close; it may be needed for a subsequent connection. This caused a
221 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
223 JH/32 Fix CHUNKING for a second message on a connection when the first was
224 rejected. Previously we did not reset the chunking-offered state, and
225 erroneously rejected the BDAT command. Investigation help from
228 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
229 an empty address. Previously the expansion returned an error.
231 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
232 proxy. Previously these were misparsed, leading to paniclog entries.
238 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
239 related applications. Previously an "H" was used where available info
240 says that "M" should be, so change to match.
242 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
243 as arguments, so an implementation trying to copy these into a local
244 buffer was taking a taint-enforcement trap. Fix by using dynamically
245 created buffers. Similar fix for radius expansion condition.
247 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
248 reasonable, eg. to count headers. Fix by using dynamically created
249 buffers rather than a local. Do similar fixes for ACL actions "dcc",
250 "log_reject_target", "malware" and "spam"; the arguments are expanded
251 so could be handling tainted values.
253 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
254 broken the (no-op) support for this sendmail command. Restore it
255 to doing nothing, silently, and returning good status.
257 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
258 record path was given (or the default used) without a leading directory
259 path, an error occurred on trying to open it. Use the transport's working
262 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
263 transport. Previously it was the name on the DNS A-record; use instead
264 the head of the CNAME chain leading there (if there is one). This seems
265 to align better with RFC 6125.
267 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
268 smtp_accept_max_per_host allocated resources which were not released
269 when the limit was exceeded. This eventually crashed the daemon. Fix
270 by adding a release action in that path.
272 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
273 expanded; previously using tainted values was rejected. Fix by using
274 dynamically-created buffers.
276 JH/09 Relax restrictions on ACL verify condition needing access to message
277 headers. Previously they were only permitted in data and non-smtp ACLs;
278 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
279 not_blind, header_sender and header_names_ascii verification.
281 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
282 Previously a macro used one argument twice; when called with the
283 argument as an expression having side-effects, incorrect operation
284 resulted. Use an inlineable function.
286 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
287 held open for a verify callout. Previously this wan not accounted for
288 and a corrupt onward SMTP conversation resulted.
290 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
291 Previously they were embedded naively in the constructed address; when
292 needed, strip the quoting and quote the entire local_part.
293 Also make the inbound_srs expansion condition handle quoting.
295 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
296 excluded, not matching the documentation.
298 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
299 was given for the sqlite_dbfile a trap resulted.
301 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
302 "name" argument resulted in a trap. There is no reason to disallow such;
303 this was a coding error.
305 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
306 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
307 spent suspended, ignoring the POSIX definition. Previously we assumed
308 it did and a constant offset from real time could be used as a correction.
309 Change to using the same clock source for the start-of-message and the
310 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
311 exists, just to get a clock slightly more aligned to reality.
313 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
314 RFC says it is optional some validators care. The missing char was not
315 intended but triggered by a line-wrap alignment. Discovery and fix by
316 Guillaume Outters, hacked on by JH.
318 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
319 name being quoted was tainted a trap would be taken. Fix by using
320 dynamically created buffers. The routine could have been called by a
321 rewrite with the "h" flag, by using the "-F" command-line option, or
322 by using a "name=" option on a control=submission ACL modifier.
324 JH/19 SPF: change the Authentication-Results expansion component to give
325 smtp.helo when the sender domain is empty. Previously it gave
328 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
329 not in 127.0.0.0/8 to help in spotting list domains taken over by a
330 domain-parking registrar.
332 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
333 Previously when a whitespace character was specified it was not inserted
334 after removing the newline.
336 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
337 the domain part of the recipient address. This overrides any tls_sni
338 option set, which was previously used.
340 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
343 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
344 is_tainted() had an off-by-one error in the overenthusiastic direction.
345 Find and fix by Gavan. Although NetBSD is not a supported platform for
346 4.94 this bug could affect other platforms.
348 PP/01 Fix default prime selection to be consistent.
349 One path used ike23 still, instead of exim.dev.20160529.3; now both
350 execution flows will use the same DH primes (currently
351 exim.dev.20160529.3).
353 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
354 option in smtp transport, to match the documentation. Previously
355 verification was not being done in some cases where it should have been.
357 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
358 than one server was defined and depending on the platform memory layout
359 details, an internal consistency trap could be hit while walking the list
362 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
363 files. The value is used by the authresults expansion item. Previously
364 if this was used in a router or transport, a crash could result.
366 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
367 have a sufficient incidence of EINTR returns from select that an
368 interaction with other operations done by the main daemon loop exposed
369 a bug in the error-handling. This was benign apart from the log
372 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
373 for consistency with delivered "=>" and failed "**" lines. While we're
374 there, handle PRX and TFO.
376 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
377 applied. This resulted, if any header-line rewrite rules were configured,
378 in a panic-log triggerable by sending a message with a long address in
379 a header. Fix by increasing the arbitrary limit to larger than a single
380 (dewrapped) 5322 header line maximum size.
382 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
383 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
384 RFC 6648 which deprecates X- options in protocols as a general practice.
385 Changeover between the implementations is handled by the mechanisms
388 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
389 a remote which uses callout/hold. Previously the local was queued.
391 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
394 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
395 log line. The asterisk is now consistently appended to the remote IP
396 (and port, if given), and will also be provided on defer and fail log
397 lines. Previously it could be placed on the local IP if that was being
398 logged, and was only provided on delivery lines.
400 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
402 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
403 authenticator client_send option. Previously the next char, after a pair
404 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
405 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
406 documentation. There is still no way to get a leading ^ immediately
407 after a NUL (ie. for the password of a PLAIN method authenticator.
409 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
410 files. For bad sizes read, delete the record and whine to paniclog.
412 JH/38 When logging an AUTH failure, as server, do not include sensitive
413 information. Previously, the credentials would be included if given
414 as part of the AUTH command line and an ACL denied authentication.
416 JH/39 Bug 2691: fix $local_part_data. When the matching list element
417 referred to a file, bad data was returned. This likely also affected
420 JH/40 The gsasl authenticator now supports caching of the salted password
421 generated by the client-side implementation. This required the addition
422 of a new variable: $auth4.
424 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
425 left undeleted; the attempt to re-create it then failed - resulting in
426 the usual "SIGHUP tp have daemon reload configuration" to not work.
427 This affected any platform not supporting "abstract" Unix-domain
428 sockets (i.e. not Linux).
430 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
431 recipients" response to RCPT in a later response, with a 250. The
432 previous coding assumed this would not happen, and under PIPELINING
433 would result in both lost and duplicate recipients for a message.
435 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
436 Previously the weighting was incorrectly applied. Similar fix for socks
437 proxies. Found and fixed by Heiko Schlichting.
439 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
440 not handle sub-lists included using the +namedlist syntax. While
441 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
442 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
445 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
446 Previously the permanent pool was used, so the sore could not be freed.
447 This meant a connection with many messages would use continually-growing
450 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
451 per-pool so as not to waste too much space. Previously a constant size
452 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
453 DOS attacks harder. The cost is wasted memory use in the larger blocks.
455 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
456 same space repeatedly, and a smaller process footprint.
458 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
459 for GnuTLS. Previously the name started "%s" which, while not a bug,
460 looked as if if might be one.
462 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
463 "continued connection") make the $proxy_* variables available. Previously
464 the information was not passed across the exec() call for subsequent
465 transport executions. This also mean that the log lines for the
466 messages can show the proxy information.
468 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
469 receive time. With modern systems the difference is significant.
470 The historical behaviour can be restored by disabling (a new) log_selector
471 "queue_time_exclusive".
473 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
474 content) were specifically tested for. Now, also cover expansions
475 resulting in ACL names and inline ACL content.
477 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
478 making it unusable in complex expressions.
480 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
481 fail with a taint-check on the spoolfile name, and leave the message
484 HS/01 Enforce absolute PID file path name.
486 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
488 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
490 PP/02 Bug 2643: Correct TLS DH constants.
491 A missing NUL termination in our code-generation tool had led to some
492 incorrect Diffie-Hellman constants in the Exim source.
493 Reported by kylon94, code-gen tool fix by Simon Arlott.
495 PP/03 Impose security length checks on various command-line options.
496 Fixes CVE-2020-SPRSS reported by Qualys.
498 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
499 better. Reported by Qualys.
501 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
502 providing a particularly obnoxious sender full name.
505 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
507 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
510 PP/08 Change default for recipients_max from unlimited to 50,000.
512 PP/09 Fix security issue with too many recipients on a message (to remove a
513 known security problem if someone does set recipients_max to unlimited,
514 or if local additions add to the recipient list).
515 Fixes CVE-2020-RCPTL reported by Qualys.
517 PP/10 Fix security issue in SMTP verb option parsing
518 Fixes CVE-2020-EXOPT reported by Qualys.
520 PP/11 Fix security issue in BDAT state confusion.
521 Ensure we reset known-good where we know we need to not be reading BDAT
522 data, as a general case fix, and move the places where we switch to BDAT
523 mode until after various protocol state checks.
524 Fixes CVE-2020-BDATA reported by Qualys.
526 HS/03 Die on "/../" in msglog file names
528 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
529 the Exim runtime user are allowed to create files.
531 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
534 QS/03 When reading the output from interpreted forward files we do not
535 pass the pipe between the parent and the interpreting process to
536 executed child processes (if any).
538 QS/04 Always die if requested from internal logging, even is logging is
541 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
542 the API; compilation noo longer completes with DMARC support included.
543 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
544 been reported on other platforms.
546 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
548 JH/56 Make the majority of info read from config files readonly, for defence-in-
549 depth against exploits. Suggestion by Qualys.
550 Not supported on Solaris 10.
552 JH/57 Fix control=fakreject for a custom message containing tainted data.
553 Previously this resulted in a log complaint, due to a re-expansion present
554 since fakereject was originally introduced.
556 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
557 modifier was given, a loop resulted.
559 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
560 pipelined SMTP command followed the BDAT LAST then it would be
561 incorrectly treated as part of the message body, causing a verification
564 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
565 require looser validation rules than those for 821-level addresses,
566 which only permit IP addresses.
572 JH/01 Avoid costly startup code when not strictly needed. This reduces time
573 for some exim process initialisations. It does mean that the logging
574 of TLS configuration problems is only done for the daemon startup.
576 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
578 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
579 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
582 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
583 DKIM signing is being done. Previously a transport_filter would always
584 disable CHUNKING, falling back to traditional DATA.
586 JH/05 Regard command-line recipients as tainted.
588 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
590 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
591 PAM library frees one of the arguments given to it, despite the
592 documentation. Therefore a plain malloc must be used.
594 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
595 on-stack buffers were used, resulting in a taint trap when DSN information
596 copied from a received message was written into the buffer.
598 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
599 the ordering of its ARC headers. This caused a crash.
601 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
602 a new record was being constructed with information from the peer, a trap
605 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
606 installation would get error messages from DMARC verify, when it hit the
607 nonexistent file indicated by the default. Distros wanting DMARC enabled
608 should both provide the file and set the option.
609 Also enforce no DMARC verification for command-line sourced messages.
611 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
612 could, depending on the platform, hang at the STARTTLS response.
614 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
615 message on a connection. Previously if one message had ARC headers and
616 the following one did not, a crash could result when adding an
617 Authentication-Results: header.
619 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
620 Exim main code and Exim-related utities. The introduction of taint
621 tracking also did many adjustments to string handling. Since then, eximon
622 frequently terminated with an assert failure.
624 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
625 check for 452 responses. This slightly helps the inefficieny of doing
626 a large alias-expansion into a recipient-limited target. The max_rcpt
627 transport option still applies (and at the current default, will override
628 the new feature). The check is done for either cause of synch, and forces
629 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
630 connection. The new facility is not tunable at this time.
632 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
633 library live data was being used, so the results became garbage. Make
634 copies while it is still usable.
636 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
637 on delivery deferred (==) and failed (**) lines (if a delivery was
638 attemtped). Previously it was only on completion (=>) lines.
640 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
641 for the expansion of the server_scram_iter and server_scram_salt options.
643 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
644 are now specifically given a NO_DATA response without hitting the system
645 resolver. The library goes on to do the now-standard TXT lookup.
646 Use of dnsdb lookups is not affected.
648 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
649 only retrieve the errormessage once. Previously two calls to dlerror()
650 were used, and the second one (for mainlog/paniclog) retrieved null
653 JH/20 Taint checking: disallow use of tainted data for
654 - the appendfile transport file and directory options
655 - the pipe transport command
656 - the autoreply transport file, log and once options
657 - file names used by the redirect router (including filter files)
659 - paths used by single-key lookups
660 Previously this was permitted.
662 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
663 adjusted the size of a major service buffer; this failed because the
664 buffer was in use at the time. Change to a compile-time increase in the
665 buffer size, when this authenticator is compiled into exim.
667 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
668 previous fast-mode was untenable in the face of glibs using mmap to
669 support larger malloc requests.
671 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
672 New values supported, if defined on system where compiled:
673 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
674 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
676 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
677 running a limited number of proceses in parallel, a benefit is gained. The
678 amount varies with the platform hardware and load. The use of the option
679 queue_run_in_order means we cannot do this, as ordering becomes
682 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
683 had introduced a string-copy (for ensuring NUL-termination) which was not
684 appropriate for that case, which can include embedded NUL bytes in the
685 block of data. Investigation showed the copy to actually be needless, the
686 data being length-specified.
688 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
689 done during a receiving connection, and both used TLS, global info was
690 used rather than per-connection info for tracking the state of data
691 queued for transmission. This could result in a connection hang.
693 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
694 Previously, when delivering serveral messages down a single connection
695 only the first would provide a SIZE. This was due to the size information
696 not being properly tracked.
698 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
699 TAI (at 37 seconds currently), pretend to be in UTC for time-related
700 expansion and logging. Previously, spurious values such as a future
701 minute could be seen.
703 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
704 it could crash from a null-deref. This could also affect the
705 ${addresses: } operator and ${readsock } item.
707 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
708 message following a mime one, the variable was not reset.
710 JH/30 When an pipelined-connect fails at the first response, assume incorrect
711 cached capability (perhaps the peer reneged?) and immediately retry in
714 JH/31 Fix spurious detection of timeout while writing to transport filter.
716 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
717 an attempt to copy the string was made before checking it.
719 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
720 taint of the lookup key was maintained; we now regard the presence in the
721 filesystem as sufficient validation.
723 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
724 argument is supplied.
726 JH/35 The dsearch lookup now requires that the directory is an absolute path.
727 Previously this was not checked, and nonempty relative paths made an
728 access under Exim's current working directory.
730 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
731 Previously no event was raised.
733 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
734 parameter supplied by the sender MAIL FROM command. Previously it was
735 ignored, and only the check_spool_space option value for the required
738 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
739 the size of the signing public-key. Previously it was instead giving
740 the size of the signature hash.
742 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
743 the default. See the (new) dkim_verify_min_keysizes option.
745 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
746 and an ACL use a lookup for checking either the local_part or domain,
747 stale data could be accessed. Ensure that variable references are
748 dropped between messages.
750 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
751 by the client was not checked as pointing within response data before
752 being used. A malicious client could thus cause an out-of-bounds read and
753 possibly gain authentication. Fix by adding the check.
755 JH/42 Internationalisation: change the default for downconversion in the smtp
756 transport to be "if needed". Previously it was "as previously set" for
757 the message, which usually meant "if needed" for message-submission but
758 "no" for everything else. However, MTAs have been seen using SMTPUTF8
759 even when the envelope addresses did not need it, resulting in forwarding
760 failures to non-supporting MTAs. A downconvert in such cases will be
761 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
762 transport. The change does mean that addresses needing conversion will
763 be converted when previously a delivery failure would occur.
765 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
766 response was received it would be used unchecked in a fail-DSN, violating
767 standards on line-length limits. Truncate if needed.
769 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
770 doesn't work. (Additionally add a new main config option to configure the
777 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
778 side, to decode a TLS 1.3 packet capture.
780 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
781 Previously the default library behaviour applied, sending two, each in
784 JH/03 Debug output for ACL now gives the config file name and line number for
787 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
789 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
791 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
792 buffer overrun for (non-chunking) other transports.
794 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
795 TLS1.3, means that a server rejecting a client certificate is not visible
796 to the client until the first read of encrypted data (typically the
797 response to EHLO). Add detection for that case and treat it as a failed
798 TLS connection attempt, so that the normal retry-in-clear can work (if
799 suitably configured).
801 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
802 and/or domain. Found and fixed by Jason Betts.
804 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
805 configuration). If a CNAME target was not a wellformed name pattern, a
808 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
809 the OS reports them interleaved with other addresses.
811 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
812 used both for input and for a verify callout, both encrypted, SMTP
813 responses being sent by the server could be lost. This resulted in
814 dropped connections and sometimes bounces generated by a peer sending
817 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
818 string. Previously it was possible to cause undefined behaviour in a
819 library routine (usually a crash). Found by "zerons".
821 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
824 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
825 API was removed, so update to use the newer ones.
827 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
828 any timeout set, is taking a long time. Previously we would hang on to a
829 rotated logfile "forever" if the input was arriving with long gaps
830 (a previous attempt to fix addressed lack, for a long time, of initial
833 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
834 shared (NFS) environment. The length of the tempfile name is now
835 4 + 16 ("hdr.$message_exim_id") which might break on file
836 systems which restrict the file name length to lower values.
839 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
840 shared (NFS) environment.
842 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
843 did for all versions <4.90). Notably -M, -m, --invert, -I may be
846 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
847 on some platforms for bit 31.
849 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
850 to changes apparently associated with TLS1.3 handling some of the APIs
851 previously used were either nonfunctional or inappropriate. Strings
852 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
853 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
854 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
855 This affects log line X= elements, the $tls_{in,out}_cipher variables,
856 and the use of specific cipher names in the encrypted= ACL condition.
858 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
860 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
861 verification result was not updated unless hosts_require_ocsp applied.
863 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
864 queue_list_requires_admin set to false, non-admin users were denied the
867 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
868 directory-of-certs mode. Previously they were advertised despite the
871 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
872 A single TCP connection by a client will now hold a TLS connection open
873 for multiple message deliveries, by default. Previously the default was to
876 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
877 default. If built with the facility, DANE will be used. The facility
878 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
880 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
881 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
882 must be defined and you must still, unless you define DISABLE_TLS, manage
883 the the include-dir and library-file requirements that go with that
884 choice. Non-TLS builds are still supported.
886 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
889 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
890 default. If the platform supports and has the facility enabled, it will
891 be requested on all coneections.
893 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
894 controlled by the build-time option SUPPORT_PIPE_CONNECT.
896 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
898 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
899 success-DSN messages. Previously the From: header was always the default
900 one for these; the option was ignored.
902 JH/28 Fix the timeout on smtp response to apply to the whole response.
903 Previously it was reset for every read, so a teergrubing peer sending
904 single bytes within the time limit could extend the connection for a
905 long time. Credit to Qualsys Security Advisory Team for the discovery.
907 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
908 delivery address, which leaked information of the results of local
909 forwarding. Change to the original envelope recipient address, per
912 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
913 requested. Previously not bounce was generated and a log entry of
914 error ignored was made.
916 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
918 JH/32 Introduce a general tainting mechanism for values read from the input
919 channel, and values derived from them. Refuse to expand any tainted
920 values, to catch one form of exploit.
922 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
923 was unused and the unexpanded text used for the test. Found and
924 fixed by Ruben Jenster.
926 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
927 an attempt to use a TLS library read routine dereffed a nul pointer,
930 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
931 them in our smtp response.
933 JH/36 Have the general router option retry_use_local_part default to true when
934 any of the restrictive preconditions are set (to anything). Previously it
935 was only for check_local user. The change removes one item of manual
936 configuration which is required for proper retries when a remote router
937 handles a subset of addresses for a domain.
939 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
940 link count into consideration.
942 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
943 caused the extension of big_buffer, the following lines were ignored.
945 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
946 accordance with RFC 2308. Previously there was no expiry, so a longlived
947 receive process (eg. due to ACL delays) versus a short SOA value could
950 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
952 JH/39 Promote DMARC support to mainline.
954 JH/40 Bug 2452: Add a References: header to DSNs.
956 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
957 parameters. The relevant library call is documented as "Deprecated: This
958 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
959 3.6.0, DH parameters are negotiated following RFC7919."
961 HS/06 Change the default of dnssec_request_domains to "*"
963 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
964 carried on and emitted a BDAT command, even when PIPELINING was not
967 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
968 buffer was used for the filename, resulting in a trap when tainted
969 arguments (eg. $domain) were used.
971 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
972 recommended to avoid a possible server-load attack. The feature can be
973 re-enabled via the openssl_options main cofiguration option.
975 JH/45 local_scan API: documented the current smtp_printf() call. This changed
976 for version 4.90 - adding a "more data" boolean to the arguments.
977 Bumped the ABI version number also, this having been missed previously;
978 release versions 4.90 to 4.92.3 inclusive were effectively broken in
979 respect of usage of smtp_printf() by either local_scan code or libraries
980 accessed via the ${dlfunc } expansion item. Both will need coding
981 adjustment for any calls to smtp_printf() to match the new function
982 signature; a FALSE value for the new argument is always safe.
984 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
985 the file-offset (which the Linux syscall does, and exim expects); this
986 resulted in an indefinite loop.
988 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
989 to do ARC verification. The Authentication-Results: header line added
990 by the configuration then had no ARC item.
992 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
993 an incorrect mode was used for reading the result, resulting in it being
1000 JH/01 Remove code calling the customisable local_scan function, unless a new
1001 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1003 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1004 non-signal-safe functions being used.
1006 JH/03 Bug 2269: When presented with a received message having a stupidly large
1007 number of DKIM-Signature headers, disable DKIM verification to avoid
1008 a resource-consumption attack. The limit is set at twenty.
1010 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1011 report of oldest_pass in ${authres } in consequence, and separate out
1012 some descriptions of reasons for verification fail.
1014 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1015 files in the spool were present and unlocked. A queue-runner could spot
1016 them, resulting in a duplicate delivery. Fix that by doing the unlock
1017 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1018 add more error-checking on spoolfile handling while that code is being
1021 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1022 No known attacks, no CVE, this is defensive hardening.
1024 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1025 a queue-runner could start a delivery while other operations were ongoing.
1026 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1027 Found and investigated by Tim Stewart. Fix by using the open message data
1028 file handle rather than opening another, and not locally closing it (which
1029 releases a lock) for that case, while creating the temporary .eml format
1030 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1032 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1033 $sender_verify_failure/$recipient_verify_failure to "random".
1035 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1038 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1039 Previously this would segfault.
1041 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1044 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1045 like zero, since the resolver should be doing this for us, But we need one
1046 as a CNAME but no MX presence gets the CNAME returned; we need to check
1047 that doesn't point to an MX to declare it "no MX returned" rather than
1048 "error, loop". A new main option is added so the older capability of
1049 following some limited number of chain links is maintained.
1051 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1053 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1054 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1055 now-deprecated earlier definitions used only specified the range up to TLS
1056 1.2 (in the older-version library docs).
1058 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1060 JH/15 Rework TLS client-side context management. Stop using a global, and
1061 explicitly pass a context around. This enables future use of TLS for
1062 connections to service-daemons (eg. malware scanning) while a client smtp
1063 connection is using TLS; with cutthrough connections this is quite likely.
1065 JH/16 Fix ARC verification to do AS checks in reverse order.
1067 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1069 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1070 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1071 Previously the "utf8" would be re-prepended for every additional message.
1073 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1074 Previously thery were accepted, resulting in issues when attempting to
1075 forward messages to a non-supporting MTA.
1077 PP/02 Let -n work with printing macros too, not just options.
1079 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1080 one parent address was copied, and bogus data was used at delivery-logging
1081 time. Either a crash (after delivery) or bogus log data could result.
1082 Discovery and analysis by Tim Stewart.
1084 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1085 Previously if the string ended mid-character, we did not insert the
1086 promised '?' replacement.
1088 PP/04 Documentation: current string operators work on bytes, not codepoints.
1090 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1091 should pack well giving a smaller memory footprint so better caching and
1092 therefore performance. Group the declarations where this can't be done so
1093 that the byte-sized flag variables are not interspersed among pointer
1094 variables, giving a better chance of good packing by the compiler.
1096 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1097 non-null, to avoid issues with sites running BATV. Previously reports were
1098 sent with an empty envelope sender so looked like bounces.
1100 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1101 The ignore_error flag wasn't being returned from the filter subprocess so
1102 was not set for later routers. Investigation and fix by Matthias Kurz.
1104 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1105 and a msg:complete for the whole, when a message is manually removed using
1106 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1108 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1109 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1110 hope that is portable enough.
1112 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1113 requiring a known-CA anchor certificate; make it now rely entirely on the
1114 TLSA as an anchor. Checking the name on the leaf cert against the name
1115 on the A-record for the host is still done for TA (but not for EE mode).
1117 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1118 included in delivery lines for non-proxied connections, when compiled with
1119 SUPPORT_SOCKS and running with proxy logging enabled.
1121 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1122 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1123 move the existing event to fire before the normal logging of message
1124 failure so that custom logging is bracketed by normal logging.
1126 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1127 msg:fail:internal event. Developement by Matthias Kurz.
1129 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1130 far too small for todays use of crypto signatures stored there. Go all
1131 the way to the max DNS message size of 64kB, even though this might be
1132 overmuch for IOT constrained device use.
1134 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1135 copy a string over itself. The library routine is documented as not
1136 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1138 JH/32 For main options check_spool_space and check_inode_space, where the
1139 platform supports 64b integers, support more than the previous 2^31 kB
1140 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1141 the previous G, M, k.
1143 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1144 $authenticated_fail_id variable on authentication failure. Previously
1147 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1148 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1149 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1150 GNUTLS_SEC_PARAM_MEDIUM.
1152 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1153 side. Previously we would continue as if no SNI had been received.
1155 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1156 "<" character, which should be a list-separator specification, we walked
1157 off past the nul-terimation.
1159 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1160 causes) even when the retry time is not yet met. Previously they were
1161 not, meaning that when (say) an account was over-quota and temp-rejecting,
1162 and multiple senders' messages were queued, only one sender would get
1163 notified on each configured delay_warning cycle.
1165 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1167 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1168 cipher-suites, an error can be left on the stack even for a succeeding
1169 accept; this results in impossible error messages when a later operation
1172 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1173 return error codes indicating retry. Under TLS1.3 this becomes required.
1175 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1176 it only wrote the new authenticators, resulting in a lack of tracking of
1177 peer changes of ESMTP extensions until the next cache flush.
1179 JH/41 Fix the loop reading a message header line to check for integer overflow,
1180 and more-often against header_maxsize. Previously a crafted message could
1181 induce a crash of the recive process; now the message is cleanly rejected.
1183 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1184 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1190 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1191 When redis_servers is set to a list of > 1 element, and the Redis servers
1192 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1193 case of MOVED into a DEFER case instead, thus moving the query onto the
1194 next server in the list. For a cluster of N elements, all N servers must
1195 be defined in redis_servers.
1197 GF/02 Catch and remove uninitialized value warning in exiqsumm
1198 Check for existence of @ARGV before looking at $ARGV[0]
1200 JH/01 Replace the store_release() internal interface with store_newblock(),
1201 which internalises the check required to safely use the old one, plus
1202 the allocate and data copy operations duplicated in both (!) of the
1203 extant use locations.
1205 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1206 modifier. This matches the restriction on the commandline.
1208 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1209 Previously only the last row was returned.
1211 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1212 we assumed that tags in the header were well-formed, and parsed the
1213 element content after inspecting only the first char of the tag.
1214 Assumptions at that stage could crash the receive process on malformed
1217 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1218 While running the DKIM ACL we operate on the Permanent memory pool so that
1219 variables created with "set" persist to the DATA ACL. Also (at any time)
1220 DNS lookups that fail create cache records using the Permanent pool. But
1221 expansions release any allocations made on the current pool - so a dnsdb
1222 lookup expansion done in the DKIM ACL releases the memory used for the
1223 DNS negative-cache, and bad things result. Solution is to switch to the
1224 Main pool for expansions.
1225 While we're in that code, add checks on the DNS cache during store_reset,
1226 active in the testsuite.
1227 Problem spotted, and debugging aided, by Wolfgang Breyha.
1229 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1230 When none of the hosts presented to a transport match an already-open
1231 connection, close it and proceed with the list. Previously we would
1232 queue the message. Spotted by Lena with Yahoo, probably involving
1235 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1236 Previously a spurious "250 OK id=" response was appended to the proper
1239 JH/08 The "support for" informational output now, which built with Content
1240 Scanning support, has a line for the malware scanner interfaces compiled
1241 in. Interface can be individually included or not at build time.
1243 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1244 by the template makefile "src/EDITME". The "STREAM" support for an older
1245 ClamAV interface method is removed.
1247 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1248 rows affected is given instead).
1250 JH/11 The runtime Berkeley DB library version is now additionally output by
1251 "exim -d -bV". Previously only the compile-time version was shown.
1253 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1254 SMTP connection. Previously, when one had more recipients than the
1255 first, an abortive onward connection was made. Move to full support for
1256 multiple onward connections in sequence, handling cutthrough connection
1257 for all multi-message initiating connections.
1259 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1260 routers. Previously, a multi-recipient message would fail to match the
1261 onward-connection opened for the first recipient, and cause its closure.
1263 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1264 a timeout on read on a GnuTLS initiating connection, resulting in the
1265 initiating connection being dropped. This mattered most when the callout
1266 was marked defer_ok. Fix to keep the two timeout-detection methods
1269 JH/15 Relax results from ACL control request to enable cutthrough, in
1270 unsupported situations, from error to silently (except under debug)
1271 ignoring. This covers use with PRDR, frozen messages, queue-only and
1274 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1276 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1277 metadata, resulting in a crash in free().
1279 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1280 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1281 Broken also in d185889f4, with init system revamp.
1283 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1284 Previously we abruptly closed the connection after reading a malware-
1285 found indication; now we go on to read the "scan ok" response line,
1288 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1289 ACL. Previously, a crash would result.
1291 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1292 macro text after a replacement (previously it was only once per line) and
1293 by skipping builtin macros when searching for an uppercase lead character.
1295 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1296 for the build is renamed.
1298 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1299 was allocated for every new TLS startup, meaning one per message. Fix
1300 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1302 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1303 reported the original. Fix to report (as far as possible) the ACL
1304 result replacing the original.
1306 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1307 OpenSSL. Certificate information is loaded for every new TLS startup,
1308 and the resources needed to be freed.
1310 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1312 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1315 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1316 DATA response info to the (existing) per-recipient response info for
1317 the "C=" log element. It can have useful tracking info from the
1318 destination system. Patch from Simon Arlott.
1320 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1321 length value. Previously this would segfault.
1323 HS/02 Support Avast multiline protoocol, this allows passing flags to
1324 newer versions of the scanner.
1326 JH/28 Ensure that variables possibly set during message acceptance are marked
1327 dead before release of memory in the daemon loop. This stops complaints
1328 about them when the debug_store option is enabled. Discovered specifically
1329 for sender_rate_period, but applies to a whole set of variables.
1330 Do the same for the queue-runner and queue-list loops, for variables set
1331 from spool message files. Do the same for the SMTP per-message loop, for
1332 certain variables indirectly set in ACL operations.
1334 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1335 as a multi-recipient message from a mailinglist manager). The coding had
1336 an arbitrary cutoff number of characters while checking for more input;
1337 enforced by writing a NUL into the buffer. This corrupted long / fast
1338 input. The problem was exposed more widely when more pipelineing of SMTP
1339 responses was introduced, and one Exim system was feeding another.
1340 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1341 receiving system, and refused recipients seen by the sending system
1342 (propating to people being dropped from mailing lists).
1343 Discovered and pinpointed by David Carter.
1345 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1346 replaced by the ${authresults } expansion.
1348 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1350 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1351 allows proper process termination in container environments.
1353 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1354 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1356 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1357 and "err_perm", deprecated since 4.83 when the RFC-defined words
1358 "temperror" and "permerror" were introduced.
1360 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1361 transport-filters or DKIM-signing. The restriction was lost in the
1362 consolidation of verify-callout and delivery SMTP handling.
1363 Extend the restriction to also cover ARC-signing.
1365 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1366 in defer=pass mode supply a 450 to the initiator. Previously the message
1369 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1370 tls_require_ciphers is used as before.
1372 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1373 "pass_unscanned". Only tmpfails from the scanner are written to
1374 the paniclog, as they may require admin intervention (permission
1375 denied, license issues). Other scanner errors (like decompression
1376 bombs) do not cause a paniclog entry.
1378 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1379 Previously it was possible to log spurious information in receive log
1382 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1383 triggered odd behaviour from Outlook Express clients.
1385 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1388 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1389 since the IETF WG has not yet settled on that versus the original
1390 "bare" representation.
1392 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1393 Previously the millisecond value corrupted the output.
1394 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1395 corrupted the output.
1401 JH/01 Rework error string handling in TLS interface so that the caller in
1402 more cases is responsible for logging. This permits library-sourced
1403 string to be attached to addresses during delivery, and collapses
1404 pairs of long lines into single ones.
1406 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1407 during configuration. Wildcards are allowed and expanded.
1409 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1410 This permits better logging.
1412 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1413 do not close it down and have the child transport start it up again on
1414 the passed-on TCP connection. Instead, proxy the child (and any
1415 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1416 affected: the continued delivery log lines do not have any DNSSEC, TLS
1417 Certificate or OCSP information. TLS cipher information is still logged.
1419 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1420 identical IP addresses on different listening ports. Will also affect
1423 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1424 add noisy ifdef guards to special-case this sillyness.
1425 Patch from Bernd Kuhls.
1427 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1428 than 255 are no longer allowed.
1430 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1431 Disable the session-cache too, which might reduce our load. Since we
1432 currrectly use a new context for every connection, both as server and
1433 client, there is no benefit for these.
1434 GnuTLS appears to not support tickets server-side by default (we don't
1435 call gnutls_session_ticket_enable_server()) but client side is enabled
1436 by default on recent versions (3.1.3 +) unless the PFS priority string
1439 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1440 <https://reproducible-builds.org/specs/source-date-epoch/>.
1442 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1443 the check for any unsuccessful recipients did not notice the limit, and
1444 erroneously found still-pending ones.
1446 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1447 MSG_MORE. Only in-clear (not on TLS connections).
1449 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1450 transport-filter is involved we can buffer the headers in memory for
1451 creating the signature, and read the spool data file once for the
1452 signature and again for transmission.
1454 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1455 4.77 as the kernel support then wasn't solid, having issues in 64bit
1456 mode. Now, it's been long enough. Add support for FreeBSD also.
1458 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1459 case where the routing stage had gathered several addresses to send to
1460 a host before calling the transport for the first, we previously failed
1461 to close down TLS in the old transport process before passing the TCP
1462 connection to the new process. The new one sent a STARTTLS command
1463 which naturally failed, giving a failed delivery and bloating the retry
1464 database. Investigation and fix prototype from Wolfgang Breyha.
1466 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1467 false-negatives in the check that the sender had not preempted a response
1468 or prompt from Exim (running as a server), due to that code's lack of
1469 awareness of the SMTP input buffering.
1471 PP/04 Add commandline_checks_require_admin option.
1472 Exim drops privileges sanely, various checks such as -be aren't a
1473 security problem, as long as you trust local users with access to their
1474 own account. When invoked by services which pass untrusted data to
1475 Exim, this might be an issue. Set this option in main configuration
1476 AND make fixes to the calling application, such as using `--` to stop
1479 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1480 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1481 responses to those, into a single TLS record each way (this usually means
1482 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1485 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1486 affects you only if you're dancing at the edge of the param size limits.
1487 If you are, and this message makes sense to you, then: raise the
1488 configured limit or use OpenSSL 1.1. Nothing we can do for older
1491 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1492 cmdline element to get the documented default one. Previously it was
1495 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1496 are unsuable for non-message-specific reasons - eg. connection timeout,
1497 banner-time rejection.
1499 JH/16 Fix logging of delivery remote port, when specified by router, under
1502 PP/06 Repair manualroute's ability to take options in any order, even if one
1503 is the name of a transport.
1506 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1508 JH/17 Change the list-building routines interface to use the expanding-string
1509 triplet model, for better allocation and copying behaviour.
1511 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1512 Previously it was constructed the first time a possibly-matching string
1513 was met in the configuration file input during startup; now it is done
1516 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1517 compatible one, to avoid the (poorly documented) possibility of a config
1518 file in the working directory redirecting the DB files, possibly correpting
1519 some existing file. CVE-2017-10140 assigned for BDB.
1521 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1522 cache-hot. Previously, although the result was properly cached, the
1523 initial verify call returned a defer.
1525 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1526 the main verify for receipient in uncached-mode.
1528 JH/22 Retire historical build files to an "unsupported" subdir. These are
1529 defined as "ones for which we have no current evidence of testing".
1531 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1532 if present. Previously it was ignored.
1534 JH/24 Start using specified-initialisers in C structure init coding. This is
1535 a C99 feature (it's 2017, so now considered safe).
1537 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1538 if was a fixed-sized field and bitmask ops via macros; it is now more
1541 PP/07 GitHub PR 56: Apply MariaDB build fix.
1542 Patch provided by Jaroslav Škarvada.
1544 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1545 during Coverity cleanups [4.87 JH/47]
1546 Diagnosis and fix provided by Michael Fischer v. Mollard.
1548 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1549 the right size to place the terminating semicolon on its own folded
1550 line, the header hash was calculated to an incorrect value thanks to
1551 the (relaxed) space the fold became.
1553 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1554 and confused the parent.
1556 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1557 which could crash as a result. This could lead to undeliverable messages.
1559 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1562 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1563 matters on fast-turnover and PID-randomising systems, which were getting
1564 out-of-order delivery.
1566 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1567 a possibly-overlapping copy. The symptom was that "Remote host closed
1568 connection in response to HELO" was logged instead of the actual 4xx
1571 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1572 Previously only that bufferd was discarded, resulting in SYMTP command
1575 JH/32 DKIM: when a message has multiple signatures matching an identity given
1576 in dkim_verify_signers, run the dkim acl once for each. Previously only
1577 one run was done. Bug 2189.
1579 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1580 "panic the current process" to "deliberately defer". The panic log is
1581 still written with the problem list name; the mail and reject logs now
1582 get a temp-reject line for the message that was being handled, saying
1583 something like "domains check lookup or other defer". The SMTP 451
1584 message is still "Temporary local problem".
1586 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1587 A crafted sequence of BDAT commands could result in in-use memory beeing
1588 freed. CVE-2017-16943.
1590 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1591 from SMTP input. Previously it was always done; now only done for DATA
1592 and not BDAT commands. CVE-2017-16944.
1594 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1595 to the message (such as an overlong header line). Previously this was
1596 not done and we did not exit BDAT mode. Followon from the previous item
1597 though a different problem.
1603 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1604 than -2003 did; needs libidn2 in addition to libidn.
1606 JH/02 The path option on a pipe transport is now expanded before use.
1608 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1609 Patch provided by "Björn", documentation fix added too.
1611 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1612 missing a wire-to-host endian conversion.
1614 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1615 close after a BDAT command line could be taken as a following command,
1616 giving a synch failure. Fix by only checking for synch immediately
1617 before acknowledging the chunk.
1619 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1620 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1621 macro. Patches provided by Josh Soref.
1623 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1624 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1625 by IANA as a service extension. However, John Klensin suggests that we
1628 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1629 the dkim code may be unix-mode line endings rather than smtp wire-format
1630 CRLF, so prepend a CR to any bare LF.
1632 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1633 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1635 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1636 the first were themselves being wrongly included in the feed into dkim
1637 processing; with most chunk sizes in use this resulted in an incorrect
1638 body hash calculated value.
1640 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1641 DKIM signature block, for verification. Although advised against by
1642 standards it is specifically not ruled illegal.
1644 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1646 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1647 missing a body hash (the bh= tag).
1649 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1650 It seems that HAProxy sends the Proxy Protocol information in clear and
1651 only then does a TLS startup, so do the same.
1653 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1654 TCP connections (such as for Spamd) unless the daemon successfully set
1655 Fast Open mode on its listening sockets. This fixes breakage seen on
1656 too-old kernels or those not configured for Fast Open, at the cost of
1657 requiring both directions being enabled for TFO, and TFO never being used
1658 by non-daemon-related Exim processes.
1660 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1661 endings, at least on the first header line. Try to canonify any that get
1662 past that check, despite the cost.
1664 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1665 now limited to an arbitrary five deep, while parsing addresses with the
1666 strip_excess_angle_brackets option enabled.
1668 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1669 instead leave the unprompted TLS handshake in socket buffer for the
1670 TLS library to consume.
1672 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1674 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1676 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1677 region in one operation, for speed, and this leaves assigned pointers
1678 dangling. Add checks run only under the testsuite which checks all
1679 variables at a store-reset and panics on a dangling pointer; add code
1680 explicitly nulling out all the variables discovered. Fixes one known
1681 bug: a transport crash, where a dangling pointer for $sending_ip_address
1682 originally assigned in a verify callout, is re-used.
1684 PP/06 Drop '.' from @INC in various Perl scripts.
1686 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1688 PP/08 Reduce a number of compilation warnings under clang; building with
1689 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1690 should be warning-free.
1692 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1694 HS/01 Fix portability problems introduced by PP/08 for platforms where
1695 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1697 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1698 chunk. This allows us to accept broken chunked messages. We need a more
1699 general solution here.
1701 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1702 already-broken messages in the queue.
1704 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1706 JH/19 Fix reference counting bug in routing-generated-address tracking.
1712 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1713 supports it and a size is available (ie. the sending peer gave us one).
1715 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1716 years of being deprecated). The replacements are the ACLs
1717 acl_smtp_mime and acl_not_smtp_mime.
1719 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1720 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1721 or even an in-clear connection were permitted. Now, if the host lookup
1722 was dnssec and dane was requested then the host is only used if the
1723 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1724 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1725 if one fails this test.
1726 This means that a poorly-configured remote DNS will make it incommunicado;
1727 but it protects against a DNS-interception attack on it.
1729 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1730 non-noisy when a race steals the message being considered.
1732 JH/05 If main configuration option tls_certificate is unset, generate a
1733 self-signed certificate for inbound TLS connections.
1735 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1736 in rewrites and routers.
1738 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1739 and logged a warning sing 4.83; now they are a configuration file error.
1741 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1742 (lacking @domain). Apply the same qualification processing as RCPT.
1744 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1746 JH/10 Support ${sha256:} applied to a string (as well as the previous
1749 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1750 a cutthrough deliver is pending, as we always want to make a connection.
1751 This also avoids re-routing the message when later placing the cutthrough
1752 connection after a verify cache hit.
1753 Do not update it with the verify result either.
1755 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1756 when routing results in more than one destination address.
1758 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1759 signing (which inhibits the cutthrough capability). Previously only
1760 the presence of an option was tested; now an expansion evaluating as
1761 empty is permissible (obviously it should depend only on data available
1762 when the cutthrough connection is made).
1764 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1765 the relevant preceding SMTP command did not note the pipelining mode.
1767 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1768 Previously they were not counted.
1770 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1771 as one having no matching records. Previously we deferred the message
1772 that needed the lookup.
1774 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1775 distinguished as "(=".
1777 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1778 for missing MX records. Previously it only worked for missing A records.
1780 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1782 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1783 after the data-go-ahead and data-ack. Patch from Jason Betts.
1785 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1786 even for a "none" policy. Patch from Tony Meyer.
1788 JH/22 Fix continued use of a connection for further deliveries. If a port was
1789 specified by a router, it must also match for the delivery to be
1792 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1793 When one of the recipients of a message was unsuitable for the connection
1794 (has no matching addresses), we lost track of needing to mark it
1795 deferred. As a result mail would be lost.
1797 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1799 JH/25 Decoding ACL controls is now done using a binary search; the source code
1800 takes up less space and should be simpler to maintain. Merge the ACL
1801 condition decode tables also, with similar effect.
1803 JH/26 Fix problem with one_time used on a redirect router which returned the
1804 parent address unchanged. A retry would see the parent address marked as
1805 delivered, so not attempt the (identical) child. As a result mail would
1808 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1809 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1810 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1813 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1814 Main config options check_{log,spool}_{inodes,space} are now
1815 100 inodes, 10MB unless set otherwise in the configuration.
1817 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1818 Previously it only applied to the main-section connection policy
1821 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1823 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1824 by me. Added RFC7919 DH primes as an alternative.
1826 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1827 are not in the system include path.
1829 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1830 GnuTLS, when a session startup failed (eg because the client disconnected)
1831 Exim did stdio operations after fclose. This was exposed by a recent
1832 change which nulled out the file handle after the fclose.
1834 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1835 signed directly by the cert-signing cert, rather than an intermediate
1836 OCSP-signing cert. This is the model used by LetsEncrypt.
1838 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1840 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1841 an incoming connection.
1843 HS/02 Bug 1802: Do not half-close the connection after sending a request
1846 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1847 fallback to "prime256v1".
1849 JH/34 SECURITY: Use proper copy of DATA command in error message.
1850 Could leak key material. Remotely exploitable. CVE-2016-9963.
1856 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1857 and 3.4.4 - once the server is enabled to respond to an OCSP request
1858 it does even when not requested, resulting in a stapling non-aware
1859 client dropping the TLS connection.
1861 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1862 support variable-length bit vectors. No functional change.
1864 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1865 The I= interface field on outgoing lines is now after the H= remote
1866 host field, same as incoming lines. There is a separate
1867 outgoing_interface log selector which allows you to disable the
1870 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1871 If not running log_selector +smtp_connection the mainlog would be held
1872 open indefinitely after a "too many connections" event, including to a
1873 deleted file after a log rotate. Leave the per net connection logging
1874 leaving it open for efficiency as that will be quickly detected by the
1875 check on the next write.
1877 HS/01 Bug 1671: Fix post transport crash.
1878 Processing the wait-<transport> messages could crash the delivery
1879 process if the message IDs didn't exist for some reason. When
1880 using 'split_spool_directory=yes' the construction of the spool
1881 file name failed already, exposing the same netto behaviour.
1883 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1884 mime_regex ACL conditions.
1886 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1887 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1888 to HELO, local diagnostic string.
1890 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1891 log line to debug. Even when configured with a tls authenticator many
1892 client connections are expected to not authenticate in this way, so
1893 an authenticate fail is not an error.
1895 HS/02 Add the Exim version string to the process info. This way exiwhat
1896 gives some more detail about the running daemon.
1898 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1899 matter for fast-change records such as DNSBLs.
1901 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1902 retry record, even if constant. There may be multiple transports with
1903 different interface settings and the retry behaviour needs to be kept
1906 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1908 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1910 JH/10 Bug 840: fix log_defer_output option of pipe transport
1912 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1913 hosts_randomize. This matters a lot when combined with mua_wrapper.
1915 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1916 ${quote_pgsql:<string>} operator.
1918 JH/13 Bug 1708: avoid misaligned access in cached lookup.
1920 JH/14 Change header file name for freeradius-client. Relevant if compiling
1921 with Radius support; from the Gentoo tree and checked under Fedora.
1923 JH/15 Bug 1712: Introduce $prdr_requested flag variable
1925 JH/16 Bug 1714: Permit an empty string as expansion result for transport
1926 option transport_filter, meaning no filtering.
1928 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
1930 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
1931 defaults to "*" (all hosts). The variable is now available when not built
1932 with TLS, default unset, mainly to enable keeping the testsuite sane.
1933 If a server certificate is not supplied (via tls_certificate) an error is
1934 logged, and clients will find TLS connections fail on startup. Presumably
1935 they will retry in-clear.
1936 Packagers of Exim are strongly encouraged to create a server certificate
1937 at installation time.
1939 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
1940 with the $config_file variable.
1942 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
1943 in transport context, after the attempt, and per-recipient. The latter type
1944 is per host attempted. The event data is the error message, and the errno
1945 information encodes the lookup type (A vs. MX) used for the (first) host,
1946 and the trailing two digits of the smtp 4xx response.
1948 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
1949 to write to mainlog (or rejectlog, paniclog) in the window between file
1950 creation and permissions/ownership being changed. Particularly affects
1951 installations where exicyclog is run as root, rather than exim user;
1952 result is that the running daemon panics and dies.
1954 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
1956 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
1957 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
1958 "pri" and "weight". Note that the previous implicit priority given by the
1959 list order is no longer honoured.
1961 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
1962 for DKIM processing.
1964 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
1965 by defining SUPPORT_SOCKS.
1967 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
1968 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
1969 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
1970 variables are renamed to proxy_{local,external}_{address,port}.
1972 JH/27 Move Internationalisation support from Experimental to mainline, enabled
1973 for a build by defining SUPPORT_I18N
1975 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
1976 of the query string, and make ${quote_redis:} do that quoting.
1978 JH/29 Move Events support from Experimental to mainline, enabled by default
1979 and removable for a build by defining DISABLE_EVENT.
1981 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
1983 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
1984 cached by the daemon.
1986 JH/32 Move Redis support from Experimental to mainline, enabled for a build
1987 by defining LOOKUP_REDIS. The libhiredis library is required.
1989 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
1990 keys are given for lookup.
1992 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
1993 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
1994 only supported when built with TLS support. The PolarSSL SHA routines
1995 are still used when the TLS library is too old for convenient support.
1997 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
1998 openssl_options), for security. OpenSSL forces this from version 1.1.0
1999 server-side so match that on older versions.
2001 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2002 allocation for $value could be released as the expansion processing
2003 concluded, but leaving the global pointer active for it.
2005 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2006 and to use the domains and local_parts ACL conditions.
2008 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2009 incorrectly not doubled on cutthrough transmission, hence seen as a
2010 body-termination at the receiving system - resulting in truncated mails.
2011 Commonly the sender saw a TCP-level error, and retransmitted the message
2012 via the normal store-and-forward channel. This could result in duplicates
2013 received - but deduplicating mailstores were liable to retain only the
2014 initial truncated version.
2016 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2018 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2020 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2021 we're in there, support oversigning also; bug 1309.
2023 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2025 HS/04 Add support for keep_environment and add_environment options.
2027 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2028 either intentional arithmetic overflow during PRNG, or testing config-
2031 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2032 delivery resulted in actual delivery. Cancel cutthrough before DATA
2035 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2036 rejects a recipient: pass the reject to the originator.
2038 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2039 Many were false-positives and ignorable, but it's worth fixing the
2042 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2043 for the new environment-manipulation done at startup. Move the routines
2044 from being local to tls.c to being global via the os.c file.
2046 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2047 an extract embedded as result-arg for a map, the first arg for extract
2048 is unavailable so we cannot tell if this is a numbered or keyed
2049 extraction. Accept either.
2055 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2058 JH/02 The smtp transport option "multi_domain" is now expanded.
2060 JH/03 The smtp transport now requests PRDR by default, if the server offers
2063 JH/04 Certificate name checking on server certificates, when exim is a client,
2064 is now done by default. The transport option tls_verify_cert_hostnames
2065 can be used to disable this per-host. The build option
2066 EXPERIMENTAL_CERTNAMES is withdrawn.
2068 JH/05 The value of the tls_verify_certificates smtp transport and main options
2069 default to the word "system" to access the system default CA bundle.
2070 For GnuTLS, only version 3.0.20 or later.
2072 JH/06 Verification of the server certificate for a TLS connection is now tried
2073 (but not required) by default. The verification status is now logged by
2074 default, for both outbound TLS and client-certificate supplying inbound
2077 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2080 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2081 Status Notification (bounce) messages are now MIME format per RFC 3464.
2082 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2083 under the control of the dsn_advertise_hosts option, and routers may
2084 have a dsn_lasthop option.
2086 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2087 default, modifiable by a malware= option. The list separator for
2088 the options can now be changed in the usual way. Bug 68.
2090 JH/10 The smtp_receive_timeout main option is now expanded before use.
2092 JH/11 The incoming_interface log option now also enables logging of the
2093 local interface on delivery outgoing connections.
2095 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2096 if the interface and destination host and port all match.
2098 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2101 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2102 Patch from Andrew Lewis.
2104 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2105 now supports optional time-restrictions, weighting, and priority
2106 modifiers per server. Patch originally by <rommer@active.by>.
2108 JH/16 The spamd_address main option now supports a mixed list of local
2109 and remote servers. Remote servers can be IPv6 addresses, and
2110 specify a port-range.
2112 JH/17 Bug 68: The spamd_address main option now supports an optional
2113 timeout value per server.
2115 JH/18 Bug 1581: Router and transport options headers_add/remove can
2116 now have the list separator specified.
2118 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2121 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2124 JH/21 Support for the A6 type of dns record is withdrawn.
2126 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2127 rather than the verbs used.
2129 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2130 from 255 to 1024 chars.
2132 JH/24 Verification callouts now attempt to use TLS by default.
2134 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2135 are generic router options now. The defaults didn't change.
2137 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2138 Original patch from Alexander Shikoff, worked over by JH.
2140 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2141 files if ZCAT_COMMAND is not executable.
2143 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2145 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2147 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2148 Normally benign, it bites when the pair was led to by a CNAME;
2149 modern usage is to not canonicalize the domain to a CNAME target
2150 (and we were inconsistent anyway for A-only vs AAAA+A).
2152 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2154 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2155 when evaluating $sender_host_dnssec.
2157 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2158 $sender_helo_dnssec variable.
2160 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2162 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2164 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2166 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2167 documented as working, but never had. Support all but $spam_report.
2169 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2170 added for tls authenticator.
2172 HS/03 Add perl_taintmode main config option
2178 TL/01 When running the test suite, the README says that variables such as
2179 no_msglog_check are global and can be placed anywhere in a specific
2180 test's script, however it was observed that placement needed to be near
2181 the beginning for it to behave that way. Changed the runtest perl
2182 script to read through the entire script once to detect and set these
2183 variables, reset to the beginning of the script, and then run through
2184 the script parsing/test process like normal.
2186 TL/02 The BSD's have an arc4random API. One of the functions to induce
2187 adding randomness was arc4random_stir(), but it has been removed in
2188 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2189 function when detected.
2191 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2192 cause callback expansion.
2194 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2195 syntax errors in an expansion can be treated as a string instead of
2196 logging or causing an error, due to the internal use of bool_lax
2197 instead of bool when processing it.
2199 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2200 server certificates when making smtp deliveries.
2202 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2204 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2206 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2208 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2209 Merged patch from Sebastian Wiedenroth.
2211 JH/05 Fix results-pipe from transport process. Several recipients, combined
2212 with certificate use, exposed issues where response data items split
2213 over buffer boundaries were not parsed properly. This eventually
2214 resulted in duplicates being sent. This issue only became common enough
2215 to notice due to the introduction of connection certificate information,
2216 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2218 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2219 size buffer was used, resulting in syntax errors when an expansion
2222 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2223 version 3.3.6 or later.
2225 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2226 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2227 both become "event_action", the variables become $event_name, $event_data
2228 and $event_defer_errno. There is a new variable $verify_mode, usable in
2229 routers, transports and related events. The tls:cert event is now also
2230 raised for inbound connections, if the main configuration event_action
2233 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2234 early OCSP support, but no stapling (appears to be less than 1.0.0).
2236 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2237 server certificate names available under the smtp transport option
2238 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2241 JH/10 Time-related extraction expansions from certificates now use the main
2242 option "timezone" setting for output formatting, and are consistent
2243 between OpenSSL and GnuTLS compilations. Bug 1541.
2245 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2246 encoded parameter in the incoming message. Bug 1558.
2248 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2249 include certificate info, eximon was claiming there were spoolfile
2252 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2254 JH/14 Log delivery-related information more consistently, using the sequence
2255 "H=<name> [<ip>]" wherever possible.
2257 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2258 are problematic for Debian distribution, omit them from the release
2261 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2263 JH/16 Fix string representation of time values on 64bit time_t architectures.
2266 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2267 output list separator was used.
2272 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2273 checkers that were complaining about end of non-void function with no
2276 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2277 This was a regression introduced in 4.83 by another bugfix.
2279 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2281 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2282 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2288 TF/01 Correctly close the server side of TLS when forking for delivery.
2290 When a message was received over SMTP with TLS, Exim failed to clear up
2291 the incoming connection properly after forking off the child process to
2292 deliver the message. In some situations the subsequent outgoing
2293 delivery connection happened to have the same fd number as the incoming
2294 connection previously had. Exim would try to use TLS and fail, logging
2295 a "Bad file descriptor" error.
2297 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2298 utilities have not been installed.
2300 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2301 temporary space as the ACL may create new global variables.
2303 TL/01 LDAP support uses per connection or global context settings, depending
2304 upon the detected version of the libraries at build time.
2306 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2307 to extract and use the src ip:port in logging and expansions as if it
2308 were a direct connection from the outside internet. PPv2 support was
2309 updated based on HAProxy spec change in May 2014.
2311 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2313 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2314 Properly escape header and check for NULL return.
2316 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2319 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2321 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2322 characters in header names, implemented as a verify condition.
2323 Contributed by Michael Fischer v. Mollard.
2325 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2326 results permerror and temperror. Previous values are deprecated but
2327 still accepted. In a future release, err_perm and err_temp will be
2328 completely removed, which will be a backward incompatibility if the
2329 ACL tests for either of these two old results. Patch contributed by
2330 user bes-internal on the mailing list.
2332 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2334 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2335 selectors, in both main and reject logs.
2337 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2340 JH/07 Add malware type "sock" for talking to simple daemon.
2342 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2344 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2345 routers/transports under cutthrough routing.
2347 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2348 numbers. Touch up "bool" conditional to keep the same definition.
2350 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2352 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2354 JH/12 Expand items in router/transport headers_add or headers_remove lists
2355 individually rather than the list as a whole. Bug 1452.
2357 Required for reasonable handling of multiple headers_ options when
2358 they may be empty; requires that headers_remove items with embedded
2359 colons must have them doubled (or the list-separator changed).
2361 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2362 view the policy declared in the DMARC record. Currently, $dmarc_status
2363 is a combined value of both the record presence and the result of the
2366 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2368 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2369 dnslookup router and the smtp transport (applying to the forward
2372 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2373 of ldap servers used for a specific lookup. Patch provided by Heiko
2376 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2377 New variable $lookup_dnssec_authenticated for observability.
2379 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2380 Patch submitted by Lars Timman.
2382 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2384 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2385 Requires trusted mode and valid format message id, aborts otherwise.
2386 Patch contributed by Heiko Schlichting.
2388 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2389 certextract with support for various fields. Bug 1358.
2391 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2392 is requested by default, modifiable by smtp transport option
2395 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2396 operate on certificate variables to give certificate fingerprints
2397 Also new ${sha256:cert_variable}.
2399 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2401 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2402 Christian Aistleitner.
2404 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2406 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2407 file. Patch from Wolfgang Breyha.
2409 JH/25 Expand the coverage of the delivery $host and $host_address to
2410 client authenticators run in verify callout. Bug 1476.
2412 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2413 align with daemon_smtp_ports. Bug 72.
2415 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2416 support and error reporting did not work properly.
2418 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2419 and is readable. Patch from Andrew Colin Kissa.
2421 TL/14 Enhance documentation of ${run expansion and how it parses the
2422 commandline after expansion, particularly in the case when an
2423 unquoted variable expansion results in an empty value.
2425 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2427 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2428 by renaming away from it. Addresses GH issue 15, reported by
2431 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2433 TL/15 SECURITY: prevent double expansion in math comparison functions
2434 (can expand unsanitized data). Not remotely exploitable.
2441 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2443 PP/02 Make -n do something, by making it not do something.
2444 When combined with -bP, the name of an option is not output.
2446 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2449 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2450 $sender_host_name and config options to manage this, and basic check
2453 PP/05 DSCP support for outbound connections and control modifier for inbound.
2455 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2456 (Only plugin which currently uses this is kerberos4, which nobody should
2457 be using, but we should make it available and other future plugins might
2458 conceivably use it, even though it would break NAT; stuff *should* be
2459 using channel bindings instead).
2461 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2462 name; added for Sendmail compatibility; requires admin caller.
2463 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2464 just ignore it); requires trusted caller.
2465 Also parse but ignore: -Ac -Am -X<logfile>
2468 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2470 TL/02 Add +smtp_confirmation as a default logging option.
2472 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2473 Patch by Magnus Holmgren from 2007-02-20.
2475 TL/04 Bugzilla 1281 - Spec typo.
2476 Bugzilla 1283 - Spec typo.
2477 Bugzilla 1290 - Spec grammar fixes.
2479 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2481 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2483 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2484 mailing list by by Dmitry Isaikin.
2486 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2488 JH/02 Support "G" suffix to numbers in ${if comparisons.
2490 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2492 NM/01 Bugzilla 1197 - Spec typo
2493 Bugzilla 1196 - Spec examples corrections
2495 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2497 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2498 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2501 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2502 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2504 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2505 "acl {{name}{arg}...}", and optional args on acl condition
2508 JH/05 Permit multiple router/transport headers_add/remove lines.
2510 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2512 JH/07 Avoid using a waiting database for a single-message-only transport.
2513 Performance patch from Paul Fisher. Bugzilla 1262.
2515 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2518 JH/09 Add $headers_added variable, with content from use of ACL modifier
2519 add_header (but not yet added to the message). Bugzilla 199.
2521 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2522 Pulled from Bugzilla 817 by Wolfgang Breyha.
2524 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2526 (nb: this is the same fix as in Exim 4.80.1)
2528 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2531 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2534 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2536 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2537 advertise SMTP AUTH mechanism to us, instead of a generic
2538 protocol violation error. Also, make Exim more robust to bad
2539 data from the Dovecot auth socket.
2541 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2543 When a queue runner is handling a message, Exim first routes the
2544 recipient addresses, during which it prunes them based on the retry
2545 hints database. After that it attempts to deliver the message to
2546 any remaining recipients. It then updates the hints database using
2549 So if a recipient address works intermittently, it can get repeatedly
2550 deferred at routing time. The retry hints record remains fresh so the
2551 address never reaches the final cutoff time.
2553 This is a fairly common occurrence when a user is bumping up against
2554 their storage quota. Exim had some logic in its local delivery code
2555 to deal with this. However it did not apply to per-recipient defers
2556 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2558 This change adds a proper retry rule check during routing so that the
2559 final cutoff time is checked against the message's age. We only do
2560 this check if there is an address retry record and there is not a
2561 domain retry record; this implies that previous attempts to handle
2562 the address had the retry_use_local_parts option turned on. We use
2563 this as an approximation for the destination being like a local
2564 delivery, as in LMTP.
2566 I suspect this new check makes the old local delivery cutoff check
2567 redundant, but I have not verified this so I left the code in place.
2569 TF/02 Correct gecos expansion when From: is a prefix of the username.
2571 Test 0254 submits a message to Exim with the header
2575 When I ran the test suite under the user fanf2, Exim expanded
2576 the header to contain my full name, whereas it should have added
2577 a Resent-Sender: header. It erroneously treats any prefix of the
2578 username as equal to the username.
2580 This change corrects that bug.
2582 GF/01 DCC debug and logging tidyup
2583 Error conditions log to paniclog rather than rejectlog.
2584 Debug lines prefixed by "DCC: " to remove any ambiguity.
2586 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2588 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2589 Bug spotted by Jeremy Harris; was flawed since initial commit.
2590 Would have resulted in OCSP responses post-SNI triggering an Exim
2591 NULL dereference and crash.
2593 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2595 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2596 Bug detection, analysis and fix by Samuel Thibault.
2597 Bugzilla 1331, Debian bug #698092.
2599 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2601 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2602 Server implementation by Todd Lyons, client by JH.
2603 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2604 config variable "prdr_enable" controls whether the server
2605 advertises the facility. If the client requests PRDR a new
2606 acl_data_smtp_prdr ACL is called once for each recipient, after
2607 the body content is received and before the acl_smtp_data ACL.
2608 The client is controlled by both of: a hosts_try_prdr option
2609 on the smtp transport, and the server advertisement.
2610 Default client logging of deliveries and rejections involving
2611 PRDR are flagged with the string "PRDR".
2613 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2614 fclose(). Diagnosis by Todd Lyons.
2616 PP/17 Update configure.default to handle IPv6 localhost better.
2617 Patch by Alain Williams (plus minor tweaks).
2620 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2621 This is now consistent with GnuTLS, and is now documented: the
2622 previous undocumented portable approach to treating the option as
2623 unset was to force an expansion failure. That still works, and
2624 an empty string is now equivalent.
2626 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2627 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2628 not performing validation itself.
2630 PP/20 Added force_command boolean option to pipe transport.
2631 Patch from Nick Koston, of cPanel Inc.
2633 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2636 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2638 PP/21 Fix eximon continuous updating with timestamped log-files.
2639 Broken in a format-string cleanup in 4.80, missed when I repaired the
2640 other false fix of the same issue.
2641 Report and fix from Heiko Schlichting.
2644 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2645 Report from Prashanth Katuri.
2647 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2648 It's SecureTransport, so affects any MacOS clients which use the
2649 system-integrated TLS libraries, including email clients.
2651 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2652 using a MIME ACL for non-SMTP local injection.
2653 Report and assistance in diagnosis by Warren Baker.
2655 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2657 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2659 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2660 last id that failed so it may be referenced in subsequent ACL's.
2662 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2665 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2666 ldap library initialization, allowing self-signed CA's to be
2667 used. Also properly sets require_cert option later in code by
2668 using NULL (global ldap config) instead of ldap handle (per
2669 session). Bug diagnosis and testing by alxgomz.
2671 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2672 the src/util/ subdirectory.
2674 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2675 renamed to Transport Post Delivery Action by Jeremy Harris, as
2678 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2679 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2680 redis_servers = needs to be configured which will be used by the redis
2681 lookup. Patch from Warren Baker, of The Packet Hub.
2683 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2685 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2686 hostname or reverse DNS when processing a host list. Used suggestions
2687 from multiple comments on this bug.
2689 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2691 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2692 Missed a few lines, added it to make the runtest require no keyboard
2695 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2696 contains upper case chars. Make router use caseful_local_part.
2698 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2699 support when GnuTLS has been built with p11-kit.
2705 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2707 This, or similar/improved, will also be change PP/11 of 4.82.
2713 PP/01 Handle short writes when writing local log-files.
2714 In practice, only affects FreeBSD (8 onwards).
2715 Bugzilla 1053, with thanks to Dmitry Isaikin.
2717 NM/01 Bugzilla 949 - Documentation tweak
2719 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2722 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2724 PP/02 Implemented gsasl authenticator.
2726 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2728 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2729 `pkg-config foo` for cflags/libs.
2731 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2732 with rest of GSASL and with heimdal_gssapi.
2734 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2735 `pkg-config foo` for cflags/libs for the TLS implementation.
2737 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2738 properties get this fed in as external SSF. A number of robustness
2739 and debugging improvements to the cyrus_sasl authenticator.
2741 PP/08 cyrus_sasl server now expands the server_realm option.
2743 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2744 Patch by Jeremy Harris.
2746 PP/10 Added dbmjz lookup type.
2748 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2750 PP/12 MAIL args handles TAB as well as SP, for better interop with
2751 non-compliant senders.
2752 Analysis and variant patch by Todd Lyons.
2754 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2755 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2756 Patch from Dirk Mueller <dmueller@suse.com>
2758 PP/13 tls_peerdn now print-escaped for spool files.
2759 Observed some $tls_peerdn in wild which contained \n, which resulted
2760 in spool file corruption.
2762 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2763 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2764 or write after TLS renegotiation, which otherwise led to messages
2767 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2768 as a tracking header (ie: a signed header comes before the signature).
2769 Patch from Wolfgang Breyha.
2771 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2772 comma-sep list; embedded commas doubled.
2774 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2776 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2778 Report and patch from Dmitry Banschikov.
2780 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2781 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2782 failures. We appear to now support TLS1.1+ with Exim.
2784 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2785 lets Exim select keys and certificates based upon TLS SNI from client.
2786 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2787 before an outbound SMTP session. New log_selector, +tls_sni.
2789 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2790 NULL dereference. Report and patch from Alun Jones.
2792 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2793 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2794 Not seeing resolver debug output on NetBSD, but suspect this is a
2795 resolver implementation change.
2797 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2798 Left warnings. Added "eximon gdb" invocation mode.
2800 PP/21 Defaulting "accept_8bitmime" to true, not false.
2802 PP/22 Added -bw for inetd wait mode support.
2804 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2805 locate the relevant includes and libraries. Made this the default.
2807 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2808 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2810 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2811 This may cause build issues on older platforms.
2813 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2814 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2815 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2816 Added SNI support via GnuTLS too.
2817 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2819 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2821 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2822 Applied second patch from Janne, implementing suggestion to default
2823 multiple-strings-in-record handling to match SPF spec.
2825 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2827 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2828 read-only, out of scope).
2829 Patch from Wolfgang Breyha, report from Stuart Northfield.
2831 PP/29 Fix three issues highlighted by clang analyser static analysis.
2832 Only crash-plausible issue would require the Cambridge-specific
2833 iplookup router and a misconfiguration.
2834 Report from Marcin Mirosław.
2836 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2838 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2839 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2840 As part of this, removing so much warning spew let me fix some minor
2841 real issues in debug logging.
2843 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2844 assignment on my part. Fixed.
2846 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2847 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2848 Janne Snabb (who went above and beyond: thank you).
2850 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2851 string otherwise requires a connection and a bunch more work and it's
2852 relatively easy to get wrong. Should also expose TLS library linkage
2855 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2856 64-bit ${eval} (JH/03).
2858 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2859 GNU libc to support some of the 64-bit stuff, should not lead to
2860 conflicts. Defined before os.h is pulled in, so if a given platform
2861 needs to override this, it can.
2863 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2864 protection layer was required, which is not implemented.
2865 Bugzilla 1254, patch from Wolfgang Breyha.
2867 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2868 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2869 tls_dhparam take prime identifiers. Also unbreak combination of
2870 OpenSSL+DH_params+TLSSNI.
2872 PP/39 Disable SSLv2 by default in OpenSSL support.
2878 PP/01 Solaris build fix for Oracle's LDAP libraries.
2879 Bugzilla 1109, patch from Stephen Usher.
2881 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2883 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2886 TF/02 Fix a couple more cases where we did not log the error message
2887 when unlink() failed. See also change 4.74-TF/03.
2889 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2890 lock up or crash if it happened to be inside a call to libc when it
2891 got a SIGUSR1 from exiwhat.
2893 The SIGUSR1 handler appends the current process status to the process
2894 log which is later printed by exiwhat. It used to use the general
2895 purpose logging code to do this, but several functions it calls are
2896 not safe for signals.
2898 The new output code in the SIGUSR1 handler is specific to the process
2899 log, and simple enough that it's easy to inspect for signal safety.
2900 Removing some special cases also simplifies the general logging code.
2901 Removing the spurious timestamps from the process log simplifies
2904 TF/04 Improved ratelimit ACL condition.
2906 The /noupdate option has been deprecated in favour of /readonly which
2907 has clearer semantics. The /leaky, /strict, and /readonly update modes
2908 are mutually exclusive. The update mode is no longer included in the
2909 database key; it just determines when the database is updated. (This
2910 means that when you upgrade Exim will forget old rate measurements.)
2912 Exim now checks that the per_* options are used with an update mode that
2913 makes sense for the current ACL. For example, when Exim is processing a
2914 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2915 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2916 must specify per_mail/readonly. If you omit the update mode it defaults to
2917 /leaky where that makes sense (as before) or /readonly where required.
2919 The /noupdate option is now undocumented but still supported for
2920 backwards compatibility. It is equivalent to /readonly except that in
2921 ACLs where /readonly is required you may specify /leaky/noupdate or
2922 /strict/noupdate which are treated the same as /readonly.
2924 A useful new feature is the /count= option. This is a generalization
2925 of the per_byte option, so that you can measure the throughput of other
2926 aggregate values. For example, the per_byte option is now equivalent
2927 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
2929 The per_rcpt option has been generalized using the /count= mechanism
2930 (though it's more complicated than the per_byte equivalence). When it is
2931 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
2932 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
2933 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
2934 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
2935 Note that using per_rcpt with a non-readonly update mode in more than
2936 one ACL will cause the recipients to be double-counted. (The per_mail
2937 and per_byte options don't have this problem.)
2939 The handling of very low rates has changed slightly. If the computed rate
2940 is less than the event's count (usually one) then this event is the first
2941 after a long gap. In this case the rate is set to the same as this event's
2942 count, so that the first message of a spam run is counted properly.
2944 The major new feature is a mechanism for counting the rate of unique
2945 events. The new per_addr option counts the number of different
2946 recipients that someone has sent messages to in the last time period. It
2947 behaves like per_rcpt if all the recipient addresses are different, but
2948 duplicate recipient addresses do not increase the measured rate. Like
2949 the /count= option this is a general mechanism, so the per_addr option
2950 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
2951 example, measure the rate that a client uses different sender addresses
2952 with the options per_mail/unique=$sender_address. There are further
2953 details in the main documentation.
2955 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
2957 TF/06 Removed a few PCRE remnants.
2959 TF/07 Automatically extract Exim's version number from tags in the git
2960 repository when doing development or release builds.
2962 PP/02 Raise smtp_cmd_buffer_size to 16kB.
2963 Bugzilla 879. Patch from Paul Fisher.
2965 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
2966 Heavily based on revision 40f9a89a from Simon Arlott's tree.
2969 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
2971 PP/05 Variable $av_failed, true if the AV scanner deferred.
2972 Bugzilla 1078. Patch from John Horne.
2974 PP/06 Stop make process more reliably on build failure.
2975 Bugzilla 1087. Patch from Heiko Schlittermann.
2977 PP/07 Make maildir_use_size_file an _expandable_ boolean.
2978 Bugzilla 1089. Patch from Heiko Schlittermann.
2980 PP/08 Handle ${run} returning more data than OS pipe buffer size.
2981 Bugzilla 1131. Patch from Holger Weiß.
2983 PP/09 Handle IPv6 addresses with SPF.
2984 Bugzilla 860. Patch from Wolfgang Breyha.
2986 PP/10 GnuTLS: support TLS 1.2 & 1.1.
2988 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
2991 PP/11 match_* no longer expand right-hand-side by default.
2992 New compile-time build option, EXPAND_LISTMATCH_RHS.
2993 New expansion conditions, "inlist", "inlisti".
2995 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
2997 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
2999 PP/14 fix log_write() format string regression from TF/03.
3000 Bugzilla 1152. Patch from Dmitry Isaikin.
3006 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3008 PP/02 Harmonised TLS library version reporting; only show if debugging.
3009 Layout now matches that introduced for other libraries in 4.74 PP/03.
3011 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3013 PP/04 New "dns_use_edns0" global option.
3015 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3018 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3019 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3021 TK/01 Updated PolarSSL code to 0.14.2.
3022 Bugzilla 1097. Patch from Andreas Metzler.
3024 PP/07 Catch divide-by-zero in ${eval:...}.
3025 Fixes bugzilla 1102.
3027 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3030 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3031 format-string attack -- SECURITY: remote arbitrary code execution.
3033 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3034 time unintentionally subject to list matching rules, letting the header
3035 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3036 arbitrary string expansion). This allowed for information disclosure.
3038 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3039 INT_MIN/-1 -- value coerced to INT_MAX.
3045 NM/01 Workaround for PCRE version dependency in version reporting
3048 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3049 This fixes portability to compilers other than gcc, notably
3050 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3052 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3053 makefiles for portability to HP-UX and POSIX correctness.
3055 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3056 Also via indented variable definition in the Makefile.
3057 (Debugging by Oliver Heesakkers).
3059 PP/02 Restore caching of spamd results with expanded spamd_address.
3060 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3062 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3063 Improves build reliability. Fix from: Frank Elsner
3065 NM/02 Fix wide character breakage in the rfc2047 coding
3066 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3068 NM/03 Allow underscore in dnslist lookups
3069 Fixes bug 1026. Patch from Graeme Fowler
3071 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3072 Code patches from Adam Ciarcinski of NetBSD.
3074 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3077 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3078 is logged, to avoid truncation. Patch from John Horne.
3080 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3081 Patch from Jakob Hirsch.
3083 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3084 SQL string expansion failure details.
3085 Patch from Andrey Oktyabrski.
3087 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3088 Patch from Simon Arlott.
3090 PP/09 New lookups functionality failed to compile on old gcc which rejects
3091 extern declarations in function scope.
3092 Patch from Oliver Fleischmann
3094 PP/10 Use sig_atomic_t for flags set from signal handlers.
3095 Check getgroups() return and improve debugging.
3096 Fixed developed for diagnosis in bug 927 (which turned out to be
3099 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3100 Patch from Mark Zealey.
3102 PP/12 Bugzilla 1056: Improved spamd server selection.
3103 Patch from Mark Zealey.
3105 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3106 Based on patch from Heiko Schlittermann.
3108 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3109 Patch from Uwe Doering, sign-off by Michael Haardt.
3111 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3112 variable declaration deep within a block. Bug and patch from
3115 PP/15 lookups-Makefile IRIX compatibility coercion.
3117 PP/16 Make DISABLE_DKIM build knob functional.
3119 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3120 Patch by Simon Arlott
3122 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3123 variable argument macros. Our copy now differs from upstream.
3129 TF/01 Failure to get a lock on a hints database can have serious
3130 consequences so log it to the panic log.
3132 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3133 controlled using the smtp_confirmation log selector.
3135 TF/03 Include the error message when we fail to unlink a spool file.
3137 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3138 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3139 for maintaining out-of-tree patches for some time.
3141 PP/01 Bugzilla 139: Documentation and portability issues.
3142 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3143 Handle per-OS dynamic-module compilation flags.
3145 PP/02 Let /dev/null have normal permissions.
3146 The 4.73 fixes were a little too stringent and complained about the
3147 permissions on /dev/null. Exempt it from some checks.
3148 Reported by Andreas M. Kirchwitz.
3150 PP/03 Report version information for many libraries, including
3151 Exim version information for dynamically loaded libraries. Created
3152 version.h, now support a version extension string for distributors
3153 who patch heavily. Dynamic module ABI change.
3155 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3156 privilege escalation vulnerability whereby the Exim run-time user
3157 can cause root to append content of the attacker's choosing to
3160 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3163 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3164 If dropping privileges for untrusted macros, we disabled normal logging
3165 on the basis that it would fail; for the Exim run-time user, this is not
3166 the case, and it resulted in successful deliveries going unlogged.
3167 Fixed. Reported by Andreas Metzler.
3173 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3174 only prepend for the Resent-* case. Fixes regression introduced in
3175 Exim 4.70 by NM/22 for Bugzilla 607.
3177 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3178 increasing numbers of administrators be bitten by this.
3180 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3182 PP/03 Bugzilla 994: added openssl_options main configuration option.
3184 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3186 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3188 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3190 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3191 without a peer certificate, leading to a segfault because of an
3192 assumption that peers always have certificates. Be a little more
3193 paranoid. Problem reported by Martin Tscholak.
3195 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3196 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3197 NB: ClamAV planning to remove STREAM in "middle of 2010".
3198 CL also introduces -bmalware, various -d+acl logging additions and
3199 more caution in buffer sizes.
3201 PP/09 Implemented reverse_ip expansion operator.
3203 PP/10 Bugzilla 937: provide a "debug" ACL control.
3205 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3207 PP/12 Bugzilla 973: Implement --version.
3209 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3211 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3213 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3215 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3216 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3217 ignore trailing whitespace.
3219 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3221 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3224 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3225 Notification from Dr Andrew Aitchison.
3227 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3228 ExtendedDetectionInfo response format.
3229 Notification from John Horne.
3231 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3234 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3235 XSL and documented dependency on system catalogs, with examples of how
3238 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3241 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3242 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3243 configuration file which is writeable by the Exim user or group.
3245 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3246 of configuration files to cover files specified with the -C option if
3247 they are going to be used with root privileges, not just the default
3250 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3251 option (effectively making it always true).
3253 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3254 files to be used while preserving root privileges.
3256 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3257 that rogue child processes cannot use them.
3259 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3260 run-time user, instead of root.
3262 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3263 Exim run-time user without dropping privileges.
3265 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3266 result string, instead of calling string_vformat() twice with the same
3269 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3270 for other users. Others should always drop root privileges if they use
3271 -C on the command line, even for a whitelisted configure file.
3273 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3275 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3281 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3282 $data_path, and $header_path variables; fixed documentation bugs and
3285 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3286 exipick to access non-standard spools, including the "frozen" queue
3289 NM/01 Bugzilla 965: Support mysql stored procedures.
3290 Patch from Alain Williams
3292 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3294 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3295 Patch from Andreas Metzler
3297 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3298 Patch from Kirill Miazine
3300 NM/05 Bugzilla 671: Added umask to procmail example.
3302 JJ/03 installed exipick 20100323.0, fixing doc bug
3304 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3305 directory. Notification and patch from Dan Rosenberg.
3307 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3309 TK/02 Improve log output when DKIM signing operation fails.
3311 MH/01 Treat the transport option dkim_domain as a colon separated
3312 list, not as a single string, and sign the message with each element,
3313 omitting multiple occurences of the same signer.
3315 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3316 Bugzilla 985, 986. Patch by Simon Arlott
3318 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3319 Patch by Simon Arlott
3321 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3322 MBX locking. Notification from Dan Rosenberg.
3328 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3330 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3332 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3334 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3336 NM/04 Bugzilla 915: Flag broken perl installation during build.
3342 TK/01 Added patch by Johannes Berg that expands the main option
3343 "spamd_address" if it starts with a dollar sign.
3345 TK/02 Write list of recipients to X-Envelope-Sender header when building
3346 the mbox-format spool file for content scanning (suggested by Jakob
3349 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3350 (http://www.dcc-servers.net/) support via dccifd. Activated by
3351 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3353 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3354 by Mark Daniel Reidel <mr@df.eu>.
3356 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3357 When building exim an external PCRE library is now needed -
3358 PCRE is a system library on the majority of modern systems.
3359 See entry on PCRE_LIBS in EDITME file.
3361 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3362 conversation. Added nologin parameter to request.
3363 Patch contributed by Kirill Miazine.
3365 TF/01 Do not log submission mode rewrites if they do not change the address.
3367 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3369 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3370 log files in place. Contributed by Roberto Lima.
3372 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3374 TF/03 Bugzilla 615: When checking the local_parts router precondition
3375 after a local_part_suffix or local_part_prefix option, Exim now
3376 does not use the address's named list lookup cache, since this
3377 contains cached lookups for the whole local part.
3379 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3380 Robert Millan. Documentation is in experimental-spec.txt.
3382 TF/04 Bugzilla 668: Fix parallel build (make -j).
3384 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3386 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3387 Patch provided by Jan Srzednicki.
3389 TF/05 Leading white space used to be stripped from $spam_report which
3390 wrecked the formatting. Now it is preserved.
3392 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3393 that they are available at delivery time.
3395 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3397 TF/08 TLS error reporting now respects the incoming_interface and
3398 incoming_port log selectors.
3400 TF/09 Produce a more useful error message if an SMTP transport's hosts
3401 setting expands to an empty string.
3403 NM/06 Bugzilla 744: EXPN did not work under TLS.
3404 Patch provided by Phil Pennock.
3406 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3407 Patch provided by Richard Godbee.
3409 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3410 acl_smtp_notquit, added index entry.
3412 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3413 Patch provided by Eugene Bujak.
3415 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3416 accept(). Patch provided by Maxim Dounin.
3418 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3419 Patch provided by Phil Pennock.
3421 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3423 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3424 Patch provided by Brad "anomie" Jorsch.
3426 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3427 Patch provided by Dean Brooks.
3429 TK/05 Add native DKIM support (does not depend on external libraries).
3431 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3432 Patch provided by Graeme Fowler.
3434 NM/16 Bugzilla 851: Documentation example syntax fix.
3436 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3438 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3441 NM/19 Bugzilla 745: TLS version reporting.
3442 Patch provided by Phil Pennock.
3444 NM/20 Bugzilla 167: bool: condition support.
3445 Patch provided by Phil Pennock.
3447 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3448 clients. Patch provided by Phil Pennock.
3450 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3451 Patch provided by Brad "anomie" Jorsch.
3453 NM/23 Bugzilla 687: Fix misparses in eximstats.
3454 Patch provided by Heiko Schlittermann.
3456 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3457 Patch provided by Heiko Schlittermann.
3459 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3460 plus update to original patch.
3462 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3464 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3465 Patch provided by David Brownlee.
3467 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3469 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3471 NM/30 Bugzilla 888: TLS documentation bugfixes.
3473 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3475 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3476 Unlike the original bugzilla I have changed all shell scripts in src tree.
3478 NM/33 Bugzilla 898: Transport filter timeout fix.
3479 Patch by Todd Rinaldo.
3481 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3482 Patch by Serge Demonchaux.
3484 NM/35 Bugzilla 39: Base64 decode bug fixes.
3485 Patch by Jakob Hirsch.
3487 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3489 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3491 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3493 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3499 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3500 ALT-N's libdkim that I have put here:
3501 http://duncanthrax.net/exim-experimental/
3503 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3504 were called 'true' and it seems that C99 defines that as a reserved
3505 keyword to be used with 'bool' variable types. That means you could
3506 not include C99-style headers which use bools without triggering
3507 build errors in sieve.c.
3509 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3510 as mailq or other aliases. Changed the --help handling significantly
3511 to do whats expected. exim_usage() emits usage/help information.
3513 SC/01 Added the -bylocaldomain option to eximstats.
3515 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3517 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3519 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3525 PH/01 Another patch from the Sieve maintainer.
3527 PH/02 When an IPv6 address is converted to a string for single-key lookup
3528 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3529 dots are used instead of colons so that keys in lsearch files need not
3530 contain colons. This was done some time before quoting was made available
3531 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3532 (notated using the quote facility) so as to distinguish them from IPv4
3533 keys. This meant that lookups for IP addresses in host lists did not work
3534 for iplsearch lookups.
3536 This has been fixed by arranging for IPv6 addresses to be expressed with
3537 colons if the lookup type is iplsearch. This is not incompatible, because
3538 previously such lookups could never work.
3540 The situation is now rather anomalous, since one *can* have colons in
3541 ordinary lsearch keys. However, making the change in all cases is
3542 incompatible and would probably break a number of configurations.
3544 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3547 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3548 conversion specification without a maximum field width, thereby enabling
3549 a rogue spamd server to cause a buffer overflow. While nobody in their
3550 right mind would setup Exim to query an untrusted spamd server, an
3551 attacker that gains access to a server running spamd could potentially
3552 exploit this vulnerability to run arbitrary code as the Exim user.
3554 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3555 $primary_hostname instead of what libspf2 thinks the hosts name is.
3557 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3558 a directory entry by the name of the lookup key. Previously, if a
3559 symlink pointed to a non-existing file or a file in a directory that
3560 Exim lacked permissions to read, a lookup for a key matching that
3561 symlink would fail. Now it is enough that a matching directory entry
3562 exists, symlink or not. (Bugzilla 503.)
3564 PH/03 The body_linecount and body_zerocount variables are now exported in the
3567 PH/04 Added the $dnslist_matched variable.
3569 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3570 This means they are set thereafter only if the connection becomes
3573 PH/06 Added the client_condition to authenticators so that some can be skipped
3574 by clients under certain conditions.
3576 PH/07 The error message for a badly-placed control=no_multiline_responses left
3577 "_responses" off the end of the name.
3579 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3581 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3582 (without spaces) instead of just copying the configuration text.
3584 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3586 PH/11 Added $max_received_linelength.
3588 PH/12 Added +ignore_defer and +include_defer to host lists.
3590 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3591 way in which PCRE > 7.0 is built.
3593 PH/14 Implemented queue_only_load_latch.
3595 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3596 MAIL command. The effect was to mangle the value on 64-bit systems.
3598 PH/16 Another patch from the Sieve maintainer.
3600 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3602 PH/18 If a system quota error occurred while trying to create the file for
3603 a maildir delivery, the message "Mailbox is full" was not appended to the
3604 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3605 only to a quota excession during the actual writing of the file.
3607 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3608 characters?) which causes problems in log lines. The DN values are now
3609 passed through string_printing() before being added to log lines.
3611 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3612 and InterBase are left for another time.)
3614 PH/21 Added message_body_newlines option.
3616 PH/22 Guard against possible overflow in moan_check_errorcopy().
3618 PH/23 POSIX allows open() to be a macro; guard against that.
3620 PH/24 If the recipient of an error message contained an @ in the local part
3621 (suitably quoted, of course), incorrect values were put in $domain and
3622 $local_part during the evaluation of errors_copy.
3628 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3629 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3632 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3633 issue a MAIL command.
3635 PH/02 In an ACL statement such as
3637 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3639 if a client was not listed at all, or was listed with a value other than
3640 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3641 the condition was not true (as it should be), so access was not denied.
3642 The bug was that the ! inversion was incorrectly passed on to the second
3643 item. This has been fixed.
3645 PH/03 Added additional dnslists conditions == and =& which are different from
3646 = and & when the dns lookup returns more than one IP address.
3648 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3649 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3651 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3652 FSYNC, which compiles an option called disable_fsync that allows for
3653 bypassing fsync(). The documentation is heavily laced with warnings.
3655 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3657 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3658 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3659 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3660 including adding "make clean"; (3) Added -fPIC when compiling the test
3661 dynamically loaded module, to get rid of a warning.
3663 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3664 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3665 = 0s. The bug is otherwise harmless.
3667 PH/07 There was a bug in the dovecot authenticator such that the value of
3668 $auth1 could be overwritten, and so not correctly preserved, after a
3669 successful authentication. This usually meant that the value preserved by
3670 the server_setid option was incorrect.
3672 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3674 PH/09 Installed PCRE release 7.0.
3676 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3677 run for batched SMTP input. It is now run at the start of every message
3678 in the batch. While fixing this I discovered that the process information
3679 (output by running exiwhat) was not always getting set for -bs and -bS
3680 input. This is fixed, and it now also says "batched" for BSMTP.
3682 PH/11 Added control=no_pipelining.
3684 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3685 patch, slightly modified), and move the expansion of helo_data till after
3686 the connection is made in the smtp transport (so it can use these
3689 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3691 PH/14 Added log_selector = +pid.
3693 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3695 PH/16 Add ${if forany and ${if forall.
3697 PH/17 Added dsn_from option to vary the From: line in DSNs.
3699 PH/18 Flush SMTP output before performing a callout, unless control =
3700 no_callout_flush is set.
3702 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3703 was true (the default) a successful delivery failed to delete the retry
3704 item, thus causing premature timeout of the address. The bug is now
3707 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3709 PH/21 Long custom messages for fakedefer and fakereject are now split up
3710 into multiline responses in the same way that messages for "deny" and
3711 other ACL rejections are.
3713 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3714 with slight modification.
3716 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3717 draft, changing the syntax and factoring some duplicate code".
3719 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3720 for deliveries of the second and subsequent messages over the same SMTP
3723 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3724 ${reduce, with only minor "tidies".
3726 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3728 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3729 expansion side effects.
3731 PH/27 When a message times out after an over-quota error from an Exim-imposed
3732 quota, the bounce message says "mailbox is full". This message was not
3733 being given when it was a system quota that was exceeded. It now should
3736 MH/03 Made $recipients available in local_scan(). local_scan() already has
3737 better access to the recipient list through recipients_list[], but
3738 $recipients can be useful in postmaster-provided expansion strings.
3740 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3741 in the case of a MAIL command with additional options following the
3742 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3743 were accidentally chopped off.
3745 PH/29 SMTP synchronization checks are implemented when a command is read -
3746 there is a check that no more input is waiting when there shouldn't be
3747 any. However, for some commands, a delay in an ACL can mean that it is
3748 some time before the response is written. In this time, more input might
3749 arrive, invalidly. So now there are extra checks after an ACL has run for
3750 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3751 pipelining has not been advertised.
3753 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3755 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3756 returned in the SMTP response when smtp_return_error_details was false.
3757 This has been fixed.
3759 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3760 instead of the C I/O that was originally supplied, because problems were
3761 reported on Solaris.
3763 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3764 Exim which did not show up earlier: it was assuming that a call to
3765 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3766 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3767 was a macro that became an assignment, so it seemed to work. This has
3768 changed to a proper function call with a void return, hence the compile
3769 error. Exim's code has been fixed.
3771 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3774 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3776 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3778 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3779 because it turns out to be too big or there is a timeout) there is no
3780 "Completed" line in the log. When some messages of this type were
3781 selected by exigrep, they were listed as "not completed". Others were
3782 picked up by some special patterns. I have improved the selection
3783 criteria to be more general.
3785 PH/38 The host_find_failed option in the manualroute router can now be set
3786 to "ignore", to completely ignore a host whose IP address cannot be
3787 found. If all hosts are ignored, the behaviour is controlled by the new
3788 host_all_ignored option.
3790 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3791 homing or because of multiple MX records with /mx) generated more than
3792 one IP address, and the following item turned out to be the local host,
3793 all the secondary addresses of the first item were incorrectly removed
3794 from the list, along with the local host and any following hosts (which
3795 is what is supposed to happen).
3797 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3798 whoever called Exim into the -H file. In the case of the daemon it was
3799 behaving confusingly. When first started, it used values for whoever
3800 started the daemon, but after a SIGHUP it used the Exim user (because it
3801 calls itself on a restart). I have changed the code so that it now always
3804 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3805 message are rejected with the same error (e.g. no authentication or bad
3806 sender address), and a DATA command is nevertheless sent (as can happen
3807 with PIPELINING or a stupid MUA), the error message that was given to the
3808 RCPT commands is included in the rejection of the DATA command. This is
3809 intended to be helpful for MUAs that show only the final error to their
3812 PH/42 Another patch from the Sieve maintainer.
3814 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3815 Eximstats - Fixed some broken HTML links and added missing column headers
3817 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3818 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3820 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3822 PH/43 Yet another patch from the Sieve maintainer.
3824 PH/44 I found a way to check for a TCP/IP connection going away before sending
3825 the response to the final '.' that terminates a message, but only in the
3826 case where the client has not sent further data following the '.'
3827 (unfortunately, this is allowed). However, in many cases there won't be
3828 any further data because there won't be any more messages to send. A call
3829 to select() can be used: if it shows that the input is "ready", there is
3830 either input waiting, or the socket has been closed. An attempt to read
3831 the next input character can distinguish the two cases. Previously, Exim
3832 would have sent an OK response which the client would never have see.
3833 This could lead to message repetition. This fix should cure that, at
3834 least in a lot of common cases.
3836 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3837 advertised in response to EHLO.
3843 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3844 fixed by 4.65/MH/01 (is this a record?) are fixed:
3846 (i) An empty string was always treated as zero by the numeric comparison
3847 operators. This behaviour has been restored.
3849 (ii) It is documented that the numeric comparison operators always treat
3850 their arguments as decimal numbers. This was broken in that numbers
3851 starting with 0 were being interpreted as octal.
3853 While fixing these problems I realized that there was another issue that
3854 hadn't been noticed. Values of message_size_limit (both the global option
3855 and the transport option) were treated as octal if they started with 0.
3856 The documentation was vague. These values are now always treated as
3857 decimal, and I will make that clear in the documentation.
3863 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3864 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3867 MH/01 Don't check that the operands of numeric comparison operators are
3868 integers when their expansion is in "skipping" mode (fixes bug
3869 introduced by 4.64-PH/07).
3871 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3872 child addresses, Exim now panics and dies. Previously, because the count
3873 is held in a short int, deliveries were likely to be lost. As such a
3874 large number of recipients for a single message is ridiculous
3875 (performance will be very, very poor), I have chosen to impose a limit
3876 rather than extend the field.
3882 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3883 leftover -K file (the existence of which was triggered by #402).
3884 While we were at it, introduced process PID as part of the -K
3885 filename. This should rule out race conditions when creating
3888 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3889 processing considerably. Previous code took too long for large mails,
3890 triggering a timeout which in turn triggers #401.
3892 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3893 in the DK code in transports.c. sendfile() is not really portable,
3894 hence the _LINUX specificness.
3896 TF/01 In the add_headers option to the mail command in an Exim filter,
3897 there was a bug that Exim would claim a syntax error in any
3898 header after the first one which had an odd number of characters
3901 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3902 callout verification, Exim cached a "reject" for the entire domain. This
3903 is correct for most verifications, but it is not correct for a recipient
3904 verification with use_sender or use_postmaster set, because in that case
3905 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3906 case of MAIL FROM:<> rejection from other early rejections (e.g.
3907 rejection of HELO). When verifying a recipient using a non-null MAIL
3908 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3909 Whatever the result of the callout, the value of the domain cache is
3910 left unchanged (for any other kind of callout, getting as far as trying
3911 RCPT means that the domain itself is ok).
3913 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3916 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3917 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3918 session, the daemon ignores ECONNECT errors and logs others; it now
3919 ignores EPIPE as well.
3921 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
3922 (quoted-printable decoding).
3924 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
3925 later the small subsequent patch to fix an introduced bug.
3927 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
3929 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
3931 PH/08 An error is now given if message_size_limit is specified negative.
3933 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
3934 to be given (somewhat) arbitrary names.
3936 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
3939 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
3940 miscellaneous code fixes
3942 PH/10 Added the log_reject_target ACL modifier to specify where to log
3945 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
3946 hostname. This is wrong, because it relates to the incoming message (and
3947 probably the interface on which it is arriving) and not to the outgoing
3948 callout (which could be using a different interface). This has been
3949 changed to use the value of the helo_data option from the smtp transport
3950 instead - this is what is used when a message is actually being sent. If
3951 there is no remote transport (possible with a router that sets up host
3952 addresses), $smtp_active_hostname is used.
3954 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
3955 tweaks were necessary in order to get it to work (see also 21 below):
3956 (a) The code assumed that strncpy() returns a negative number on buffer
3957 overflow, which isn't the case. Replaced with Exim's string_format()
3959 (b) There were several signed/unsigned issues. I just did the minimum
3960 hacking in of casts. There is scope for a larger refactoring.
3961 (c) The code used strcasecmp() which is not a standard C function.
3962 Replaced with Exim's strcmpic() function.
3963 (d) The code set only $1; it now sets $auth1 as well.
3964 (e) A simple test gave the error "authentication client didn't specify
3965 service in request". It would seem that Dovecot has changed its
3966 interface. Fortunately there's a specification; I followed it and
3967 changed what the client sends and it appears to be working now.
3969 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
3972 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
3973 address A is aliased to B and C, where B exists and C does not. Without
3974 -v the output is "A verified" because verification stops after a
3975 successful redirection if more than one address is generated. However,
3976 with -v the child addresses are also verified. Exim was outputting "A
3977 failed to verify" and then showing the successful verification for C,
3978 with its parentage. It now outputs "B failed to verify", showing B's
3979 parentage before showing the successful verification of C.
3981 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
3982 look up a TXT record in a specific list after matching in a combined
3985 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
3986 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
3987 they consult the DNS. I had assumed they would set it the way they
3988 wanted; and indeed my experiments on Linux seem to show that in some
3989 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
3990 To be on the safe side, however, I have now made the interface to
3991 host_find_byname() similar to host_find_bydns(), with an argument
3992 containing the DNS resolver options. The host_find_byname() function now
3993 sets these options at its start, just as host_find_bydns() does. The smtp
3994 transport options dns_qualify_single and dns_search_parents are passed to
3995 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
3996 of host_find_byname() use the default settings of RES_DEFNAMES
3997 (qualify_single) but not RES_DNSRCH (search_parents).
3999 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4000 spool_read_header() do less string testing, by means of a preliminary
4001 switch on the second character of optional "-foo" lines. (This is
4002 overdue, caused by the large number of possibilities that now exist.
4003 Originally there were few.) While I was there, I also converted the
4004 str(n)cmp tests so they don't re-test the leading "-" and the first
4005 character, in the hope this might squeeze out yet more improvement.
4007 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4008 flag allowing group syntax was set by the header_syntax check but not
4009 turned off, possible causing trouble later; (2) The flag was not being
4010 set at all for the header_verify test, causing "group"-style headers to
4011 be rejected. I have now set it in this case, and also caused header_
4012 verify to ignore an empty address taken from a group. While doing this, I
4013 came across some other cases where the code for allowing group syntax
4014 while scanning a header line wasn't quite right (mostly, not resetting
4015 the flag correctly in the right place). These bugs could have caused
4016 trouble for malformed header lines. I hope it is now all correct.
4018 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4019 with the "reply" argument non-NULL. The code, however (which originally
4020 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4021 but it didn't always do it. This confused somebody who was copying the
4022 code for some other use. I have removed all the tests.
4024 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4025 feature that was used to support insecure browsers during the U.S. crypto
4026 embargo. It requires special client support, and Exim is probably the
4027 only MTA that supported it -- and would never use it because real RSA is
4028 always available. This code has been removed, because it had the bad
4029 effect of slowing Exim down by computing (never used) parameters for the
4030 RSA_EXPORT functionality.
4032 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4033 authenticator to fail if there's a tab character in the incoming data
4034 (there should never be unless someone is messing about, as it's supposed
4035 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4036 if the connection is using TLS or if the remote IP is the same as the
4037 local IP, and the "valid-client-cert option" if a client certificate has
4040 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4041 authenticators. This can be used for authorization after authentication
4042 succeeds. (In the case of plaintext, it servers for both authentication
4045 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4046 if any retry times were supplied.
4048 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4049 connection, where there is no client IP address to check. In this
4050 situation, the verify now always succeeds.
4052 PH/25 Applied John Jetmore's -Mset patch.
4054 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4056 PH/27 In a string expansion for a processed (not raw) header when multiple
4057 headers of the same name were present, leading whitespace was being
4058 removed from all of them, but trailing whitespace was being removed only
4059 from the last one. Now trailing whitespace is removed from each header
4060 before concatenation. Completely empty headers in a concatenation (as
4061 before) are ignored.
4063 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4064 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4066 PH/29 [Removed. This was a change that I later backed out, and forgot to
4067 correct the ChangeLog entry (that I had efficiently created) before
4068 committing the later change.]
4070 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4071 address errors (4xx response to RCPT) over the same connection as other
4072 messages routed to the same hosts. Such deliveries are always "forced",
4073 so retry times are not inspected. This resulted in far too many retries
4074 for the affected addresses. The effect occurred only when there were more
4075 hosts than the hosts_max_try setting in the smtp transport when it had
4076 the 4xx errors. Those hosts that it had tried were not added to the list
4077 of hosts for which the message was waiting, so if all were tried, there
4078 was no problem. Two fixes have been applied:
4080 (i) If there are any address or message errors in an SMTP delivery, none
4081 of the hosts (tried or untried) are now added to the list of hosts
4082 for which the message is waiting, so the message should not be a
4083 candidate for sending over the same connection that was used for a
4084 successful delivery of some other message. This seems entirely
4085 reasonable: after all the message is NOT "waiting for some host".
4086 This is so "obvious" that I'm not sure why it wasn't done
4087 previously. Hope I haven't missed anything, but it can't do any
4088 harm, as the worst effect is to miss an optimization.
4090 (ii) If, despite (i), such a delivery is accidentally attempted, the
4091 routing retry time is respected, so at least it doesn't keep
4092 hammering the server.
4094 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4095 in ${readsocket because some servers need this prod.
4097 PH/32 Added some extra debug output when updating a wait-xxx database.
4099 PH/33 The hint "could be header name not terminated by colon", which has been
4100 given for certain expansion errors for a long time, was not being given
4101 for the ${if def:h_colon_omitted{... case.
4103 PH/34 The spec says: "With one important exception, whenever a domain list is
4104 being scanned, $domain contains the subject domain." There was at least
4105 one case where this was not true.
4107 PH/35 The error "getsockname() failed: connection reset by peer" was being
4108 written to the panic log as well as the main log, but it isn't really
4109 panic-worthy as it just means the connection died rather early on. I have
4110 removed the panic log writing for the ECONNRESET error when getsockname()
4113 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4114 runs only) independently of the message's sender address. This meant
4115 that, if the 4xx error was in fact related to the sender, a different
4116 message to the same recipient with a different sender could confuse
4117 things. In particular, this can happen when sending to a greylisting
4118 server, but other circumstances could also provoke similar problems.
4119 I have changed the default so that the retry time for these errors is now
4120 based a combination of the sender and recipient addresses. This change
4121 can be overridden by setting address_retry_include_sender=false in the
4124 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4125 remote server are returned as part of bounce messages. This was not
4126 happening for LMTP over a pipe (the lmtp transport), but now it is the
4127 same for both kinds of LMTP.
4129 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4130 in header lines that were in fact CNAMEs. This is no longer the case.
4132 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4133 and queue runs started by the daemon processed all messages. This has
4134 been fixed so that -R and -S can now usefully be given with -q<time>.
4136 PH/40 Import PCRE release 6.7 (fixes some bugs).
4138 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4140 PH/42 Give an error if -q is specified more than once.
4142 PH/43 Renamed the variables $interface_address and $interface_port as
4143 $received_ip_address and $received_port, to make it clear that these
4144 values apply to message reception, and not to the outgoing interface when
4145 a message is delivered. (The old names remain recognized, of course.)
4147 PH/44 There was no timeout on the connect() call when using a Unix domain
4148 socket in the ${readsocket expansion. There now is.
4150 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4151 be meaningful with "accept".
4153 SC/01 Eximstats V1.43
4154 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4156 SC/02 Eximstats V1.44
4157 Use a glob alias rather than an array ref in the generated
4158 parser. This improves both readability and performance.
4160 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4161 Collect SpamAssassin and rejection statistics.
4162 Don't display local sender or destination tables unless
4163 there is data to show.
4164 Added average volumes into the top table text output.
4166 SC/04 Eximstats V1.46
4167 Collect data on the number of addresses (recipients)
4168 as well as the number of messages.
4170 SC/05 Eximstats V1.47
4171 Added 'Message too big' to the list of mail rejection
4172 reasons (thanks to Marco Gaiarin).
4174 SC/06 Eximstats V1.48
4175 Mainlog lines which have GMT offsets and are too short to
4176 have a flag are now skipped.
4178 SC/07 Eximstats V1.49 (Alain Williams)
4179 Added the -emptyok flag.
4181 SC/08 Eximstats V1.50
4182 Fixes for obtaining the IP address from reject messages.
4184 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4185 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4186 whitespace changes from 4.64-PH/27
4188 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4191 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4192 are found, allow negative numbers in numeric criteria)
4194 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4196 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4197 to match changes made in 4.64-PH/43
4199 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4201 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4202 "domain" is actually an IP address, and force a failure. This locks out
4203 those revolvers/nameservers that support "A-for-A" lookups, in
4204 contravention of the specifications.
4206 PH/48 When a host name was looked up from an IP address, and the subsequent
4207 forward lookup of the name timed out, the host name was left in
4208 $sender_host_name, contrary to the specification.
4210 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4211 restricted to single-key lookups, Exim was not diagnosing an error if
4212 * or *@ was used with a query-style lookup.
4214 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4216 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4217 long ago, but noone interested enough thought of it. Let's just say that
4218 the "1.1" means that there are some new functions that weren't there at
4219 some point in the past.
4221 PH/51 Error processing for expansion failure of helo_data from an smtp
4222 transport during callout processing was broken.
4224 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4225 tested/used via the -bh/-bhc/-bs options.
4227 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4228 bug, fixed in subsequent PCRE releases).
4230 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4231 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4233 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4239 SC/01 Use a glob alias rather than an array ref in eximstats generated
4240 parser. This improves both readability and performance.
4242 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4243 Don't display local sender or destination tables in eximstats unless
4244 there is data to show.
4245 Added average volumes into the eximstats top table text output.
4247 SC/03 Collect data on the number of addresses (recipients) as well
4248 as the number of messages in eximstats.
4250 TF/01 Correct an error in the documentation for the redirect router. Exim
4251 does (usually) call initgroups() when daemonizing.
4253 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4254 with consistent privilege compared to when running as a daemon.
4256 TF/03 Note in the spec that $authenticated_id is not set for local
4257 submissions from trusted users.
4259 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4260 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4262 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4263 by adding some example configuration directives to the default
4264 configuration file. A little bit of work is required to uncomment the
4265 directives and define how usernames and passwords are checked, but
4266 there is now a framework to start from.
4268 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4269 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4270 without this. I don't know how relevant this is to other LDAP libraries.
4272 PH/02 Add the verb name to the "unknown ACL verb" error.
4274 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4276 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4278 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4279 directory not expanded when it should be if an expanded home directory
4280 was set for the address (which is overridden by the transport).
4282 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4285 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4286 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4287 because it is too late at that time, and has no effect.
4289 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4290 security issue with \' (bugzilla #107). I could not use the
4291 PQescapeStringConn() function, because it needs a PGconn value as one of
4294 PH/08 When testing addresses using -bt, indicate those final addresses that
4295 are duplicates that would not cause an additional delivery. At least one
4296 person was confused, thinking that -bt output corresponded to deliveries.
4297 (Suppressing duplicates isn't a good idea as you lose the information
4298 about possibly different redirections that led to the duplicates.)
4300 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4301 systems where poll() doesn't work, in particular OS X.
4303 PH/10 Added more information to debugging output for retry time not reached.
4305 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4306 operations in malware.c.
4308 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4311 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4312 syslog_duplication set false, log lines that would normally be written
4313 both the the main log and to the reject log were not written to syslog at
4316 PH/14 In the default configuration, change the use of "message" in ACL warn
4317 statements to "add_header".
4319 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4320 not followed by a command (e.g. "seen endif").
4322 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4323 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4326 PH/17 Added extra conditions to the default value of delay_warning_condition
4330 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4331 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4332 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4335 The Auto-Submitted: and various List- headers are standardised, whereas I
4336 don't think Precedence: ever was.
4338 PH/18 Refactored debugging code in route_finduser() to show more information,
4339 in particular, the error code if getpwnam() issues one.
4341 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4342 This is apparently needed in addition to the PH/07 change above to avoid
4343 any possible encoding problems.
4345 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4346 but not after initializing Perl.
4348 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4349 output them only if debugging. By default they are written stderr,
4350 apparently, which is not desirable.
4352 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4355 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4358 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4360 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4361 authenticated or an ident call has been made. Suppress the default
4362 values for $authenticated_id and $authenticated_sender (but permit -oMai
4363 and -oMas) when testing with -bh.
4365 PH/24 Re-jigged the order of the tests in the default configuration so that the
4366 tests for valid domains and recipients precede the DNS black list and CSA
4367 tests, on the grounds that those ones are more expensive.
4369 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4370 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4371 This bug exists in every version of Exim that I still have, right back to
4374 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4375 However, an attempt to turn on case-sensitivity in a regex key by
4376 including (?-i) didn't work because the subject string was already
4377 lowercased, and the effects were non-intuitive. It turns out that a
4378 one-line patch can be used to allow (?-i) to work as expected.
4384 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4385 other effects) broke the use of negated acl sub-conditions.
4387 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4390 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4391 "Deny" causes Exim to reject the incoming connection with a 554 error.
4392 Unfortunately, if there is a major crisis, such as a disk failure,
4393 tcp-wrappers gives "deny", whereas what one would like would be some
4394 kind of temporary error. A kludge has been added to help with this.
4395 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4396 554 error is used if errno is still zero or contains ENOENT (which occurs
4397 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4400 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4402 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4403 errors. Otherwise a message that provokes a temporary error (when other
4404 messages do not) can cause a whole host to time out.
4406 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4407 addresses were routed directly to files or pipes from a redirect router.
4408 File deliveries just didn't batch; pipe deliveries might have suffered
4411 PH/06 A failure to get a lock for a hints database would erroneously always say
4412 "Failed to get write lock", even when it was really a read lock.
4414 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4415 of 0600. This has been changed to use the value of the lockfile_mode
4416 option (which defaults to 0600).
4418 PH/08 Applied small patch from the Sieve maintainer.
4420 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4421 folder from quota calculations, a direct delivery into this folder messed
4422 up the contents of the maildirsize file. This was because the regex was
4423 used only to exclude .Trash (or whatever) when the size of the mailbox
4424 was calculated. There was no check that a delivery was happening into an
4425 excluded directory. This bug has been fixed by ignoring all quota
4426 processing for deliveries into excluded directories.
4428 PH/10 Added the maildirfolder_create_regex option to appendfile.
4434 PH/01 The code for finding all the local interface addresses on a FreeBSD
4435 system running IPv6 was broken. This may well have applied to all BSD
4436 systems, as well as to others that have similar system calls. The broken
4437 code found IPv4 interfaces correctly, but gave incorrect values for the
4438 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4439 that it would not match correctly against @[] and not recognize the IPv6
4442 PH/02 The ipliteral router was not recognizing addresses of the form user@
4443 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4445 PH/03 Added disable_ipv6.
4447 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4448 decoded form, because it is most often used to construct To: headers
4449 lines in autoreplies, and the decoded form may well be syntactically
4450 invalid. However, $reply_address has leading white space removed, and all
4451 newlines turned into spaces so that the autoreply transport does not
4454 PH/05 If group was specified without a user on a router, and no group or user
4455 was specified on a transport, the group from the router was ignored.
4457 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4458 for visible compile-time settings that can be used to change these
4459 numbers, for those that want even more. Backwards compatibility with old
4460 spool files has been maintained. However, going back to a previous Exim
4461 release will lost any variables that are in spool files.
4463 PH/07 Two small changes when running in the test harness: increase delay when
4464 passing a TCP/IP connection to a new process, in case the original
4465 process has to generate a bounce, and remove special handling of
4466 127.0.0.2 (sic), which is no longer necessary.
4468 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4469 be the same on different OS.
4471 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4474 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4475 whether --show-vars was specified or not
4477 JJ/02 exipick: Added support for new ACL variable spool format introduced
4480 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4481 syntactically invalid From: or Reply-to: line, and a filter used this to
4482 generate an autoreply, and therefore failed to obtain an address for the
4483 autoreply, Exim could try to deliver to a non-existent relative file
4484 name, causing unrelated and misleading errors. What now happens is that
4485 it logs this as a hard delivery error, but does not attempt to create a
4488 PH/11 The exinext utility has a -C option for testing purposes, but although
4489 the given file was scanned by exinext itself; it wasn't being passed on
4490 when Exim was called.
4492 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4493 an end-of-file indication when reading a command response.
4495 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4496 compiled. In many other places in Exim, IPv6 addresses are always
4497 recognized, so I have changed this. It also means that IPv4 domain
4498 literals of the form [IPV4:n.n.n.n] are now always recognized.
4500 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4501 used if the router is not running as root, for example, when verifying at
4502 ACL time, or when using -bh. The debugging output from this situation was
4503 non-existent - all you got was a failure to exec. I have made two
4506 (a) Failures to set uid/gid, the current directory, or a process leader
4507 in a subprocess such as that created by queryprogram now generate
4508 suitable debugging output when -d is set.
4510 (b) The queryprogram router detects when it is not running as root,
4511 outputs suitable debugging information if -d is set, and then runs
4512 the subprocess without attempting to change uid/gid.
4514 PH/15 Minor change to Makefile for building test_host (undocumented testing
4517 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4518 additional section of a DNS packet that returns MX or SRV records.
4519 Instead, it always explicitly searches for A/AAAA records. This avoids
4520 major problems that occur when a DNS server includes only records of one
4521 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4522 fixed another bug: if SRV records were looked up and the corresponding
4523 address records were *not* found in the additional section, the port
4524 values from the SRV records were lost.
4526 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4527 using the correct key (the original address) when searching the retry
4528 rules in order to find which one to use for generating the retry hint.
4530 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4531 adding the default one. Similarly, if it contains a Reply-To: header, the
4532 errors_reply_to option, if set, is not used.
4534 PH/19 When calculating a retry time, Exim used to measure the "time since
4535 failure" by looking at the "first failed" field in the retry record. Now
4536 it does not use this if it is later than than the arrival time of the
4537 message. Instead it uses the arrival time. This makes for better
4538 behaviour in cases where some deliveries succeed, thus re-setting the
4539 "first failed" field. An example is a quota failure for a huge message
4540 when small messages continue to be delivered. Without this change, the
4541 "time since failure" will always be short, possible causing more frequent
4542 delivery attempts for the huge message than are intended.
4543 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4545 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4546 $1, $2, $3) because the numerical variables can be reset during some
4547 expansion items (e.g. "match"), thereby losing the authentication data.
4549 PH/21 Make -bV show the size of off_t variables so that the test suite can
4550 decide whether to run tests for quotas > 2G.
4552 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4553 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4554 filecount value is greater than 2G or if a quota value is greater than 2G
4555 on a system where the size of off_t is not greater than 4, a panic error
4558 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4559 never match. The debug and -bh output now contains an explicit error
4560 message indicating a malformed IPv4 address or mask.
4562 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4563 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4564 PH/23 above applies.
4566 PH/25 Do not write to syslog when running in the test harness. The only
4567 occasion when this arises is a failure to open the main or panic logs
4568 (for which there is an explicit test).
4570 PH/26 Added the /no_tell option to "control=freeze".
4572 PH/27 If a host name lookup failed very early in a connection, for example, if
4573 the IP address matched host_lookup and the reverse lookup yielded a name
4574 that did not have a forward lookup, an error message of the form "no IP
4575 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4576 could be logged. Now it outputs the IP address instead of "NULL".
4578 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4579 allows the sender and the authenticated sender to be set when
4580 submitting a message from within Exim. Since child_open_exim() is
4581 documented for local_scan(), the new function should be too.
4583 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4584 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4585 results in an empty string is now treated as unset.
4587 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4589 PH/31 Added sender_verify_fail logging option.
4591 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4592 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4593 tidied the source and removed it altogether.
4595 PH/33 When a queue run was abandoned because the load average was too high, a
4596 log line was always written; now it is written only if the queue_run log
4597 selector is set. In addition, the log line for abandonment now contains
4598 information about the queue run such as the pid. This is always present
4599 in "start" and "stop" lines but was omitted from the "abandon" line.
4601 PH/34 Omit spaces between a header name and the colon in the error message that
4602 is given when verify = headers_syntax fails (if there are lots of them,
4603 the message gets confusing).
4605 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4606 names, as there are now some PTR records that contain slashes. This check
4607 is only to protect against broken name servers that fall over on strange
4608 characters, so the fact that it applies to all lookups doesn't matter.
4610 PH/36 Now that the new test suite is complete, we can remove some of the
4611 special code in Exim that was needed for the old test suite. For example,
4612 sorting DNS records because real resolvers return them in an arbitrary
4613 order. The new test suite's fake resolver always returns records in the
4616 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4617 bounces) except when queue_only is set, to avoid logging races between
4618 the different processes.
4620 PH/38 Panic-die if .include specifies a non-absolute path.
4622 PH/39 A tweak to the "H" retry rule from its user.
4624 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4625 a label. They prevented compilation on older perls.
4627 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4628 a warning to be raised on newish perls.
4630 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4631 on queue. Changes to match documented behaviour of showing count of
4632 messages matching specified criteria.
4634 PH/40 Changed the default ident timeout from 30s to 5s.
4636 PH/41 Added support for the use of login_cap features, on those BSD systems
4637 that have them, for controlling the resources used by pipe deliveries.
4639 PH/42 The content-scanning code uses fopen() to create files in which to put
4640 message data. Previously it was not paying any attention to the mode of
4641 the files. Exim runs with umask(0) because the rest of the code creates
4642 files with open(), and sets the required mode explicitly. Thus, these
4643 files were ending up world-writeable. This was not a big issue, because,
4644 being within the spool directory, they were not world-accessible. I have
4645 created a function called modefopen, which takes an additional mode
4646 argument. It sets umask(777), creates the file, chmods it to the required
4647 mode, then resets the umask. All the relevant calls to fopen() in the
4648 content scanning code have been changed to use this function.
4650 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4651 to 24 hours. This avoids potential overflow problems when processing G
4652 and H retry rules. I suspect nobody ever tinkers with this value.
4654 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4656 PH/45 When the plaintext authenticator is running as a client, the server's
4657 challenges are checked to ensure they are valid base64 strings. By
4658 default, the authentication attempt is cancelled if an invalid string is
4659 received. Setting client_ignore_invalid_base64 true ignores these errors.
4660 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4661 they are received. Thus, the responses can be made to depend on the
4662 challenges. If an invalid string is ignored, an empty string is placed in
4665 PH/46 Messages that are created by the autoreply transport now contains a
4666 References: header, in accordance with RFCs 2822 and 3834.
4668 PH/47 Added authenticated_sender_force to the smtp transport.
4670 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4672 PH/49 Installed latest patch from the Sieve maintainer.
4674 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4675 was also set, the appendfile transport was unnecessarily scanning a
4676 directory of message files (e.g. for maildir delivery) to find the count
4677 of files (along with the size), even though it did not need this
4678 information. It now does the scan only if it needs to find either the
4679 size of the count of files.
4681 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4683 PH/52 Two bugs concerned with error handling when the smtp transport is
4686 (i) Exim was not creating retry information for temporary errors given
4687 for individual recipients after the DATA command when the smtp transport
4688 was used in LMTP mode. This meant that they could be retried too
4689 frequently, and not timed out correctly.
4691 (ii) Exim was setting the flag that allows error details to be returned
4692 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4693 recipients that were returned after the DATA command.
4695 PH/53 This is related to PH/52, but is more general: for any failing address,
4696 when detailed error information was permitted to be returned to the
4697 sender, but the error was temporary, then after the final timeout, only
4698 "retry timeout exceeded" was returned. Now it returns the full error as
4699 well as "retry timeout exceeded".
4701 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4702 do this, and (what is worse) MTAs that accept it.
4704 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4705 will now be deprecated.
4707 PH/56 New os.c-cygwin from the Cygwin maintainer.
4709 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4710 formats (previously only available in exim formats via -bpr, -bpru,
4711 and -bpra. Now also available in native and exiqgrep formats)
4713 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4714 with very large, slow to parse queues
4716 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4718 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4720 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4721 responses to authentication challenges, though it was showing the
4722 challenges; (ii) I've removed the CR characters from the debug output for
4725 PH/58 Allow for the insertion of a newline as well as a space when a string
4726 is turned into more than one encoded-word during RFC 2047 encoding. The
4727 Sieve code now uses this.
4729 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4730 data_4xx, lost_connection, tls_required.
4732 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4733 message was being sent as an SMTP response.
4735 PH/61 Add -l and -k options to exicyclog.
4737 PH/62 When verifying, if an address was redirected to one new address, so that
4738 verification continued, and the new address failed or deferred after
4739 having set something in $address_data, the value of $address_data was not
4740 passed back to the ACL. This was different to the case when no
4741 redirection occurred. The value is now passed back in both cases.
4743 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4744 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4745 use login_cap.h, so on its own it isn't the distinguishing feature. The
4746 new name refers directly to the setclassresources() function.
4748 PH/65 Added configuration files for NetBSD3.
4750 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4752 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4753 is preferred over IPv4.
4755 PH/68 The bounce_return_message and bounce_return_body options were not being
4756 honoured for bounces generated during the reception of non-SMTP messages.
4757 In particular, this applied to messages rejected by the ACL. This bug has
4758 been fixed. However, if bounce_return_message is true and bounce_return_
4759 body is false, the headers that are returned for a non-SMTP message
4760 include only those that have been read before the error was detected.
4761 (In the case of an ACL rejection, they have all been read.)
4763 PH/69 The HTML version of the specification is now built in a directory called
4764 spec_html instead of spec.html, because the latter looks like a path with
4765 a MIME-type, and this confuses some software.
4767 PH/70 Catch two compiler warnings in sieve.c.
4769 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4770 function verify_get_ident() calls ip_connect() to connect a socket, but
4771 if the "connect()" function timed out, ip_connect() used to close the
4772 socket. However, verify_get_ident() also closes the socket later, and in
4773 between Exim writes to the log, which may get opened at this point. When
4774 the socket was closed in ip_connect(), the log could get the same file
4775 descriptor number as the socket. This naturally causes chaos. The fix is
4776 not to close the socket in ip_connect(); the socket should be closed by
4777 the function that creates it. There was only one place in the code where
4778 this was missing, in the iplookup router, which I don't think anybody now
4779 uses, but I've fixed it anyway.
4781 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4782 well as to direct DNS lookups. Otherwise the handling of names in host
4783 lists is inconsistent and therefore confusing.
4789 PH/01 Two changes to the default runtime configuration:
4791 (1) Move the checks for relay_from_hosts and authenticated clients from
4792 after to before the (commented out) DNS black list checks.
4794 (2) Add control=submission to the relay_from_hosts and authenticated
4795 clients checks, on the grounds that messages accepted by these
4796 statements are most likely to be submissions.
4798 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4800 (1) Generate an error if the third argument for the ${prvs expansion is
4803 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4806 (3) Reset the variables that are obtained from the first argument of
4807 ${prvscheck and used in the second argument before leaving the code,
4808 because their memory is reclaimed, so using them afterwards may do
4811 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4812 one (it's much easier than Tom thought :-).
4814 (5) Because of (4), we can now allow for the use of $prvscheck_result
4815 inside the third argument.
4817 PH/03 For some reason, the default setting of PATH when running a command from
4818 a pipe transport was just "/usr/bin". I have changed it to
4821 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4822 anything to be listed in the output from -bV.
4824 PH/05 When a filter generated an autoreply, the entire To: header line was
4825 quoted in the delivery log line, like this:
4827 => >A.N.Other <ano@some.domain> <original@ddress> ...
4829 This has been changed so that it extracts the operative address. There
4830 may be more than one such address. If so, they are comma-separated, like
4833 => >ano@some.domain,ona@other.domain <original@ddress> ...
4835 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4836 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4837 not being looked up in the rDNS to get a host name, Exim was showing the
4838 IP address twice in Received: lines, even though the IP addresses were
4839 identical. For example:
4841 Received: from [1.2.3.4] (helo=[1.2.3.4])
4843 However, if the real host name was known, it was omitting the HELO data
4844 if it matched the actual IP address. This has been tidied up so that it
4845 doesn't show the same IP address twice.
4847 PH/07 When both +timestamp and +memory debugging was on, the value given by
4848 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4849 called by the debug printing, thereby overwriting the timestamp buffer.
4850 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4852 PH/08 When the original message was included in an autoreply transport, it
4853 always said "this is a copy of the message, including all the headers",
4854 even if body_only or headers_only was set. It now gives an appropriate
4857 PH/09 Applied a patch from the Sieve maintainer which:
4859 o fixes some comments
4860 o adds the (disabled) notify extension core
4861 o adds some debug output for the result of if/elsif tests
4862 o points to the current vacation draft in the documentation
4863 and documents the missing references header update
4867 o fixes a bug in processing the envelope test (when testing
4868 multiple envelope elements, the last element determined the
4871 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4872 Electronic Mail") by including:
4874 Auto-submitted: auto-generated
4876 in the messages that it generates (bounce messages and others, such as
4877 warnings). In the case of bounce messages for non-SMTP messages, there was
4878 also a typo: it was using "Auto_submitted" (underscore instead of
4879 hyphen). Since every message generated by Exim is necessarily in response
4880 to another message, thes have all been changed to:
4882 Auto-Submitted: auto-replied
4884 in accordance with these statements in the RFC:
4886 The auto-replied keyword:
4888 - SHOULD be used on messages sent in direct response to another
4889 message by an automatic process,
4891 - MUST NOT be used on manually-generated messages,
4893 - MAY be used on Delivery Status Notifications (DSNs) and Message
4894 Disposition Notifications (MDNs),
4896 - MUST NOT be used on messages generated by automatic or periodic
4897 processes, except for messages which are automatic responses to
4900 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4901 to the default Received: header definition.
4903 PH/12 Added log selector acl_warn_skipped (default on).
4905 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4906 variables because (a) they are in the wrong storage pool and (b) even if
4907 they were copied, it wouldn't work properly because of the caching.
4909 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4910 checking when decoding. Apparently there are clients that generate
4911 overlong encoded strings. Why am I not surprised?
4913 PH/15 If the first argument of "${if match_address" was not empty, but did not
4914 contain an "@" character, Exim crashed. Now it writes a panic log message
4915 and treats the condition as false.
4917 PH/16 In autoreply, treat an empty string for "once" the same as unset.
4919 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
4920 extension "envelope-auth". The code is finished and in agreement with
4921 other implementations, but there is no documentation so far and in fact,
4922 nobody wrote the draft yet. This extension is currently #undef'ed, thus
4923 not changing the active code.
4925 Print executed "if" and "elsif" statements when debugging is used. This
4926 helps a great deal to understand what a filter does.
4928 Document more things not specified clearly in RFC3028. I had all this
4929 sorted out, when out of a sudden new issues came to my mind. Oops."
4931 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
4934 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
4935 canonical form (as documented). However, after a host name lookup from
4936 the IP address, check_host() was doing a simple string comparison with
4937 addresses acquired from the DNS when checking that the found name did
4938 have the original IP as one of its addresses. Since any found IPv6
4939 addresses are likely to be in abbreviated form, the comparison could
4940 fail. Luckily, there already exists a function for doing the comparison
4941 by converting both addresses to binary, so now that is used instead of
4942 the text comparison.
4944 PH/20 There was another similar case to PH/19, when a complete host name was
4945 given in a host list; looking up its IP address could give an abbreviated
4946 form, whereas the current host's name might or might not be abbreviated.
4947 The same fix has been applied.
4953 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
4954 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
4957 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
4958 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
4960 PH/03 Typo: missing ".o" in src/pcre/Makefile.
4962 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
4963 header line, restrict the check to what is listed in RFCs 2369 and 2929.
4964 Also, for "Auto-Submitted", treat anything other than "no" as
4965 non-personal, in accordance with RFC 3834. (Previously it treated
4966 anything starting "auto-" as non-personal.)
4968 TF/01 The control=submission/name=... option had a problem with syntax
4969 errors if the name included a slash character. The /name= option
4970 now slurps the rest of the string, so it can include any characters
4971 but it must come last in the list of options (after /sender_retain
4974 PH/05 Some modifications to the interface to the fake nameserver for the new
4982 TK/01 Added the "success_on_redirect" address verification option. See
4983 NewStuff for rationale and an example.
4985 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
4987 PH/02 Patch to exigrep to allow it to work on syslog lines.
4989 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
4990 fread() to skip over the body file's header line, because in Cygwin the
4991 header line is locked and is inaccessible.
4993 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
4994 co-exist for some time) to make it clear that it is the Exim ID that is
4995 referenced, not the Message-ID: header line.
4997 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
4998 string_format() function, because snprintf() does not exist on all
5001 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5002 ${stat: expansion item.
5004 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5005 protocol synchronization error", to keep the pedants happy.
5007 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5008 well as for IRIX systems, when gcc is being used. See the host.c source
5011 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5013 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5016 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5017 transport and to the smtp transport in LMTP mode.
5019 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5021 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5022 run a filter in a subprocess. This could lead to confusion in subsequent
5023 lookups in the parent process. There should also be a search_tidyup() at
5024 the end of the subprocess.
5026 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5027 only if the host matched helo_try_verify_hosts, which caused the
5028 verification to occur when the EHLO/HELO command was issued. The ACL just
5029 tested the remembered result. Now, if a previous verification attempt has
5030 not happened, "verify = helo" does it there and then.
5032 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5034 TK/03 Fix log output including CR from clamd.
5036 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5037 exist provoked a memory error which could cause a segfault.
5039 PH/15 Installed PCRE 6.2
5041 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5043 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5044 of the problem. Specifically, suggested +O2 rather than +O1 for the
5047 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5049 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5050 record, the port was not correctly logged when the outgoing_port log
5051 selector was set (it logged the transort's default port).
5053 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5054 fallback_hosts, and "hosts" in the smtp transport.
5056 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5057 host errors such as "Connection refused".
5059 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5060 authentication with radiusclient 0.4.9:
5062 - Error returned from rc_read_config was caught wrongly
5063 - Username/password not passed on to radius server due to wrong length.
5065 The presumption is that some radiusclient API changes for 4.51/PH/17
5066 were not taken care of correctly. The code is still untested by me (my
5067 Linux distribution still has 0.3.2 of radiusclient), but it was
5068 contributed by a Radius user.
5070 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5071 expanding the "port" option of the smtp transport.
5073 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5074 while reading a MIME header. Thanks to Tom Hughes for a patch.
5076 PH/24 Include config.h inside local_scan.h so that configuration settings are
5079 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5080 that in an ACL for RCPT (for example), you can examine exactly what was
5083 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5084 commands, but it was not correctly comparing the address with the actual
5085 client host address. Thus, it would show the EHLO address in Received:
5086 header lines when this was not necessary.
5088 PH/27 Added the % operator to ${eval:}.
5090 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5091 it should be ignoring failures (because with -C, for example, it has lost
5092 privilege). It wasn't ignoring creation failures other than "already
5095 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5098 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5099 because an input line was too long, either on its own, or by virtue of
5100 too many continuations, the temporary file was not being removed, and the
5101 return code was incorrect.
5103 PH/31 Missing "BOOL" in function definition in filtertest.c.
5105 PH/32 Applied Sieve patches from the maintainer.
5107 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5109 PH/33 Added "verify = not_blind".
5111 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5112 Local/Makefile (with some defaults set). These are used in built scripts
5113 such as exicyclog, but they have never been used in the exim_install
5114 script (though there are many overriding facilities there). I have
5115 arranged that the exim_install script now takes note of these two
5118 PH/35 Installed configuration files for Dragonfly.
5120 PH/36 When a locally submitted message by a trusted user did not contain a
5121 From: header, and the sender address was obtained from -f or from an SMTP
5122 MAIL command, and the trusted user did not use -F to supply a sender
5123 name, $originator_name was incorrectly used when constructing a From:
5124 header. Furthermore, $originator_name was used for submission mode
5125 messages from external hosts without From: headers in a similar way,
5126 which is clearly wrong.
5128 PH/37 Added control=suppress_local_fixups.
5130 PH/38 When log_selector = +received_sender was set, and the addition of the
5131 sender made the log line's construction buffer exactly full, or one byte
5132 less than full, an overflow happened when the terminating "\n" was
5135 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5136 when the result of a list match is failure because a DNS lookup failed.
5138 PH/40 RM_COMMAND is now used in the building process.
5140 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5141 the "build-* directories that it finds.
5143 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5144 address in a domain literal was a prefix of an interface address.
5146 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5147 when verifying a sender address, unless rewrite_headers is false.
5149 PH/44 Wrote a long comment about why errors_to addresses are verified as
5150 recipients, not senders.
5152 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5153 the ratelimit ACL was added.
5155 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5157 PH/46 Added extra information about PostgreSQL errors to the error string.
5159 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5160 avoiding the need to install special zones in a real server. This is
5161 backwards compatible; if it can't find the fake resolver, it drops back.
5162 Thus, both old and new test suites can be run.
5164 TF/02 Added util/ratelimit.pl
5166 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5167 clock is set back in time.
5169 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5170 Candler <B.Candler@pobox.com>.
5172 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5173 OK for addresses that are the result of redirections.
5175 PH/48 A number of further additions for the benefit of the new test suite,
5176 including a fake gethostbyname() that interfaces to the fake DNS resolver
5179 TF/06 The fix for widen_domains has also been applied to qualify_single and
5180 search_parents which are the other dnslookup options that can cause
5183 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5186 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5188 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5189 service" if it encountered an MX record with an empty target hostname.
5190 The message is now "an MX or SRV record indicated no SMTP service".
5192 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5193 if the DNS of the sending site is misconfigured. This is quite a
5194 common situation. This change restores the behaviour of treating a
5195 helo verification defer as a failure.
5197 PH/51 If self=fail was set on a router, the bounce message did not include the
5198 actual error message.
5204 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5206 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5207 command itself ended in error, the underlying message about the transport
5208 filter timeout was being overwritten with the pipe command error. Now the
5209 underlying error message should be appended to the second error message.
5211 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5213 PH/02 Reduce the amount of output that "make" produces by default. Full output
5214 can still be requested.
5216 PH/03 The warning log line about a condition test deferring for a "warn" verb
5217 was being output only once per connection, rather than after each
5218 occurrence (because it was using the same function as for successful
5219 "warn" verbs). This seems wrong, so I have changed it.
5221 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5222 it should not have, which might have caused a crash in the right
5223 circumstances, but probably never did.
5225 PH/04 Installed a modified version of Tony Finch's patch to make submission
5226 mode fix the return path as well as the Sender: header line, and to
5227 add a /name= option so that you can make the user's friendly name appear
5230 TF/03 Added the control = fakedefer ACL modifier.
5232 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5233 Mark Lowes for thorough testing.
5235 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5237 TK/03 Merged latest SRS patch from Miles Wilton.
5239 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5240 with the definition in sysexits.h (which is #included earlier).
5241 Fortunately, Exim does not actually use EX_OK. The code used to try to
5242 preserve the sysexits.h value, by assuming that macro definitions were
5243 scanned for macro replacements. I have been disabused of this notion,
5244 so now the code just undefines EX_OK before #including unistd.h.
5246 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5247 in the smtp transport. When a block could not be written in a single
5248 write() function, the timeout was being re-applied to each part-write.
5249 This seems wrong - if the receiver was accepting one byte at a time it
5250 would take for ever. The timeout is now adjusted when this happens. It
5251 doesn't have to be particularly precise.
5253 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5254 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5256 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5257 without a domain if the check to <postmaster@domain> fails.
5259 SC/01 Eximstats: added -xls and the ability to specify output files
5260 (patch written by Frank Heydlauf).
5262 SC/02 Eximstats: use FileHandles for outputting results.
5264 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5266 SC/04 Eximstats: fixed display of large numbers with -nvr option
5268 SC/05 Eximstats: fixed merging of reports with empty tables.
5270 SC/06 Eximstats: added the -include_original_destination flag
5272 SC/07 Eximstats: removed tabs and trailing whitespace.
5274 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5276 TK/06 MBOX spool code: Add real "From " MBOX separator line
5277 so the .eml file is really in mbox format (even though
5278 most programs do not really care). Patch from Alex Miller.
5280 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5281 The latter is generated from $received_to and is only set if the
5282 message has one envelope recipient. SA can use these headers,
5283 obviously out-of-the-box. Patch from Alex Miller.
5285 PH/08 The ${def test on a variable was returning false if the variable's
5286 value was "0", contrary to what the specification has always said!
5287 The result should be true unless the variable is empty.
5289 PH/09 The syntax error of a character other than { following "${if
5290 def:variable_name" (after optional whitespace) was not being diagnosed.
5291 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5292 accidental colon was present, for example, could give incorrect results.
5294 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5295 result is used (not including appendfile, where other changes are about
5298 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5299 This involved changing a lot of size variables from int to off_t. It
5300 should work with maildirs and everything.
5302 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5303 spamd dying while we are connected to it.
5305 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5308 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5309 function for BDB 4.3.
5311 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5313 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5314 the sources for PCRE have been re-arranged and more files are now
5317 PH/15 The code I had for printing potentially long long variables in PH/11
5318 above was not the best (it lost precision). The length of off_t variables
5319 is now inspected at build time, and an appropriate printing format (%ld
5320 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5321 to be "long long int" or "long int". This is needed for the internal
5322 formatting function string_vformat().
5324 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5325 the configuration file to be ":syslog", then the script "guesses" where
5326 the logs files are, rather than using the compiled in default. In our
5327 case the guess is not the same as the compiled default, so the script
5328 suddenly stopped working when I started to use syslog. The patch checks
5329 to see if log_file_path is "". If so, it attempts to read it from exim
5330 with no configuration file to get the compiled in version, before it
5331 falls back to the previous guessing code."
5333 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5334 implementing BATV in an Exim configuration. See NewStuff for the gory
5337 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5338 Makefile that are specific to HP-UX.
5340 PH/18 If the "use_postmaster" option was set for a recipient callout together
5341 with the "random" option, the postmaster address was used as the MAIL
5342 FROM address for the random test, but not for the subsequent recipient
5343 test. It is now used for both.
5345 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5346 patch removes a few documentation additions to RFC 3028, because the
5347 latest draft now contains them. It adds the new en;ascii-case comparator
5348 and a new error check for 8bit text in MIME parts. Comparator and
5349 require names are now matched exactly. I enabled the subaddress
5350 extension, but it is not well tested yet (read: it works for me)."
5352 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5353 rework some of the code of TK/09 above to avoid the hardwired use of
5354 "%lld" and "long long". Replaced the call to snprintf() with a call to
5357 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5358 records point to non-existent hosts", "retry timeout exceeded", and
5359 "retry time not reached for any host after a long failure period".
5361 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5362 experimental DomainKeys support:
5364 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5365 (2) On an error such as an illegally used "control", the wrong name for
5366 the control was given.
5368 These problems did NOT occur unless DomainKeys support was compiled.
5370 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5372 PH/24 Added ${if match_ip condition.
5374 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5375 databases so that it will be absolutely obvious if a crash occurs in the
5376 DB library. This is a regular occurrence (often caused by mis-matched
5379 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5380 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5381 picked up on a user's system that detects such things. There doesn't seem
5382 to be a gcc warning option for this - only an attribute that has to be
5383 put on the function's prototype. It seems that in Fedora Core 4 they have
5384 set this on a number of new functions. No doubt there will be more in due
5387 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5388 specify a transport. However, if an address that was verified by such a
5389 router was the subject of a callout, Exim crashed because it tried to
5390 read the rcpt_include_affixes from the non-existent transport. Now it
5391 just assumes that the setting of that option is false. This bug was
5392 introduced by 4.51/PH/31.
5394 PH/28 Changed -d+all to exclude +memory, because that information is very
5395 rarely of interest, but it makes the output a lot bigger. People tend to
5396 do -d+all out of habit.
5398 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5399 code in os-type was giving problems when libc.so lives in lib64, like on
5402 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5403 aren't the modern standard, and it seems that some systems' include files
5404 don't always have them. Exim was already checking for some of the newer
5405 ones like T_AAAA, and defining it itself. I've added checks for all the
5406 record types that Exim uses.
5408 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5409 not automatically generating a new one, as it is supposed to. This
5410 prevented TLS from working. If the file did exist, but contained invalid
5411 data, a new version was generated, as expected. It was only the case of a
5412 non-existent file that was broken.
5414 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5415 with a change in libdomainkeys > 0.64.
5417 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5418 from DNS. If the selector record carries the flag, it now has
5419 precedence over the domain-wide flag.
5421 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5423 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5424 the use of a port name that isn't defined in /etc/services) occurred, the
5425 message was deferred as in a normal delivery, and thus remained on the
5426 spool, instead of being failed because of the mua_wrapper setting. This
5427 is now fixed, and I tidied up some of the mua_wrapper messages at the
5430 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5431 the messages in a hash of arrays rather than using individual hashes.
5432 This is a bit cleaner and results in dramatic memory savings, albeit
5433 at a slight CPU cost.
5435 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5436 as requested by Marc Sherman.
5438 SC/10 Eximstats: added histograms for user specified patterns as requested
5441 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5443 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5444 fopen() in the content-scanning modules that did not already have it.
5450 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5451 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5453 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5455 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5457 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5458 to test Sieve filters that use "vacation".
5460 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5461 that changes the way the GnuTLS parameters are stored in the cache file.
5462 The new format can be generated externally. For backward compatibility,
5463 if the data in the cache doesn't make sense, Exim assumes it has read an
5464 old-format file, and it generates new data and writes a new file. This
5465 means that you can't go back to an older release without removing the
5468 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5469 work if there are any delivery delays because "one_time" forces the
5470 parent to be marked "delivered", so its unseen clone is never tried
5471 again. For this reason, Exim now forbids the simultaneous setting of
5474 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5475 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5476 entry if you want to know the details. The fix, however, neglected to
5477 consider the case when local delivery batching is involved. The test for
5478 "previously delivered" was not happening when checking to see if an
5479 address could be batched with a previous (undelivered) one; under
5480 certain circumstances this could lead to multiple deliveries to the same
5483 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5484 in its include files, and this causes problems building Exim.
5486 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5487 header_syntax) but Exim was just ignoring anything given after a slash.
5488 In particular, this caused confusion with an attempt to use "verify =
5489 reverse_host_lookup/defer_ok". An error is now given when options are
5490 supplied for verify items that do not have them. (Maybe reverse_host_
5491 lookup should have a defer_ok option, but that's a different point.)
5493 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5494 defined by RFC 821) to 2048, because there were problems with some AUTH
5495 commands, and RFC 1869 says the size should be increased for extended
5496 SMTP commands that take arguments.
5498 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5501 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5502 "unknown" error; now it says that the functionality isn't in the binary.
5504 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5505 an address' error message when a string expansion fails (syntax or
5506 whatever). Otherwise the password may appear in the log. Following change
5507 PH/42 below, there is no longer a chance of it appearing in a bounce
5510 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5512 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5513 of its addresses was ever tried. (Bugzilla bug #2.)
5515 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5516 the result incorrectly in the debug output. (It correctly added a newline
5517 to what was transported.)
5519 TF/01 Added $received_time.
5521 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5522 commented out examples of how to interface to a virus scanner and to
5523 SpamAssassin. Also added commented examples of av_scanner and
5524 spamd_address settings.
5526 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5527 and controls are allowed in which ACLs. There were a couple of minor
5528 errors. Some of the entries in the conditions table (which is a table of
5529 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5530 negation of where the condition IS allowed.
5532 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5534 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5535 header file does not have a version number, so I've had to invent a new
5536 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5537 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5538 radiusclient), but it was contributed by a Radius user.
5540 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5541 files or directories, for OpenSSL.
5543 PH/19 When an Exim process that is running as root has to create an Exim log
5544 file, it does so in a subprocess that runs as exim:exim so as to get the
5545 ownership right at creation (otherwise, other Exim processes might see
5546 the file with the wrong ownership). There was no test for failure of this
5547 fork() call, which would lead to the process getting stuck as it waited
5548 for a non-existent subprocess. Forks do occasionally fail when resources
5549 run out. I reviewed all the other calls to fork(); they all seem to check
5552 PH/20 When checking for unexpected SMTP input at connect time (before writing
5553 the banner), Exim was not dealing correctly with a non-positive return
5554 from the read() function. If the client had disconnected by this time,
5555 the result was a log entry for a synchronization error with an empty
5556 string after "input=" when read() returned zero. If read() returned -1
5557 (an event I could not check), uninitialized data bytes were printed.
5558 There were reports of junk text (parts of files, etc) appearing after
5561 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5563 PH/22 Added support for macro redefinition, and (re)definition in between
5564 driver and ACL definitions.
5566 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5567 forgetting to use the resulting value; it was using the unexpanded value.
5569 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5570 hadn't been configured. The fix is from Juergen Kreileder, who
5571 understands it better than I do:
5573 "Here's what I see happening with three configured cyrus_sasl
5574 authenticators configured (plain, login, cram-md5):
5576 On startup auth_cyrus_sasl_init() gets called for each of these.
5577 This means three calls to sasl_listmech() without a specified mech_list.
5578 => SASL tests which mechs of all available mechs actually work
5579 => three warnings about OTP not working
5580 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5582 With the patch, sasl_listmech() also gets called three times. But now
5583 SASL's mech_list option is set to the server_mech specified in the the
5584 authenticator. Or in other words, the answer from sasl_listmech()
5585 gets limited to just the mech you're testing for (which is different
5587 => the return list contains just 'plain' or 'login', 'cram-md5' or
5588 nothing depending on the value of ob->server_mech.
5590 I've just tested the patch: Authentication still works fine,
5591 unavailable mechs specified in the exim configuration are still
5592 caught, and the auth.log warnings about OTP are gone."
5594 PH/25 When debugging is enabled, the contents of the command line are added
5595 to the debugging output, even when log_selector=+arguments is not
5598 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5599 answer is "GNU", and only if the return is "GNU/something" is the answer
5602 PH/27 $acl_verify_message is now set immediately after the failure of a
5603 verification in an ACL, and so is available in subsequent modifiers. In
5604 particular, the message can be preserved by coding like this:
5606 warn !verify = sender
5607 set acl_m0 = $acl_verify_message
5609 Previously, $acl_verify_message was set only while expanding "message"
5610 and "log_message" when a very denied access.
5612 PH/28 Modified OS/os.c-Linux with
5614 -#ifndef OS_LOAD_AVERAGE
5615 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5617 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5618 nomenclature these days.)
5620 PH/29 Installed patch from the Sieve maintainer that adds the options
5621 sieve_useraddress and sieve_subaddress to the redirect router.
5623 PH/30 In these circumstances:
5624 . Two addresses routed to the same list of hosts;
5625 . First host does not offer TLS;
5626 . First host accepts first address;
5627 . First host gives temporary error to second address;
5628 . Second host offers TLS and a TLS session is established;
5629 . Second host accepts second address.
5630 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5631 and peerdn, if requested) that were in fact used only for the second
5634 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5635 attention to any local part prefix or suffix that was matched by the
5636 router that accepted the address. It now behaves in the same way as it
5637 does for delivery: the affixes are removed from the local part unless
5638 rcpt_include_affixes is set on the transport.
5640 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5641 timeout during the DATA phase of an incoming message.
5643 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5644 applied a patch sanctioned by the Sieve maintainer.
5646 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5647 the uid or gid is negative. A case of a negative gid caused this to be
5648 noticed. The fix allows for either to be negative.
5650 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5651 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5654 PH/36 The change PH/12 above was broken. Fixed it.
5656 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5657 the grounds that routing the same address twice would always produce the
5658 same answer. This might have been true once, but it is certainly no
5659 longer true now. Routing a child address may depend on the previous
5660 routing that produced that child. Some complicated redirection strategies
5661 went wrong when messages had multiple recipients, and made Exim's
5662 behaviour dependent on the order in which the addresses were given.
5664 I have moved the duplicate checking until after the routing is complete.
5665 Exim scans the addresses that are assigned to local and remote
5666 transports, and removes any duplicates. This means that more work will be
5667 done, as duplicates will always all be routed, but duplicates are
5668 presumably rare, so I don't expect this is of any significance.
5670 For deliveries to pipes, files, and autoreplies, the duplicate checking
5671 still happens during the routing process, since they are not going to be
5674 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5675 It corrects a timeout issue with spamd. This is Ian's comment: "The
5676 background is that sometimes spamd either never reads data from a
5677 connection it has accepted, or it never writes response data. The exiscan
5678 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5679 blindly assumes that writes won't block so it may never time out."
5681 PH/39 Allow G after quota size as well as K and M.
5683 PH/40 The value set for $authenticated_id in an authenticator may not contain
5684 binary zeroes or newlines because the value is written to log lines and
5685 to spool files. There was no check on this. Now the value is run through
5686 the string_printing() function so that such characters are converted to
5687 printable escape sequences.
5689 PH/41 $message_linecount is a new variable that contains the total number of
5690 lines in the message. Compare $body_linecount, which is the count for the
5693 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5694 bounce and delay warning messages, except in certain special cases, which
5697 (a) An SMTP error message from a remote host;
5698 (b) A message specified in a :fail: redirection;
5699 (c) A message specified in a "fail" command in a system filter;
5700 (d) A message specified in a FAIL return from the queryprogram router;
5701 (e) A message specified by the cannot_route_message router option.
5703 In these cases only, Exim does include the error details in bounce and
5704 warning messages. There are also a few cases where bland messages such
5705 as "unrouteable address" or "local delivery error" are given.
5707 PH/43 $value is now also set for the "else" part of a ${run expansion.
5709 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5710 being worked on, but at least Exim now implements the latest version to
5713 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5714 process to complete was treated as a delivery failure, a timeout while
5715 writing the message to the pipe was logged, but erroneously treated as a
5716 successful delivery. Such timeouts include transport filter timeouts. For
5717 consistency with the overall process timeout, these timeouts are now
5718 treated as errors, giving rise to delivery failures by default. However,
5719 there is now a new Boolean option for the pipe transport called
5720 timeout_defer, which, if set TRUE, converts the failures into defers for
5721 both kinds of timeout. A transport filter timeout is now identified in
5724 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5725 systems where "make" and "gmake" are different, calling "gmake" at top
5726 level broke things. I've arranged for the value of $(MAKE) to be passed
5727 from the Makefile to this script so that it can call the same version of
5731 A note about Exim versions 4.44 and 4.50
5732 ----------------------------------------
5734 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5735 changes of various kinds. As a consequence, a big documentation update was
5736 needed. This delayed the release for rather longer than seemed good, especially
5737 in the light of a couple of (minor) security issues. Therefore, the changes
5738 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5739 release. So 4.44 and 4.50 are in effect two different branches that both start
5742 I have left the 4.50 change log unchanged; it contains all the changes since
5743 4.43. The change log for 4.44 is below; many of its items are identical to
5744 those for 4.50. This seems to be the most sensible way to preserve the
5745 historical information.
5751 1. Minor wording change to the doc/README.SIEVE file.
5753 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5754 computation of the current number of files was incorrect.
5756 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5757 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5760 4. Give more explanation in the error message when the command for a transport
5761 filter fails to execute.
5763 5. There are several places where Exim runs a non-Exim command in a
5764 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5765 was being done only for the command run by the queryprogram router. It is
5766 now done for all such subprocesses. The other cases are: ${run, transport
5767 filters, and the commands run by the lmtp and pipe transports.
5769 6. Added CONFIGURE_GROUP build-time option.
5771 7. Some older OS have a limit of 256 on the maximum number of file
5772 descriptors. Exim was using setrlimit() to set 1000 as a large value
5773 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5774 systems. I've change it so that if it can't get 1000, it tries for 256.
5776 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5777 was an oversight, and furthermore, ever since the addition of extra
5778 controls (e.g. 4.43/32), the checks on when to allow different forms of
5779 "control" were broken. There should now be diagnostics for all cases when a
5780 control that does not make sense is encountered.
5782 9. Added the /retain_sender option to "control=submission".
5784 10. $recipients is now available in the predata ACL (oversight).
5786 11. Tidy the search cache before the fork to do a delivery from a message
5787 received from the command line. Otherwise the child will trigger a lookup
5788 failure and thereby defer the delivery if it tries to use (for example) a
5789 cached ldap connection that the parent has called unbind on.
5791 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5792 of $address_data from the recipient verification was clobbered by the
5793 sender verification.
5795 13. The value of address_data from a sender verification is now available in
5796 $sender_address_data in subsequent conditions in the ACL statement.
5798 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5800 15. Added a new option "connect=<time>" to callout options, to set a different
5803 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5804 was its contents. (It was OK if the option was not defined at all.)
5806 17. A "Completed" log line is now written for messages that are removed from
5807 the spool by the -Mrm option.
5809 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5810 information about exactly what failed.
5812 19. Added -dd to debug only the daemon process.
5814 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5815 handles timeouts, both on the server side and network timeouts. Renamed the
5816 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5818 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5819 It is now set to "smtps".
5821 22. $host_address is now set to the target address during the checking of
5822 ignore_target_hosts.
5824 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5825 being passed; this would have caused $sender_host_name to have been used if
5826 matching the list had actually called for a host name (not very likely,
5827 since this list is usually IP addresses). A host name is now passed as
5830 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5831 code that specifies a non-restarting handler (typically sigaction() in
5832 modern systems) in an attempt to fix a rare and obscure crash bug.
5834 25. Narrowed the window for a race in the daemon that could cause it to ignore
5835 SIGCHLD signals. This is not a major problem, because they are used only to
5836 wake it up if nothing else does.
5838 26. A malformed maildirsize file could cause Exim to calculate negative values
5839 for the mailbox size or file count. Odd effects could occur as a result.
5840 The maildirsize information is now recalculated if the size or filecount
5843 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5844 support for a long time. Removed HAVE_SYS_VFS_H.
5846 28. Installed the latest version of exipick from John Jetmore.
5848 29. In an address list, if the pattern was not a regular expression, an empty
5849 subject address (from a bounce message) matched only if the pattern was an
5850 empty string. Non-empty patterns were not even tested. This was the wrong
5851 because it is perfectly reasonable to use an empty address as part of a
5852 database query. An empty address is now tested by patterns that are
5853 lookups. However, all the other forms of pattern expect the subject to
5854 contain a local part and a domain, and therefore, for them, an empty
5855 address still always fails if the pattern is not itself empty.
5857 30. Exim went into a mad DNS loop when attempting to do a callout where the
5858 host was specified on an smtp transport, and looking it up yielded more
5859 than one IP address.
5861 31. Re-factored the code for checking spool and log partition space into a
5862 function that finds that data and another that does the check. The former
5863 is then used to implement four new variables: $spool_space, $log_space,
5864 $spool_inodes, and $log_inodes.
5866 32. The RFC2047 encoding function was originally intended for short strings
5867 such as real names; it was not keeping to the 75-character limit for
5868 encoded words that the RFC imposes. It now respects the limit, and
5869 generates multiple encoded words if necessary. To be on the safe side, I
5870 have increased the buffer size for the ${rfc2047: expansion operator from
5873 33. It is now permitted to omit both strings after an "if" condition; if the
5874 condition is true, the result is "true". As before, when the second string
5875 is omitted, a false condition yields an empty string. This makes it less
5876 cumbersome to write custom ACL and router conditions.
5878 34. Failure to deliver a bounce message always caused it to be frozen, even if
5879 there was an errors_to setting on the router. The errors_to setting is now
5882 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5883 canonical form (fully expanded) before being placed in
5884 $sender_host_address.
5886 36. The table in the code that translates DNS record types into text (T_A to
5887 "A" for instance) was missing entries for NS and CNAME. It is just possible
5888 that this could have caused confusion if both these types were looked up
5889 for the same domain, because the text type is used as part of Exim's
5890 per-process caching. But the chance of anyone hitting this buglet seems
5893 37. The dnsdb lookup has been extended in a number of ways.
5895 (1) There is a new type, "zns", which walks up the domain tree until it
5896 finds some nameserver records. It should be used with care.
5898 (2) There is a new type, "mxh", which is like "mx" except that it returns
5899 just the host names, not the priorities.
5901 (3) It is now possible to give a list of domains (or IP addresses) to be
5902 looked up. The behaviour when one of the lookups defers can be
5903 controlled by a keyword.
5905 (4) It is now possible to specify the separator character for use when
5906 multiple records are returned.
5908 38. The dnslists ACL condition has been extended: it is now possible to supply
5909 a list of IP addresses and/or domains to be looked up in a particular DNS
5912 39. Added log_selector=+queue_time_overall.
5914 40. When running the queue in the test harness, wait just a tad after forking a
5915 delivery process, to get repeatability of debugging output.
5917 41. Include certificate and key file names in error message when GnuTLS fails
5918 to set them up, because the GnuTLS error message doesn't include the name
5919 of the failing file when there is a problem reading it.
5921 42. Allow both -bf and -bF in the same test run.
5923 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
5925 44. The "Exiscan patch" is now merged into the mainline Exim source.
5927 45. Sometimes the final signoff response after QUIT could fail to get
5928 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5929 < 0 before doing a fflush(). This bug looks as though it goes back to the
5930 introduction of TLS in release 3.20, but "sometimes" must have been rare
5931 because the tests only now provoked it.
5933 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
5934 (this can affect the format of dates).
5936 47. exim_tidydb, when checking for the continued existence of a message for
5937 which it has found a message-specific retry record, was not finding
5938 messages that were in split spool directories. Consequently, it was
5939 deleting retry records that should have stayed in existence.
5941 48. Steve fixed some bugs in eximstats.
5943 49. The SPA authentication driver was not abandoning authentication and moving
5944 on to the next authenticator when an expansion was forced to fail,
5945 contradicting the general specification for all authenticators. Instead it
5946 was generating a temporary error. It now behaves as specified.
5948 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
5949 (the order specifies the preference for clients). The order is now AES256,
5950 AES128, 3DES, ARCFOUR128.
5952 51. Small patch to Sieve code - explicitly set From: when generating an
5955 52. Exim crashed if a remote delivery caused a very long error message to be
5956 recorded - for instance if somebody sent an entire SpamAssassin report back
5957 as a large number of 550 error lines. This bug was coincidentally fixed by
5958 increasing the size of one of Exim's internal buffers (big_buffer) that
5959 happened as part of the Exiscan merge. However, to be on the safe side, I
5960 have made the code more robust (and fixed the comments that describe what
5963 53. Now that there can be additional text after "Completed" in log lines (if
5964 the queue_time_overall log selector is set), a one-byte patch to exigrep
5965 was needed to allow it to recognize "Completed" as not the last thing in
5968 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
5969 patch that reportedly fixes this has been added. I am not expert enough to
5970 create a test for it. This is what the patch creator wrote:
5972 "I found a little strange behaviour of ldap code when working with
5973 Windows 2003 AD Domain, where users was placed in more than one
5974 Organization Units. When I tried to give exim partial DN, the exit code
5975 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
5976 But simultaneously result of request was absolutely normal ldap result,
5977 so I produce this patch..."
5979 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
5980 REFERENCE, so I have modified the code to exclude the patch when that macro
5983 55. Some experimental protocols are using DNS PTR records for new purposes. The
5984 keys for these records are domain names, not reversed IP addresses. The
5985 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5986 leaves it alone. Component reversal etc. now happens only for IP addresses.
5989 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
5991 57. Double the size of the debug message buffer (to 2048) so that more of very
5992 long debug lines gets shown.
5994 58. The exicyclog utility now does better if the number of log files to keep
5995 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
5997 59. Two changes related to the smtp_active_hostname option:
5999 (1) $smtp_active_hostname is now available as a variable.
6000 (2) The default for smtp_banner uses $smtp_active_hostname instead
6001 of $primary_hostname.
6003 60. The host_aton() function is supposed to be passed a string that is known
6004 to be a valid IP address. However, in the case of IPv6 addresses, it was
6005 not checking this. This is a hostage to fortune. Exim now panics and dies
6006 if the condition is not met. A case was found where this could be provoked
6007 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6008 components; fortuitously, this particular loophole had already been fixed
6009 by change 4.50/55 above.
6011 If there are any other similar loopholes, the new check in host_aton()
6012 itself should stop them being exploited. The report I received stated that
6013 data on the command line could provoke the exploit when Exim was running as
6014 exim, but did not say which command line option was involved. All I could
6015 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6016 running as the user.
6019 61. There was a buffer overflow vulnerability in the SPA authentication code
6020 (which came originally from the Samba project). I have added a test to the
6021 spa_base64_to_bits() function which I hope fixes it.
6024 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6025 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6027 63. The daemon start-up calls getloadavg() while still root for those OS that
6028 need the first call to be done as root, but it missed one case: when
6029 deliver_queue_load_max is set with deliver_drop_privilege. This is
6030 necessary for the benefit of the queue runner, because there is no re-exec
6031 when deliver_drop_privilege is set.
6033 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6034 This has been fixed.
6036 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6037 was in use, was not putting the data itself into the right store pool;
6038 consequently, it could be overwritten for a subsequent message in the same
6039 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6042 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6044 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6045 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6046 the function were treating the return as a boolean value, which happened to
6047 work because 0=false and not-0=true, but is not correct code.
6049 68. The host_aton() function was not handling scoped IPv6 addresses (those
6050 with, for example, "%eth0" on the end) correctly.
6052 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6053 negated items (that is, ~something) in unsigned ints. Some compilers
6054 apparently mutter when there is no cast.
6056 70. If an address verification called from an ACL failed, and did not produce a
6057 user-specific message (i.e. there was only a "system" message), nothing was
6058 put in $acl_verify_message. In this situation, it now puts the system
6061 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6062 session; change 4.31/43 added the unwanted input to the log line - except
6063 that it did not do this in the start of session case. It now does.
6065 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6066 This is wrong and can cause the other end to generate a synchronization
6067 error if it is another Exim or anything else that does the synchronization
6068 check. A QUIT command is no longer sent after a timeout.
6070 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6071 during host lookups.
6073 74. The defer_ok option of callout verification was not working if it was used
6074 when verifying addresses in header lines, that is, for this case:
6076 verify = header_sender/callout=defer_ok
6078 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6079 those file descriptors could be used for SMTP connections. If anything
6080 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6081 could be sent to the SMTP client, causing chaos. The daemon now opens
6082 stdin, stdout, and stderr to /dev/null when it puts itself into the
6085 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6086 log by default. The user can override this with suitable Perl magic.
6088 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6089 the log message when discard triggers, was not working for the DATA ACL or
6090 for the non-SMTP ACL.
6092 78. Error message wording change in sieve.c.
6094 79. If smtp_accept_max_per_host was set, the number of connections could be
6095 restricted to fewer than expected, because the daemon was trying to set up
6096 a new connection before checking whether the processes handling previous
6097 connections had finished. The check for completed processes is now done
6098 earlier. On busy systems, this bug wouldn't be noticed because something
6099 else would have woken the daemon, and it would have reaped the completed
6102 80. If a message was submitted locally by a user whose login name contained one
6103 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6104 It caused a spool format error. I have fixed the spool reading code. A
6105 related problem was that the "from" clause in the Received: line became
6106 illegal because of the space(s). It is now covered by ${quote_local_part.
6108 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6111 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6113 83. Patch from Sieve maintainer to fix unterminated string problem in
6114 "vacation" handling.
6116 84. Some minor changes to the Linux configuration files to help with other
6117 OS variants using glibc.
6119 85. One more patch for Sieve to update vacation handling to latest spec.
6122 ----------------------------------------------------
6123 See the note above about the 4.44 and 4.50 releases.
6124 ----------------------------------------------------
6130 1. Change 4.43/35 introduced a bug that caused file counts to be
6131 incorrectly computed when quota_filecount was set in an appendfile
6134 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6135 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6138 3. Give more explanation in the error message when the command for a transport
6139 filter fails to execute.
6141 4. There are several places where Exim runs a non-Exim command in a
6142 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6143 was being done only for the command run by the queryprogram router. It is
6144 now done for all such subprocesses. The other cases are: ${run, transport
6145 filters, and the commands run by the lmtp and pipe transports.
6147 5. Some older OS have a limit of 256 on the maximum number of file
6148 descriptors. Exim was using setrlimit() to set 1000 as a large value
6149 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6150 systems. I've change it so that if it can't get 1000, it tries for 256.
6152 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6153 was an oversight, and furthermore, ever since the addition of extra
6154 controls (e.g. 4.43/32), the checks on when to allow different forms of
6155 "control" were broken. There should now be diagnostics for all cases when a
6156 control that does not make sense is encountered.
6158 7. $recipients is now available in the predata ACL (oversight).
6160 8. Tidy the search cache before the fork to do a delivery from a message
6161 received from the command line. Otherwise the child will trigger a lookup
6162 failure and thereby defer the delivery if it tries to use (for example) a
6163 cached ldap connection that the parent has called unbind on.
6165 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6166 of $address_data from the recipient verification was clobbered by the
6167 sender verification.
6169 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6170 was its contents. (It was OK if the option was not defined at all.)
6172 11. A "Completed" log line is now written for messages that are removed from
6173 the spool by the -Mrm option.
6175 12. $host_address is now set to the target address during the checking of
6176 ignore_target_hosts.
6178 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6179 being passed; this would have caused $sender_host_name to have been used if
6180 matching the list had actually called for a host name (not very likely,
6181 since this list is usually IP addresses). A host name is now passed as
6184 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6185 code that specifies a non-restarting handler (typically sigaction() in
6186 modern systems) in an attempt to fix a rare and obscure crash bug.
6188 15. Narrowed the window for a race in the daemon that could cause it to ignore
6189 SIGCHLD signals. This is not a major problem, because they are used only to
6190 wake it up if nothing else does.
6192 16. A malformed maildirsize file could cause Exim to calculate negative values
6193 for the mailbox size or file count. Odd effects could occur as a result.
6194 The maildirsize information is now recalculated if the size or filecount
6197 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6198 support for a long time. Removed HAVE_SYS_VFS_H.
6200 18. Updated exipick to current release from John Jetmore.
6202 19. Allow an empty sender to be matched against a lookup in an address list.
6203 Previously the only cases considered were a regular expression, or an
6206 20. Exim went into a mad DNS lookup loop when doing a callout where the
6207 host was specified on the transport, if the DNS lookup yielded more than
6210 21. The RFC2047 encoding function was originally intended for short strings
6211 such as real names; it was not keeping to the 75-character limit for
6212 encoded words that the RFC imposes. It now respects the limit, and
6213 generates multiple encoded words if necessary. To be on the safe side, I
6214 have increased the buffer size for the ${rfc2047: expansion operator from
6217 22. Failure to deliver a bounce message always caused it to be frozen, even if
6218 there was an errors_to setting on the router. The errors_to setting is now
6221 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6222 canonical form (fully expanded) before being placed in
6223 $sender_host_address.
6225 24. Updated eximstats to version 1.33
6227 25. Include certificate and key file names in error message when GnuTLS fails
6228 to set them up, because the GnuTLS error message doesn't include the name
6229 of the failing file when there is a problem reading it.
6231 26. Expand error message when OpenSSL has problems setting up cert/key files.
6234 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6235 (this can affect the format of dates).
6237 28. exim_tidydb, when checking for the continued existence of a message for
6238 which it has found a message-specific retry record, was not finding
6239 messages that were in split spool directories. Consequently, it was
6240 deleting retry records that should have stayed in existence.
6242 29. eximstats updated to version 1.35
6243 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6244 1.35 - bugfix such that pie charts by volume are generated correctly
6246 30. The SPA authentication driver was not abandoning authentication and moving
6247 on to the next authenticator when an expansion was forced to fail,
6248 contradicting the general specification for all authenticators. Instead it
6249 was generating a temporary error. It now behaves as specified.
6251 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6252 (the order specifies the preference for clients). The order is now AES256,
6253 AES128, 3DES, ARCFOUR128.
6255 31. Small patch to Sieve code - explicitly set From: when generating an
6258 32. Exim crashed if a remote delivery caused a very long error message to be
6259 recorded - for instance if somebody sent an entire SpamAssassin report back
6260 as a large number of 550 error lines. This bug was coincidentally fixed by
6261 increasing the size of one of Exim's internal buffers (big_buffer) that
6262 happened as part of the Exiscan merge. However, to be on the safe side, I
6263 have made the code more robust (and fixed the comments that describe what
6266 33. Some experimental protocols are using DNS PTR records for new purposes. The
6267 keys for these records are domain names, not reversed IP addresses. The
6268 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6269 leaves it alone. Component reversal etc. now happens only for IP addresses.
6272 34. The host_aton() function is supposed to be passed a string that is known
6273 to be a valid IP address. However, in the case of IPv6 addresses, it was
6274 not checking this. This is a hostage to fortune. Exim now panics and dies
6275 if the condition is not met. A case was found where this could be provoked
6276 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6277 components; fortuitously, this particular loophole had already been fixed
6278 by change 4.50/55 or 4.44/33 above.
6280 If there are any other similar loopholes, the new check in host_aton()
6281 itself should stop them being exploited. The report I received stated that
6282 data on the command line could provoke the exploit when Exim was running as
6283 exim, but did not say which command line option was involved. All I could
6284 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6285 running as the user.
6288 35. There was a buffer overflow vulnerability in the SPA authentication code
6289 (which came originally from the Samba project). I have added a test to the
6290 spa_base64_to_bits() function which I hope fixes it.
6293 36. The daemon start-up calls getloadavg() while still root for those OS that
6294 need the first call to be done as root, but it missed one case: when
6295 deliver_queue_load_max is set with deliver_drop_privilege. This is
6296 necessary for the benefit of the queue runner, because there is no re-exec
6297 when deliver_drop_privilege is set.
6299 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6300 was in use, was not putting the data itself into the right store pool;
6301 consequently, it could be overwritten for a subsequent message in the same
6302 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6305 38. Sometimes the final signoff response after QUIT could fail to get
6306 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6307 < 0 before doing a fflush(). This bug looks as though it goes back to the
6308 introduction of TLS in release 3.20, but "sometimes" must have been rare
6309 because the tests only now provoked it.
6315 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6316 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6317 Now it returns an integer. A number of calls were still expecting a T/F
6318 return. Fortuitously, in all cases, the tests worked in OK situations,
6319 which is the norm. However, things would have gone wrong on any write
6320 failures on the smtp file descriptor. This function is used when sending
6321 messages over SMTP and also when doing verify callouts.
6323 2. When Exim is called to do synchronous delivery of a locally submitted
6324 message (the -odf or -odi options), it no longer closes stderr before doing
6327 3. Implemented the mua_wrapper option.
6329 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6331 5. Implemented the functions header_remove(), header_testname(),
6332 header_add_at_position(), and receive_remove_recipient(), and exported them
6335 6. If an ACL "warn" statement specified the addition of headers, Exim already
6336 inserted X-ACL-Warn: at the start if there was no header name. However, it
6337 was not making this test for the second and subsequent header lines if
6338 there were newlines in the string. This meant that an invalid header could
6339 be inserted if Exim was badly configured.
6341 7. Allow an ACL "warn" statement to add header lines at the start or after all
6342 the Received: headers, as well as at the end.
6344 8. Added the rcpt_4xx retry error code.
6346 9. Added postmaster_mailfrom=xxx to callout verification option.
6348 10. Added mailfrom=xxxx to the callout verification option, for verify=
6351 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6352 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6354 12. Inserted some casts to stop certain compilers warning when using pointer
6355 differences as field lengths or precisions in printf-type calls (mostly
6356 affecting debugging statements).
6358 13. Added optional readline() support for -be (dynamically loaded).
6360 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6361 same clock tick as a message's arrival, so that its received time was the
6362 same as the "first fail" time on the retry record, and that message
6363 remained on the queue past the ultimate address timeout, every queue runner
6364 would try a delivery (because it was past the ultimate address timeout) but
6365 after another failure, the ultimate address timeout, which should have then
6366 bounced the address, did not kick in. This was a "< instead of <=" error;
6367 in most cases the first failure would have been in the next clock tick
6368 after the received time, and all would be well.
6370 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6371 being recognized when the domain list was tested by the match_domain
6372 condition in an expansion string.
6374 16. Added the ${str2b64: operator.
6376 17. Exim was always calling setrlimit() to set a large limit for the number of
6377 processes, without checking whether the existing limit was already
6378 adequate. (It did check for the limit on file descriptors.) Furthermore,
6379 errors from getrlimit() and setrlimit() were being ignored. Now they are
6380 logged to the main and panic logs, but Exim does carry on, to try to do its
6381 job under whatever limits there are.
6383 18. Imported PCRE 5.0.
6385 19. Trivial typo in log message " temporarily refused connection" (the leading
6388 20. If the log selector return_path_on_delivery was set and an address was
6389 redirected to /dev/null, the delivery process crashed because it assumed
6390 that a return path would always be set for a "successful" delivery. In this
6391 case, the whole delivery is bypassed as an optimization, and therefore no
6394 21. Internal re-arrangement: the function for sending a challenge and reading
6395 a response while authentication was assuming a zero-terminated challenge
6396 string. It's now changed to take a pointer and a length, to allow for
6397 binary data in such strings.
6399 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6401 23. Exim was not respecting finduser_retries when seeking the login of the
6402 uid under which it was called; it was always trying 10 times. (The default
6403 setting of finduser_retries is zero.) Also, it was sleeping after the final
6404 failure, which is pointless.
6406 24. Implemented tls_on_connect_ports.
6408 25. Implemented acl_smtp_predata.
6410 26. If the domain in control=submission is set empty, Exim assumes that the
6411 authenticated id is a complete email address when it generates From: or
6412 Sender: header lines.
6414 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6415 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6416 chown and chgrp in /bin and hostname in /usr/bin.
6418 28. Exim was keeping the "process log" file open after each use, just as it
6419 does for the main log. This opens the possibility of it remaining open for
6420 long periods when the USR1 signal hits a daemon. Occasional processlog
6421 errors were reported, that could have been caused by this. Anyway, it seems
6422 much more sensible not to leave this file open at all, so that is what now
6425 29. The long-running daemon process does not normally write to the log once it
6426 has entered its main loop, and it closes the log before doing so. This is
6427 so that log files can straightforwardly be renamed and moved. However,
6428 there are a couple of unusual error situations where the daemon does write
6429 log entries, and I had neglected to close the log afterwards.
6431 30. The text of an SMTP error response that was received during a remote
6432 delivery was being truncated at 512 bytes. This is too short for some of
6433 the long messages that one sometimes sees. I've increased the limit to
6436 31. It is now possible to make retry rules that apply only when a message has a
6437 specific sender, in particular, an empty sender.
6439 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6440 it possible to be selective about when SMTP synchronization is enforced.
6442 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6444 32. Implemented hosts_connection_nolog.
6446 33. Added an ACL for QUIT.
6448 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6451 35. Added mailbox_size and mailbox_filecount to appendfile.
6453 36. Added control = no_multiline_responses to ACLs.
6455 37. There was a bug in the logic of the code that waits for the clock to tick
6456 in the case where the clock went backwards by a substantial amount such
6457 that the microsecond fraction of "now" was more than the microsecond
6458 fraction of "then" (but the whole seconds number was less).
6460 38. Added support for the libradius Radius client library this is found on
6461 FreeBSD (previously only the radiusclient library was supported).
6467 1. When certain lookups returned multiple values in the form name=value, the
6468 quoting of the values was not always being done properly. Specifically:
6469 (a) If the value started with a double quote, but contained no whitespace,
6471 (b) If the value contained whitespace other than a space character (i.e.
6472 tabs or newlines or carriage returns) it was not quoted.
6473 This fix has been applied to the mysql and pgsql lookups by writing a
6474 separate quoting function and calling it from the lookup code. The fix
6475 should probably also be applied to nisplus, ibase and oracle lookups, but
6476 since I cannot test any of those, I have not disturbed their existing code.
6478 2. A hit in the callout cache for a specific address caused a log line with no
6479 reason for rejecting RCPT. Now it says "Previous (cached) callout
6480 verification failure".
6482 3. There was an off-by-one bug in the queryprogram router. An over-long
6483 return line was truncated at 256 instead of 255 characters, thereby
6484 overflowing its buffer with the terminating zero. As well as fixing this, I
6485 have increased the buffer size to 1024 (and made a note to document this).
6487 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6488 when Exim is waiting for an SMTP response from a remote server, Exim
6489 restarts its select() call on the socket, thereby resetting its timeout.
6490 This is not a problem when such interrupts are rare. Somebody set up a cron
6491 job to run exiwhat every 2 minutes, which is less than the normal select()
6492 timeout (5 or 10 minutes). This meant that the select() timeout never
6493 kicked in because it was always reset. I have fixed this by comparing the
6494 time when an interrupt arrives with the time at the start of the first call
6495 to select(). If more time than the timeout has elapsed, the interrupt is
6496 treated as a timeout.
6498 5. Some internal re-factoring in preparation for the addition of Sieve
6499 extensions (by MH). In particular, the "personal" test is moved to a
6500 separate function, and given an option for scanning Cc: and Bcc: (which is
6501 not set for Exim filters).
6503 6. When Exim created an email address using the login of the caller as the
6504 local part (e.g. when creating a From: or Sender: header line), it was not
6505 quoting the local part when it contained special characters such as @.
6507 7. Installed new OpenBSD configuration files.
6509 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6510 try to make them clearer.
6512 9. Callout options, other than the timeout value, were being ignored when
6513 verifying sender addresses in header lines. For example, when using
6515 verify = header_sender/callout=no_cache
6517 the cache was (incorrectly) being used.
6519 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6520 only the Cygwin environment.
6522 11. When return_path_on_delivery was set as a log selector, if different remote
6523 addresses in the same message used different return paths and parallel
6524 remote delivery occurred, the wrong values would sometimes be logged.
6525 (Whenever a remote delivery process finished, the return path value from
6526 the most recently started remote delivery process was logged.)
6528 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6529 lines when AUTH and/or TLS are in use. This is the "received protocol"
6530 field. Exim used to use "asmtp" for authenticated SMTP, without any
6531 indication (in the protocol name) for TLS use. Now it follows the RFC and
6532 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6533 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6534 names appear in log lines as well as in Received: header lines.
6536 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6537 extensions, and comparison tests, and to fix some bugs.
6539 14. Changes to the "personal" filter test:
6541 (1) The test was buggy in that it was just doing the equivalent of
6542 "contains" tests on header lines. For example, if a user's address was
6543 anne@some.where, the "personal" test would incorrectly be true for
6545 To: susanne@some.where
6547 This test is now done by extracting each address from the header in turn,
6548 and checking the entire address. Other tests that are part of "personal"
6549 are now done using regular expressions (for example, to check local parts
6550 of addresses in From: header lines).
6552 (2) The list of non-personal local parts in From: addresses has been
6553 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6554 taken from the Sieve specification recommendations.
6556 (3) If the message contains any header line starting with "List-" it is
6557 treated as non-personal.
6559 (4) The test for "circular" in the Subject: header line has been removed
6560 because it now seems ill-conceived.
6562 15. Minor typos in src/EDITME comments corrected.
6564 16. Installed latest exipick from John Jetmore.
6566 17. If headers_add on a router specified a text string that was too long for
6567 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6568 of string_sprintf() is now avoided.
6570 18. $message_body_size was not set (it was always zero) when running the DATA
6571 ACL and the local_scan() function.
6573 19. For the "mail" command in an Exim filter, no default was being set for
6574 the once_repeat time, causing a random time value to be used if "once" was
6575 specified. (If the value happened to be <= 0, no repeat happened.) The
6576 default is now 0s, meaning "never repeat". The "vacation" command was OK
6577 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6578 (I found it when inspecting the code).
6580 20. There is now an overall timeout for performing a callout verification. It
6581 defaults to 4 times the callout timeout, which applies to individual SMTP
6582 commands during the callout. The overall timeout applies when there is more
6583 than one host that can be tried. The timeout is checked before trying the
6584 next host. This prevents very long delays if there are a large number of
6585 hosts and all are timing out (e.g. when the network connections are timing
6586 out). The value of the overall timeout can be changed by specifying an
6587 additional sub-option for "callout", called "maxwait". For example:
6589 verify = sender/callout=5s,maxwait=20s
6591 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6592 to the end before writing, but this should make it even safer).
6594 22. Exim was forgetting that it had advertised PIPELINING for the second and
6595 subsequent messages on an SMTP connection. It was also not resetting its
6596 memory on STARTTLS and an internal HELO.
6598 23. When Exim logs an SMTP synchronization error within a session, it now
6599 records whether PIPELINING has been advertised or not.
6601 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6602 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6603 rather than long int.
6605 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6607 26. Added the never_mail option to autoreply.
6613 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6614 crash if the getsockname() call failed; this can happen if a connection is
6615 closed very soon after it is established. The problem was simply in the
6616 order in which certain operations were done, causing Exim to try to write
6617 to the SMTP stream before it had set up the file descriptor. The bug has
6618 been fixed by making things happen in the correct order.
6624 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6625 before the connection was closed, thus losing the rejection response.
6627 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6628 some early Solaris releases, but causes trouble in current releases where
6629 socklen_t is defined.
6631 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6634 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6637 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6638 that releasing the top of it at the end releases what was used for sub-
6639 expansions (unless the block got too big). However, discard this block if
6640 the first thing is a variable or header, so that we can use its block when
6641 it is dynamic (useful for very large $message_headers, for example).
6643 6. Lookups now cache *every* query, not just the most recent. A new, separate
6644 store pool is used for this. It can be recovered when all lookup caches are
6645 flushed. Lookups now release memory at the end of their result strings.
6646 This has involved some general refactoring of the lookup sources.
6648 7. Some code has been added to the store_xxx() functions to reduce the amount
6649 of flapping under certain conditions.
6651 8. log_incoming_interface used to affect only the <= reception log lines. Now
6652 it causes the local interface and port to be added to several more SMTP log
6653 lines, for example "SMTP connection from", and rejection lines.
6655 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6657 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6659 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6660 could be overwritten at the end of the current message (or the start of a
6661 new message if it was set in a HELO ACL). The value is now preserved for
6662 the duration of the SMTP connection.
6664 12. If a transport had a headers_rewrite setting, and a matching header line
6665 contained an unqualified address, that address was qualified, even if it
6666 did not match any rewriting rules. The underlying bug was that the values
6667 of the flags that permit the existence of unqualified sender and recipient
6668 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6669 non-local messages, and by -bnq for local messages) were not being
6670 preserved with the message after it was received.
6672 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6673 "next input=" as part of the text comprising the host identity instead of
6674 the correct text. The code was using the same buffer for two different
6675 strings. However, depending on which order the printing function evaluated
6676 its arguments, the bug did not always show up. Under Linux, for example, my
6677 test suite worked just fine.
6679 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6680 doesn't work with some older versions of Perl. It has been changed to "my",
6681 which in any case is probably the better facility to use.
6683 15. A really picky compiler found some instances of statements for creating
6684 error messages that either had too many or two few arguments for the format
6687 16. The size of the buffer for calls to the DNS resolver has been increased
6688 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6689 for addresses that have a lot of PTR records. This alleviates a problem; it
6690 does not fully solve it.
6692 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6693 buffer now truncates the list and logs the incident, which is the same
6694 action as happens when Exim is looking up a host name and its aliases.
6695 Previously in this situation something unpredictable would happen;
6696 sometimes it was "internal error: store_reset failed".
6698 18. If a server dropped the connection unexpectedly when an Exim client was
6699 using GnuTLS and trying to read a response, the client delivery process
6700 crashed while trying to generate an error log message.
6702 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6703 string, for example:
6705 warn message = H1: something\nH2: something
6707 the text was added as a single header line from Exim's point of view
6708 though it ended up OK in the delivered message. However, searching for the
6709 second and subsequent header lines using $h_h2: did not work. This has been
6710 fixed. Similarly, if a system filter added multiple headers in this way,
6711 the routers could not see them.
6713 20. Expanded the error message when iplsearch is called with an invalid key to
6714 suggest using net-iplsearch in a host list.
6716 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6717 ACLs are no longer actually imposed (and a message to that effect is
6720 22. If a "gecos" field in a passwd entry contained escaped characters, in
6721 particular, if it contained a \" sequence, Exim got it wrong when building
6722 a From: or a Sender: header from that name. A second bug also caused
6723 incorrect handling when an unquoted " was present following a character
6724 that needed quoting.
6726 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6727 was not being matched caselessly.
6729 24. Arranged for all hyphens in the exim.8 source to be escaped with
6732 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6733 properly. Recipient callout cache records were still being keyed to include
6734 the sender, even when use_sender was set false. This led to far more
6735 callouts that were necessary. The sender is no longer included in the key
6736 when use_sender is false.
6738 26. Added "control = submission" modifier to ACLs.
6740 27. Added the ${base62d: operator to decode base 62 numbers.
6742 28. dnsdb lookups can now access SRV records.
6744 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6745 the configuration file.
6747 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6748 (-v) mode. This makes the output for a verbose queue run more intelligible.
6750 31. Added a use_postmaster feature to recipient callouts.
6752 32. Added the $body_zerocount variable, containing the number of binary zero
6753 bytes in the message body.
6755 33. The time of last modification of the "new" subdirectory is now used as the
6756 "mailbox time last read" when there is a quota error for a maildir
6759 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6761 35. Added +ignore_unknown as a special item in host lists.
6763 36. Code for decoding IPv6 addresses in host lists is now included, even if
6764 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6765 address was recognized as an IP address, but was then not correctly decoded
6766 into binary, causing unexpected and incorrect effects when compared with
6773 1. Very minor rewording of debugging text in manualroute to say "list of
6774 hosts" instead of "hostlist".
6776 2. If verify=header_syntax was set, and a header line with an unqualified
6777 address (no domain) and a large number of spaces between the end of the
6778 name and the colon was received, the reception process suffered a buffer
6779 overflow, and (when I tested it) crashed. This was caused by some obsolete
6780 code that should have been removed. The fix is to remove it!
6782 3. When running in the test harness, delay a bit after writing a bounce
6783 message to get a bit more predictability in the log output.
6785 4. Added a call to search_tidyup() just before forking a reception process. In
6786 theory, someone could use a lookup in the expansion of smtp_accept_max_
6787 per_host which, without the tidyup, could leave open a database connection.
6789 5. Added the variables $recipient_data and $sender_data which get set from a
6790 lookup success in an ACL "recipients" or "senders" condition, or a router
6791 "senders" option, similar to $domain_data and $local_part_data.
6793 6. Moved the writing of debug_print from before to after the "senders" test
6796 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6797 problems for message scanning, either using a data ACL, or using
6798 local_scan() because the Received: header was not generated till after they
6799 were called (in order to set the time as the time of reception completion).
6800 I have revised the way this works. The header is now generated after the
6801 body is received, but before the ACL or local_scan() are called. After they
6802 are run, the timestamp in the header is updated.
6808 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6809 before starting a queue runner without re-exec. This happened only when
6810 deliver_drop_privilege was set or when the Exim user was set to root. The
6811 effect of the bug was that timeouts during subsequent deliveries caused
6812 crashes instead of being properly handled. The handler is now left at its
6813 default (and expected) setting.
6815 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6816 message, again when deliver_drop_privilege is set or Exim is run as root.
6817 The bug described in (1) was not present in this case, but the tidying up
6818 of the other signals was missing. I have made the two cases consistent.
6820 3. The ignore_target_hosts setting on a manualroute router was being ignored
6821 for hosts that were looked up using the /MX notation.
6823 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6826 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6827 operated on the sender address. After changing the $sender_address to <>
6828 for the sender address verify, Exim was re-instated it as the original
6829 (before rewriting) address, but remembering that it had rewritten it, so it
6830 wasn't rewriting it again. This bug also had the effect of breaking the
6831 sender address verification caching when the sender address was rewritten.
6833 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6834 This has been changed so that if the ip literal address matches
6835 ignore_target_hosts, the router declines.
6837 7. Added expansion conditions match_domain, match_address, and match_local_
6838 part (NOT match_host).
6840 8. The placeholder for the Received: header didn't have a length field set.
6842 9. Added code to Exim itself and to exim_lock to test for a specific race
6843 condition that could lead to file corruption when using MBX delivery. The
6844 issue is with the lockfile that is created in /tmp. If this file is removed
6845 after a process has opened it but before that process has acquired a lock,
6846 there is the potential for a second process to recreate the file and also
6847 acquire a lock. This could lead to two Exim processes writing to the file
6848 at the same time. The added code performs the same test as UW imapd; it
6849 checks after acquiring the lock that its file descriptor still refers to
6850 the same named file.
6852 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6853 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6856 11. Added the smtp_active_hostname option. If used, this will typically be made
6857 to depend on the incoming interface address. Because $interface_address is
6858 not set up until the daemon has forked a reception process, error responses
6859 that can happen earlier (such as "too many connections") no longer contain
6862 12. If an expansion in a condition on a "warn" statement fails because a lookup
6863 defers, the "warn" statement is abandoned, and the next ACL statement is
6864 processed. Previously this caused the whole ACL to be aborted.
6866 13. Added the iplsearch lookup type.
6868 14. Added ident_timeout as a log selector.
6870 15. Added tls_certificate_verified as a log selector.
6872 16. Added a global option tls_require_ciphers (compare the smtp transport
6873 option of the same name). This controls incoming TLS connections.
6875 17. I finally figured out how to make tls_require_ciphers do a similar thing
6876 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6877 before starting the TLS session.
6879 18. Tabs are now shown as \t in -bP output.
6881 19. If the log selector return_path_on_delivery was set, Exim crashed when
6882 bouncing a message because it had too many Received: header lines.
6884 20. If two routers both had headers_remove settings, and the first one included
6885 a superfluous trailing colon, the final name in the first list and the
6886 first name in the second list were incorrectly joined into one item (with a
6887 colon in the middle).
6893 1. Added -C and -D options to the exinext utility, mainly to make it easier
6894 to include in the automated testing, but these could be helpful when
6895 multiple configurations are in use.
6897 2. The exinext utility was not formatting the output nicely when there was
6898 an alternate port involved in the retry record key, nor when there was a
6899 message id as well (for retries that were specific to a specific message
6900 and a specific host). It was also confused by IPv6 addresses, because of
6901 the additional colons they contain. I have fixed the IPv4 problem, and
6902 patched it up to do a reasonable job for IPv6.
6904 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6905 delivery, the log line now contains "pipelined" if PIPELINING was used.
6907 4. An SMTP transport process used to panic and die if the bind() call to set
6908 an explicit outgoing interface failed. This has been changed; it is now
6909 treated in the same way as a connect() failure.
6911 5. A reference to $sender_host_name in the part of a conditional expansion
6912 that was being skipped was still causing a DNS lookup. This no longer
6915 6. The def: expansion condition was not recognizing references to header lines
6916 that used bh_ and bheader_.
6918 7. Added the _cache feature to named lists.
6920 8. The code for checking quota_filecount in the appendfile transport was
6921 allowing one more file than it should have been.
6923 9. For compatibility with Sendmail, the command line option
6931 and sets the incoming protocol and host name (for trusted callers). The
6932 host name and its colon can be omitted when only the protocol is to be set.
6933 Note the Exim already has two private options, -pd and -ps, that refer to
6934 embedded Perl. It is therefore impossible to set a protocol value of "d" or
6935 "s", but I don't think that's a major issue.
6937 10. A number of refactoring changes to the code, none of which should affect
6940 (a) The number of logging options was getting close to filling up the
6941 32-bit word that was used as a bit map. I have split them into two classes:
6942 those that are passed in the argument to log_write(), and those that are
6943 only ever tested independently outside of that function. These are now in
6944 separate 32-bit words, so there is plenty of room for expansion again.
6945 There is no change in the user interface or the logging behaviour.
6947 (b) When building, for example, log lines, the code previously used a
6948 macro that called string_cat() twice, in order to add two strings. This is
6949 not really sufficiently general. Furthermore, there was one instance where
6950 it was actually wrong because one of the argument was used twice, and in
6951 one call a function was used. (As it happened, calling the function twice
6952 did not affect the overall behaviour.) The macro has been replaced by a
6953 function that can join an arbitrary number of extra strings onto a growing
6956 (c) The code for expansion conditions now uses a table and a binary chop
6957 instead of a serial search (which was left over from when there were very
6958 few conditions). Also, it now recognizes conditions like "pam" even when
6959 the relevant support is not compiled in: a suitably worded error message is
6960 given if an attempt is made to use such a condition.
6962 11. Added ${time_interval:xxxxx}.
6964 12. A bug was causing one of the ddress fields not to be passed back correctly
6965 from remote delivery subprocesses. The field in question was not being
6966 subsequently used, so this caused to problems in practice.
6968 13. Added new log selectors queue_time and deliver_time.
6970 14. Might have fixed a bug in maildirsizefile handling that threw up
6971 "unexpected character" debug warnings, and recalculated the data
6972 unnecessarily. In any case, I expanded the warning message to give more
6975 15. Added the message "Restricted characters in address" to the statements in
6976 the default ACL that block characters like @ and % in local parts.
6978 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
6979 Three changes have been made:
6981 (a) There was a serious bug; a negative response to MAIL caused the whole
6982 recipient domain to be cached as invalid, thereby blocking all messages
6983 to all local parts at the same domain, from all senders. This bug has
6984 been fixed. The domain is no longer cached after a negative response to
6985 MAIL if the sender used is not empty.
6987 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
6990 (c) A new callout option, "use_sender" has been added for people who want
6991 the modified behaviour.
6997 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7000 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7001 indeed breaks things for older releases.
7003 3. Added additional logging to the case where there is a problem reading data
7004 from a filter that is running in a subprocess using a pipe, in order to
7005 try to track down a specific problem.
7007 4. Testing facility fudge: when running in the test harness and attempting
7008 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7009 getting "No route to host". Convert this to a timeout.
7011 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7014 6. Some OS don't have socklen_t but use size_t instead. This affects the
7015 fifth argument of getsockopt() amongst other things. This is now
7016 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7017 can be set for individual OS. I have set it for SunOS5, OSF1, and
7018 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7019 some earlier ones do not.
7021 7. Change 4.30/15 was not doing the test caselessly.
7023 8. The standard form for an IPv6 address literal was being rejected by address
7024 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7025 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7026 this, as well as the form without the "IPv6" on the front (but only when
7027 address literals are enabled, of course).
7029 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7031 10. Exim crashed if a message with an empty sender address specified by -f
7032 encountered a router with an errors_to setting. This could be provoked only
7033 by a command such as
7037 where an empty string was supplied; "<>" did not hit this bug.
7039 11. Installed PCRE release 4.5.
7041 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7042 remained set. It is now erased.
7044 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7045 times from message ids (which are base 36 rather than the normal 62).
7047 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7048 were being counted as actual protocol errors, and logged if the log
7049 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7050 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7051 and DATA following a set of rejected RCPTs do not count as protocol errors.
7052 In other words, Exim assumes they were pipelined, though this may not
7053 actually be the case. Of course, in all cases the client gets an
7054 appropriate error code.
7056 15. If a lookup fails in an ACL condition, a message about the failure may
7057 be available; it is used if testing the ACL cannot continue, because most
7058 such messages specify what the cause of the deferral is. However, some
7059 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7060 that caused an old message to be retained and used if a later statement
7061 caused a defer, replacing the real cause of the deferral.
7063 16. If an IP address had so many PTR records that the DNS lookup buffer
7064 was not large enough to hold them, Exim could crash while trying to process
7065 the truncated data. It now detects and logs this case.
7067 17. Further to 4.21/58, another change has been made: if (and only if) the
7068 first line of a message (the first header line) ends with CRLF, a bare LF
7069 in a subsequent header line has a space inserted after it, so as not to
7070 terminate the header.
7072 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7073 unnecessarily, used atoi() instead of strtol(), and didn't check the
7074 termination when getting file sizes from file names by regex.
7076 19. Completely re-implemented the support for maildirsize files, in the light
7077 of a number of problems with the previous contributed implementation
7078 (4.30/29). In particular:
7080 . If the quota is zero, the maildirsize file is maintained, but no quota is
7083 . If the maildir directory does not exist, it is created before any attempt
7084 to write a maildirsize file.
7086 . The quota value in the file is just a cache; if the quota is changed in
7087 the transport, the new value overrides.
7089 . A regular expression is available for excluding directories from the
7092 20. The autoreply transport checks the characters in options that define the
7093 message's headers; it allows continued headers, but it was checking with
7094 isspace() after an embedded newline instead of explicitly looking for a
7097 21. If all the "regular" hosts to which an address was routed had passed their
7098 expiry times, and had not reached their retry times, the address was
7099 bounced, even if fallback hosts were defined. Now Exim should go on to try
7102 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7103 equivalent code in the SMTP transport. Some hosts send humungous responses
7104 to HELO/EHLO, more than 1024 it seems.
7106 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7107 gives compiler warnings in some environments. I've now done it "properly",
7110 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7111 (because of problems with the built-in one) was declared to return uschar *
7112 instead of char *, causing compiler failure.
7114 25. Fixed a file descriptor leak when processing alias/forward files.
7116 26. Fixed a minor format string issue in dbfn.c.
7118 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7120 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7121 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7124 29. When a sender address is verified, it is cached, to save repeating the test
7125 when there is more than one recipient in a message. However, when the
7126 verification involves a callout, it is possible for different callout
7127 options to be set for different recipients. It is too complicated to keep
7128 track of this in the cache, so now Exim always runs a verification when a
7129 callout is required, relying on the callout cache for the optimization.
7130 The overhead is duplication of the address routing, but this should not be
7133 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7134 to be verified with callout=postmaster, and the main callout worked but the
7135 postmaster check failed, the verification correctly failed. However, if a
7136 subsequent RCPT command asked for sender verification *without* the
7137 postmaster check, incorrect caching caused this verification also to fail,
7140 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7141 it was not caching the DNS options (qualify_single, search_parents) that
7142 were used when the lookup failed. A subsequent lookup with different
7143 options therefore always gave the same answer, though there were cases
7144 where it should not have. (Example: a "domains = !$mx_any" option on a
7145 dnslookup router: the "domains" option is always processed without any
7146 widening, but the router might have qualify_single set.) Now Exim uses the
7147 cached value only when the same options are set.
7149 32. Added John Jetmore's "exipick" utility to the distribution.
7151 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7152 than a timeout (e.g. a certificate is required, and is not provided), an
7153 Exim server now closes the connection immediately. Previously it waited for
7154 the client to close - but if the client is SSL, it seems that they each
7155 wait for each other, leading to a delay before one of them times out.
7157 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7158 maintained 0.8.x compatibility because I don't think many are using it, and
7159 it is clearly obsolete.
7161 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7164 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7165 client certificate was expired. A simple patch fixes this, though I don't
7166 understand the full logic of why the verify callback is called multiple
7169 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7170 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7171 which causes problems with some clients (such as the Certicom SSL Plus
7172 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7173 disables the coutermeasure allowing Eudora to connect."
7175 38. Exim was not checking that a write() to a log file succeeded. This could
7176 lead to Bad Things if a log got too big, in particular if it hit a file
7177 size limit. Exim now panics and dies if it cannot write to a log file, just
7178 as it does if it cannot open a log file.
7180 39. Modified OS/Makefile-Linux so that it now contains
7182 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7184 The two -D definitions ensure that Exim is compiled with large file
7185 support, which makes it possible to handle log files that are bigger than
7188 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7189 instance) a domain was checked against a named list that involved a lookup,
7190 causing $domain_data to be set, then another domain was checked against the
7191 same list, then the first domain was re-checked, the value of $domain_data
7192 after the final check could be wrong. In particular, if the second check
7193 failed, it could be set empty. This bug probably also applied to
7196 41. The strip_trailing_dot option was not being applied to the address given
7197 with the -f command-line option.
7199 42. The code for reading a message's header from the spool was incrementing
7200 $received_count, but never initializing it. This meant that the value was
7201 incorrect (doubled) while delivering a message in the same process in which
7202 it was received. In the most common configuration of Exim, this never
7203 happens - a fresh exec is done - but it can happen when
7204 deliver_drop_privilege is set.
7206 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7207 it now includes up to 150 characters of the unexpected data in the log
7210 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7211 and building data strings. The size of both of these buffers was 10 000
7212 bytes - far larger than anybody would *ever* want, thought I. Needless to
7213 say, somebody hit the limit. I have increased the maximum line length to
7214 20 000 and the maximum data length of concatenated lines to 100 000. I have
7215 also fixed two bugs, because there was no checking on these buffers. Tsk,
7216 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7217 buffer is too small.
7219 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7220 lsearch lookups. Now it does.
7222 46. When parsing a route_list item in a manualroute router, a fixed-length
7223 buffer was used for the list of hosts. I made this 1024 bytes long,
7224 thinking that nobody would ever have a list of hosts that long. Wrong.
7225 Somebody had a whole pile of complicated expansion conditions, and the
7226 string was silently truncated, leading to an expansion error. It turns out
7227 that it is easier to change to an unlimited length (owing to other changes
7228 that have happened since this code was originally written) than to build
7229 structure for giving a limitation error. The length of the item that
7230 expands into the list of hosts is now unlimited.
7232 47. The lsearch lookup could not handle data where the length of text line was
7233 more than 4095 characters. Such lines were truncated, leading to shortened
7234 data being returned. It should now handle lines of any length.
7236 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7237 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7240 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7241 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7242 no longer happens in comment lines. A list of replacements is now placed
7243 at the head of all of the source files, except those whose only change is
7244 to replace PERL_COMMAND in the very first #! line.
7246 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7247 messages on the queue, with a bottom-up merge sort, using code contributed
7248 by Michael Haardt. This should make operations like -bp somewhat faster on
7249 large queues. It won't affect queue runners, except when queue_run_in_order
7252 51. Installed eximstats 1.31 in the distribution.
7254 52. Added support for SRV lookups to the dnslookup router.
7256 53. If an ACL referred to $message_body or $message_body_end, the value was not
7257 reset for any messages that followed in the same SMTP session.
7259 54. The store-handling optimization for building very long strings was not
7260 differentiating between the different store pools. I don't think this
7261 actually made any difference in practice, but I've tidied it.
7263 55. While running the routers to verify a sender address, $sender_address
7264 was still set to the sender address. This is wrong, because when routing to
7265 send a bounce to the sender, it would be empty. Therefore, I have changed
7266 it so that, while verifying a sender address, $sender_address is set to <>.
7267 (There is no change to what happens when verifying a recipient address.)
7269 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7270 target A or AAAA records (if not already returned) without resetting the
7271 qualify_single or search_parents options of the DNS resolver. These are
7272 inappropriate in this case because the targets of MX and SRV records must
7273 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7274 target that, when qualified, matched something in the local domain. These
7275 two options are now turned off when doing these lookups.
7277 57. It seems that at least some releases of Reiserfs (which does not have the
7278 concept of a fixed number of inodes) returns zero and not -1 for the
7279 number of available inodes. This interacted badly with check_spool_inodes,
7280 which assumed that -1 was the "no such thing" setting. What I have done is
7281 to check that the total number of inodes is greater than zero before doing
7282 the test of how many are available.
7284 58. When a "warn" ACL statement has a log_message modifier, the message is
7285 remembered, and not repeated. This is to avoid a lot of repetition when a
7286 message has many recipients that cause the same warning to be written.
7287 However, Exim was preserving the list of already written lines for an
7288 entire SMTP session, which doesn't seem right. The memory is now reset if a
7289 new message is started.
7291 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7292 debugging output unless log_rewrite was also set.
7294 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7295 of (int)(handle) when we know that handle contains (void *)(-1).
7297 61. The Exim daemon panic-logs an error return when it closes the incoming
7298 connection. However "connection reset by peer" seems to be common, and
7299 isn't really an error worthy of noting specially, so that particular error
7302 62. When Exim is trying to find all the local interfaces, it used to panic and
7303 die if the ioctl to get the interface flags failed. However, it seems that
7304 on at least one OS (Solaris 9) it is possible to have an interface that is
7305 included in the list of interfaces, but for which you get a failure error
7306 for this call. This happens when the interface is not "plumbed" into a
7307 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7308 failure of the "get flags" call assumes that the interface is down.
7310 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7311 makes life easier for people who are doing arithmetic on fields extracted
7312 from dates, where you often get leading zeros that should not be
7313 interpreted as octal.
7315 64. Added qualify_domain to the redirect router, to override the global
7318 65. If a pathologically long header line contained very many addresses (the
7319 report of this problem mentioned 10 000) and each of them was rewritten,
7320 Exim could use up a very large amount of memory. (It kept on making new
7321 copies of the header line as it rewrote, and never released the old ones.)
7322 At the expense of a bit more processing, the header rewriting function has
7323 been changed so that it no longer eats memory in this way.
7325 66. The generation of the Received: header has been moved from the time that a
7326 message starts to be received, to the time that it finishes. The timestamp
7327 in the Received: header should now be very close to that of the <= log
7328 line. There are two side-effects of this change:
7330 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7331 logged header lines no longer include the local Received: line, because
7332 it has not yet been created. The same applies to a copy of the message
7333 that is returned to a non-SMTP sender when a message is rejected.
7335 (b) When a filter file is tested using -bf, no additional Received: header
7336 is added to the test message. After some thought, I decided that this
7339 This change does not affect the value of $received_for. It is still set
7340 after address rewriting, but before local_scan() is called.
7342 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7344 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7345 gave an unhelpful panic error message, and a defer error. I have managed to
7346 change this behaviour so that it now rejects any supplied certificate,
7347 which seems right, as the list of acceptable certificates is empty.
7349 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7350 gave an unhelpful defer error. I have not managed to make this reject any
7351 supplied certificates, but the error message it gives is "no certificate
7352 supplied", which is not helpful.
7354 70. exigrep's output now also includes lines that are not associated with any
7355 message, but which match the given pattern. Implemented by a patch from
7356 Martin Sluka, which also tidied up the Perl a bit.
7358 71. Recipient callout verification, like sender verification, was using <> in
7359 the MAIL FROM command. This isn't really the right thing, since the actual
7360 sender may affect whether the remote host accepts the recipient or not. I
7361 have changed it to use the actual sender in the callout; this means that
7362 the cache record is now keyed on a recipient/sender pair, not just the
7363 recipient address. There doesn't seem to be a real danger of callout loops,
7364 since a callout by the remote host to check the sender would use <>.
7365 [SEE ABOVE: changed after hitting problems.]
7367 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7368 temporary errors. However, in the case of such a code being given after
7369 the end of a data transmission (i.e. after ".") Exim was failing to write
7370 a retry record for the message. (Yes, there was some broken host that was
7371 actually sending 8xx at this point.)
7373 73. An unknown lookup type in a host list could cause Exim to panic-die when
7374 the list was checked. (An example that provoked this was putting <; in the
7375 middle of a list instead of at the start.) If this happened during a DATA
7376 ACL check, a -D file could be left lying around. This kind of configuration
7377 error no longer causes Exim to die; instead it causes a defer error. The
7378 incident is still logged to the main and panic logs.
7380 74. Buglet left over from Exim 3 conversion. The message "too many messages
7381 in one connection" was written to the rejectlog but not the mainlog, except
7382 when address rewriting (yes!) was being logged.
7384 75. Added write_rejectlog option.
7386 76. When a system filter was run not as root (that is, when system_filter_user
7387 was set), the values of the $n variables were not being returned to the
7388 main process; thus, they were not subsequently available in the $sn
7391 77. Added +return_path_on_delivery log selector.
7393 78. A connection timeout was being treated differently from recipients deferred
7394 when testing hosts_max_try with a message that was older than the host's
7395 retry timeout. (The host should not be counted, thus allowing all hosts to
7396 be tried at least once before bouncing.) This may have been the cause of an
7397 occasionally reported bug whereby a message would remain on the queue
7398 longer than the retry timeout, but would be bounced if a delivery was
7399 forced. I say "may" because I never totally pinned down the problem;
7400 setting up timeout/retry tests is difficult. See also the next item.
7402 79. The ultimate address timeout was not being applied to errors that involved
7403 a combination of host plus message (for example, a timeout on a MAIL
7404 command). When an address resolved to a number of possible hosts, and they
7405 were not all tried for each delivery (e.g. because of hosts_max_try), a
7406 message could remain on the queue longer than the retry timeout.
7408 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7411 81. Fixed an obscure SMTP outgoing bug which required at least the following
7412 conditions: (a) there was another message waiting for the same server;
7413 (b) the server returned 5xx to all RCPT commands in the first message so
7414 that the message was not completed; (c) the server dropped the connection
7415 or gave a negative response to the RSET that Exim sends to abort the
7416 transaction. The observed case was a dropped connection after DATA that had
7417 been sent in pipelining mode. That is, the server had advertised PIPELINING
7418 but was not implementing it correctly. The effect of the bug was incorrect
7419 behaviour, such as trying another host, and this could lead to a crash.
7425 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7426 and daemon.c were passed as pointers to ints; they should have been
7427 pointers to socklen_t variables (which are typically unsigned ints).
7429 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7432 3. Fixed a really odd bug that affected only the testing scheme; patching a
7433 certain fixed string in the binary changed the value of another string that
7434 happened to be identical to the end of the original first string.
7436 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7437 name", it returns that address as the IP address. On some operating
7438 systems (e.g. Solaris), it also passes back the IP address string as the
7439 "host name". However, on others (e.g. Linux), it passes back an empty
7440 string. Exim wasn't checking for this, and was changing the host name to an
7441 empty string, assuming it had been canonicalized.
7443 5. Although rare, it is permitted to have more than one PTR record for a given
7444 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7445 all the names associated with an address, because they do in Solaris.
7446 However, it seems that they do not in Linux for data that comes from the
7447 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7448 I found this out when I moved to a new Linux workstation and tried to run
7449 the Exim test suite.
7451 To get round this problem I have changed the code so that it now does its
7452 own call to the DNS to look up PTR records when searching for a host name.
7453 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7454 addresses that are only in /etc/hosts are still found.
7456 This behaviour is, however, controlled by an option called host_lookup_
7457 order, which defaults to "bydns:byaddr". If people want to use the other
7458 order, or indeed, just use one or the other means of lookup, they can
7459 specify it in this variable.
7461 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7462 some operating systems, this comes back from gethostbyaddr() as an empty
7463 string, and this is what Exim used to test for. However, it seems that in
7464 other systems, "." is yielded. Exim now tests for this case too.
7466 7. The values of check_spool_space and check_log_space are now held internally
7467 as a number of kilobytes instead of an absolute number of bytes. If a
7468 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7469 kilobyte. This means that much larger values can be stored.
7471 8. Exim monitor: an attempt to get the action menu when not actually pointing
7472 at a message produces an empty menu entitled "No message selected". This
7473 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7474 no entries in it ("Shell widget menu has zero width and/or height"). So I
7475 have added a single, blank menu entry in this case.
7477 9. Added ${quote_local_part.
7479 10. MIME decoding is now applied to the contents of Subject: header lines when
7482 11. Now that a reference to $sender_host_address automatically causes a reverse
7483 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7484 host lookup before query-style lookups in lists that might use this
7485 variable. This has therefore been abolished, and the "net-" prefix is no
7486 longer necessary for query-style lookups.
7488 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7489 have been a typo for LFLAGS, so it has been changed.
7491 13. The install script calls Exim with "-C /dev/null" in order to find the
7492 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7493 to be output. However, since Exim outputs its version number before the
7494 error, it didn't break the script. It just looked ugly. I fixed this by
7495 always allowing "-C /dev/null" if the caller is root.
7497 14. Ignore overlarge ACL variable number when reading spool file - insurance
7498 against a later release with more variables having written the file.
7500 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7501 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7502 this, as well as the form without the "IPv6" on the front.
7504 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7505 OS/Makefile-Darwin file.
7507 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7508 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7510 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7511 variable. However, some OS use ints for inodes. I've added cast to long int
7512 to get rid of the compiler warning.
7514 19. I had forgotten to lock out "/../" in configuration file names when
7515 ALT_CONFIG_PREFIX was set.
7517 20. Routers used for verification do not need to specify transports. However,
7518 if such a router generated a host list, and callout was configured, Exim
7519 crashed, because it could not find a port number from the (non-existent)
7520 transport. It now assumes port 25 in this circumstance.
7522 21. Added the -t option to exigrep.
7524 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7525 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7526 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7527 them is set, LOOKUP_LSEARCH is forced.
7529 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7531 24. Added sender and host information to the "rejected by local_scan()" log
7532 line; previously there was no indication of these.
7534 25. Added .include_if_exists.
7536 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7537 Linux. It turns out that not all file systems support this. Apparently some
7538 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7539 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7540 not supported on the file descriptor, is now ignored when Exim is trying to
7541 sync a directory. This applies only to Linux.
7543 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7545 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7546 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7549 29. Added support for maildirsize files from supplied patch (modified a bit).
7551 30. The use of :fail: followed by an empty string could lead Exim to respond to
7552 sender verification failures with (e.g.):
7554 550 Verification failed for <xxx>
7555 550 Sender verify failed
7557 where the first response line was missing the '-' that indicates it is not
7558 the final line of the response.
7560 31. The loop for finding the name of the user that called Exim had a hardwired
7561 limit of 10; it now uses the value of finduser_retries, which is used for
7562 all other user lookups.
7564 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7567 33. Exim was neglecting to zero errno before one call of strtol() when
7568 expanding a string and expecting an integer value. On some systems this
7569 resulted in spurious "integer overflow" errors. Also, it was casting the
7570 result into an int without checking.
7572 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7573 did not work. The code looks as if it has *never* worked, though it appears
7574 to have been documented since at least release 1.62. I have made it work.
7576 35. The "timeout_DNS" error in retry rules, also documented since at least
7577 1.62, also never worked. As it isn't clear exactly what this means, and
7578 clearly it isn't a major issue, I have abolished the feature by treating it
7579 as "timeout", and writing a warning to the main and panic logs.
7581 36. The display of retry rules for -brt wasn't always showing the error code
7584 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7585 timeout_connect_A, timeout_connect_MX.
7587 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7588 to the empty sender.
7590 39. The daemon was not analysing the content of -oX till after it had closed
7591 stderr and disconnected from the controlling terminal. This meant that any
7592 syntax errors were only noted on the panic log, and the return code from
7593 the command was 0. By re-arranging the code a little, I've made the
7594 decoding happen first, so such errors now appear on stderr, and the return
7595 code is 1. However, the actual setting up of the sockets still happens in
7596 the disconnected process, so errors there are still only recorded on the
7599 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7600 connections (as happens on some IP stacks) was logged at start up time as
7601 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7602 it from "IPv6 and IPv4", which means that two separate sockets are being
7605 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7606 says whether AF_INET or AF_INET6 was passed as an argument.
7608 42. Exiwhat output was messed up when time zones were included in log
7611 43. Exiwhat now gives more information about the daemon's listening ports,
7612 and whether -tls-on-connect was used.
7614 44. The "port" option of the smtp transport is now expanded.
7616 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7617 silently ignored. Now an error message is written to the main and panic
7620 46. There's a new ACL modifier called "logwrite" which writes to a log file
7621 as soon as it is encountered.
7623 47. Added $local_user_uid and $local_user_gid at routing time.
7625 48. Exim crashed when trying to verify a sender address that was being
7628 49. Exim was recognizing only a space character after ".include". It now also
7629 recognizes a tab character.
7631 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7632 extracting the relevant information from the specification. The man page no
7633 longer contains scrambled data for the -d option, and I've added a section
7634 at the front about calling Exim under different names.
7636 51. Added "extra_headers" argument to the "mail" command in filter files.
7638 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7641 53. Installed eximstats 1.29.
7643 54. Added transport_filter_timeout as a generic transport option.
7645 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7646 Cc: header lines. This was required by RFC 822, but it not required by RFC
7649 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7650 incoming messages that did not have them. Now it does so only if the
7651 message originates locally, that is, if there is no associated remote host
7652 address. When Resent- header lines are present, this applies to the Resent-
7653 lines rather than the non-Resent- lines.
7655 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7656 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7658 58. Messages for configuration errors now include the name of the main
7659 configuration file - useful now that there may be more than one file in a
7660 list (.included file names were always shown).
7662 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7663 for those rare installations that do not start the daemon as root or run it
7664 setuid root. I've cut out the call to initgroups() if the daemon is not
7667 60. The Exim user and group can now be bound into the binary as text strings
7668 that are looked up at the start of Exim's processing.
7670 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7672 62. Added $mailstore_basename variable.
7674 63. Installed patch to sieve.c from Michael Haardt.
7676 64. When Exim failed to open the panic log after failing to open the main log,
7677 the original message it was trying to log was written to stderr and debug
7678 output, but if they were not available (the usual case in production), it
7679 was lost. Now it is written to syslog before the two lines that record the
7680 failures to open the logs.
7682 65. Users' Exim filters run in subprocesses under the user's uid. It is
7683 possible for a "deliver" command or an alias in a "personal" command to
7684 provoke an address rewrite. If logging of address rewriting is configured,
7685 this fails because the process is not running as root or exim. There may be
7686 a better way of dealing with this, but for the moment (because 4.30 needs
7687 to be released), I have disabled address rewrite logging when running a
7688 filter in a non-root, non-exim process.
7694 1. The buildconfig auxiliary program wasn't quoting the value set for
7695 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7696 not defined. This bug was masked in 4.22 by the effect that was fixed in
7699 2. Some messages that were rejected after a message id was allocated were
7700 shown as "incomplete" by exigrep. It no longer does this for messages that
7701 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7703 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7704 have allow_domain_literals set, the ID did not get logged in the <= line.
7705 Domain literals are now always recognized in Message-ID: header lines.
7707 4. The first argument for a ${extract expansion item is the key name or field
7708 number. Leading and trailing spaces in this item were not being ignored,
7709 causing some misleading effects.
7711 5. When deliver_drop_privilege was set, single queue runner processes started
7712 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7713 same command in the process it spins off) were not dropping privilege.
7715 6. When the daemon running as "exim" started a queue runner, it always
7716 re-executed Exim in the spun-off process. This is a waste of effort when
7717 deliver_drop_privilege is set. The new process now just calls the
7718 queue-runner function directly.
7724 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7727 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7728 ignored. Though the use of -oP was forcing the writing of a pid file, it
7729 was always written to the default place.
7731 3. If the message "no IP address found for host xxxx" is generated during
7732 incoming verification, it is now followed by identification of the incoming
7733 connection (so you can more easily find what provoked it).
7735 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7737 5. Added some features to "harden" Exim a bit more against certain attacks:
7739 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7740 be put in Local/Makefile. This is like the never_users runtime option,
7741 but it cannot be overridden. The default setting is "root".
7743 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7744 prefix string with which any file named in a -C command line option
7747 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7748 is retained for -C and -D only if the caller of Exim is root. Without
7749 it, the exim user may also use -C and -D and retain privilege.
7751 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7752 command line option is disabled.
7754 6. Macro names set by the -D option must start with an upper case letter, just
7755 like macro names defined in the configuration file.
7757 7. Added "dereference=" facility to LDAP.
7759 8. Two instances of the typo "uknown" in the source files are fixed.
7761 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7762 the Configure-Makefile script screwed up while processing it.
7764 10. Incorporated PCRE 4.4.
7766 11. The SMTP synchronization check was not operating right at the start of an
7767 SMTP session. For example, it could not catch a HELO sent before the client
7768 waited for the greeting. There is now a check for outstanding input at the
7769 point when the greeting is written. Because of the duplex, asynchronous
7770 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7771 way, but not yet received, when the check is performed.
7773 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7774 on TCP/IP sockets, because this apparently causes some broken clients to
7777 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7778 unchanged) from the Cygwin maintainer.
7780 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7781 was supported instead of testing for mbx. Effectively a typo.
7783 15. The spa authenticator server code was not checking that the input it
7784 received was valid base64.
7786 16. The debug output line for the "set" modifier in ACLs was not showing the
7787 name of the variable that was being set.
7789 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7791 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7792 lookup on its own. Something else was needed to trigger the lookup. For
7793 example, a match in host_lookup or the need for a host name in a host list.
7794 Now, if $sender_host_name is referenced and the host name has not yet been
7795 looked up, a lookup is performed. If the lookup fails, the variable remains
7796 empty, and $host_lookup_failed is set to "1".
7798 19. Added "eqi" as a case-independent comparison operator.
7800 20. The saslauthd authentication condition could segfault if neither service
7801 nor realm was specified.
7803 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7804 error message that was logged was misleading, and incoming SMTP
7805 connections were dropped. The message is now more accurate, and temporary
7806 errors are given to SMTP connections.
7808 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7809 (except RSET) with a 421 error, until QUIT is received. However, it was
7810 failing to send a response to QUIT.
7812 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7813 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7814 ACL is now run after the helo_try_verify_hosts code.
7816 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7817 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7818 case-independent, so other case variants are also recognized). Apparently
7819 some systems use these upper case variants.
7821 25. If more than two messages were waiting for the same host, and a transport
7822 filter was specified for the transport, Exim sent two messages over the
7823 same TCP/IP connection, and then failed with "socket operation on non-
7824 socket" when it tried to send the third.
7826 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7828 27. The extern definition of crypt16() in expand.c was not being excluded when
7829 the OS had its own crypt16() function.
7831 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7832 as a preferred synonym for return_size_limit, both as an option and as an
7835 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7837 30. Changed the default configuration ACL to relax the local part checking rule
7838 for addresses that are not in any local domains. For these addresses,
7839 slashes and pipe symbols are allowed within local parts, but the sequence
7840 /../ is explicitly forbidden.
7842 31. SPA server authentication was not clearing the challenge buffer before
7845 32. log_message in a "warn" ACL statement was writing to the reject log as
7846 well as to the main log, which contradicts the documentation and doesn't
7847 seem right (because no rejection is happening). So I have stopped it.
7849 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7850 However, I am unable to do any testing of this.
7852 34. Fixed an infelicity in the appendfile transport. When checking directories
7853 for a mailbox, to see if any needed to be created, it was accidentally
7854 using path names with one or more superfluous leading slashes; tracing
7855 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7857 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7858 contents are added to the log line that is written for every discarded
7859 recipient. (Previously a log_message setting was ignored.)
7861 36. The ${quote: operator now quotes the string if it is empty.
7863 37. The install script runs exim in order to find its version number. If for
7864 some reason other than non-existence or emptiness, which it checks, it
7865 could not run './exim', it was installing it with an empty version number,
7866 i.e. as "exim-". This error state is now caught, and the installation is
7869 38. An argument was missing from the function that creates an error message
7870 when Exim fails to connect to the socket for saslauthd authentication.
7871 This could cause Exim to crash, or give a corrupted message.
7873 39. Added isip, isip4, and isip6 to ${if conditions.
7875 40. The ACL variables $acl_xx are now saved with the message, and can be
7876 accessed later in routers, transports, and filters.
7878 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7879 strings in the file are not string-expanded.
7881 42. If a MAIL command specified a SIZE value that was too large to fit into an
7882 int variable, the check against message_size_limit failed. Such values are
7883 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7884 day this will have to be increased, but I don't think I want to be around
7885 when emails are that large.
7892 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7893 iconv() is not standard in FreeBSD.
7895 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7896 IPv6 enabled. The observed symptom was a segmentation fault on return from
7897 the function os_common_find_running_interfaces() in src/os.c.
7899 3. In the check_special_case() function in daemon.c I had used "errno" as an
7900 argument name, which causes warnings on some systems. This was basically a
7901 typo, since it was named "eno" in the comments!
7903 4. The code that waits for the clock to tick (at a resolution of some fraction
7904 of a second) so as to ensure message-id uniqueness was always waiting for
7905 at least one whole tick, when it could have waited for less. [This is
7906 almost certainly not relevant at current processor speeds, where it is
7907 unlikely to ever wait at all. But we try to future-proof.]
7909 5. The function that sleeps for a time interval that includes fractions of a
7910 second contained a race. It did not block SIGALRM between setting the
7911 timer, and suspending (a couple of lines later). If the interval was short
7912 and the sigsuspend() was delayed until after it had expired, the suspension
7913 never ended. On busy systems this could lead to processes getting stuck for
7916 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7917 process, before it forks any delivery processes. The open lookup caching
7918 mechanism meant that the open file or database connection was passed into
7919 the delivery process. The problem was that delivery processes always tidy
7920 up cached lookup data. This could cause a problem for the next delivery
7921 process started by the queue runner, because the external queue runner
7922 process does not know about the closure. So the next delivery process
7923 still has data in the lookup cache. In the case of a file lookup, there was
7924 no problem because closing a file descriptor in a subprocess doesn't affect
7925 the parent. However, if the lookup was caching a connection to a database,
7926 the connection was closed, and the second delivery process was likely to
7927 see errors such as "PGSQL: query failed: server closed the connection
7928 unexpectedly". The problem has been fixed by closing all cached lookups
7929 in a queue runner before running a delivery process.
7931 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
7932 seem to have the "const" qualifier which it has on other OS. I've
7935 8. Change 4.21/2 was too strict. It is only if there are two authenticators
7936 *of the same type* (client or server) with the same public name that an
7937 error should be diagnosed.
7939 9. When Exim looked up a host name for an IP address, but failed to find the
7940 original IP address when looking up the host name (a safety check), it
7941 output the message "<ip address> does not match any IP for NULL", which was
7942 confusing, to say the least. The bug was that the host name should have
7943 appeared instead of "NULL".
7945 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
7946 user that is built into the binary, and the -C or -D options is used, root
7947 privilege is dropped before the configuration file is read. In addition,
7948 logging is switched to stderr instead of the normal log files. If the
7949 configuration then re-defines the Exim user, the unprivileged environment
7950 is probably not what is expected, so Exim logs a panic warning message (but
7953 However, if deliver_drop_privilege is set, the unprivileged state may well
7954 be exactly what is intended, so the warning has been cut out in that case,
7955 and Exim is allowed to try to write to its normal log files.
7961 1. smtp_return_error_details was not giving details for temporary sender
7962 or receiver verification errors.
7964 2. Diagnose a configuration error if two authenticators have the same public
7967 3. Exim used not to create the message log file for a message until the first
7968 delivery attempt. This could be confusing when incoming messages were held
7969 for policy or load reasons. The message log file is now created at the time
7970 the message is received, and an initial "Received" line is written to it.
7972 4. The automatically generated man page for command line options had a minor
7973 bug that caused no ill effects; however, a more serious problem was that
7974 the procedure for building the man page automatically didn't always
7975 operate. Consequently, release 4.20 contains an out-of-date version. This
7976 shouldn't happen again.
7978 5. When building Exim with embedded Perl support, the script that builds the
7979 Makefile was calling 'perl' to find its compile-time parameters, ignoring
7980 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
7982 6. The freeze_tell option was not being used for messages that were frozen on
7983 arrival, either by an ACL or by local_scan().
7985 7. Added the smtp_incomplete_transaction log selector.
7987 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
7988 was accepting AUTH without a new EHLO.
7990 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
7991 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
7994 10. Logging of TCP/IP connections (when configured) now happens in the main
7995 daemon process instead of the child process, so that the TCP/IP connection
7996 count is more accurate (but it can never be perfect).
7998 11. The use of "drop" in a nested ACL was not being handled correctly in the
7999 outer ACL. Now, if condition failure induced by the nested "drop" causes
8000 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8001 or "require"), the connection is dropped.
8003 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8004 that yield "discard" can now be used with an "accept" or a "discard" verb,
8005 but an error is generated for any others (because I can't see a useful way
8006 to define what should happen).
8008 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8009 are now processed in the same way as lines in the Exim configuration file.
8010 In particular, continuation lines are supported.
8012 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8014 15. Added -ti meaning -t -i.
8016 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8017 domains, and warn by logging if others are found.
8019 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8020 returned from GIFCONF when Exim is trying to find the list of interfaces on
8021 a host. The code in os.c has been modified to copy each ifreq to an aligned
8022 structure in all cases.
8024 Also, in some cases, the returned ifreq's were being copied to a 'struct
8025 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8026 means the last couple of bytes of an IPv6 address could be chopped if the
8027 ifreq contained only a normal sockaddr (14 bytes storage).
8029 18. Named domain lists were not supported in the hosts_treat_as_local option.
8030 An entry such as +xxxx was not recognized, and was treated as a literal
8033 19. Ensure that header lines added by a DATA ACL are included in the reject log
8034 if the ACL subsequently rejects the message.
8036 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8037 MD5 (which is deprecated).
8039 21. When testing a filter file using -bf, Exim was writing a message when it
8040 took the sender from a "From " line in the message, but it was not doing so
8041 when it took $return_path from a Return-Path: header line. It now does.
8043 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8044 with a valid header line field name (a series of printing characters
8045 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8047 23. Changed "disc" in the source to "disk" to conform to the documentation and
8048 the book and for uniformity.
8050 24. Ignore Sendmail's -Ooption=value command line item.
8052 25. When execve() failed while trying to run a command in a pipe transport,
8053 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8054 could be confused with a return value of 69 from the command itself. This
8055 has been changed to 127, the value the shell returns if it is asked to run
8056 a non-existent command. The wording for the related log line suggests a
8057 non-existent command as the problem.
8059 26. If received_header_text expands to an empty string, do not add a Received:
8060 header line to the message. (Well, it adds a token one on the spool, but
8061 marks it "old" so that it doesn't get used or transmitted.)
8063 27. Installed eximstats 1.28 (addition of -nt option).
8065 28. There was no check for failure on the call to getsockname() in the daemon
8066 code. This can fail if there is a shortage of resources on the system, with
8067 ENOMEM, for example. A temporary error is now given on failure.
8069 29. Contrary to the C standard, it seems that in some environments, the
8070 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8071 program. Exim now does this explicitly; it affects the formatting of
8072 timestamps using strftime().
8074 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8075 complaints. I've now initialized all the variables, to avoid this.
8077 32. Header lines added by a system filter were not being "seen" during
8078 transport-time rewrites.
8080 33. The info_callback() function passed to OpenSSL is set up with type void
8081 (*)(SSL *, int, int), as described somewhere. However, when calling the
8082 function (actually a macro) that sets it up, the type void(*)() is
8083 expected. I've put in a cast to prevent warnings from picky compilers.
8085 34. If a DNS black list lookup found a CNAME record, but there were no A
8086 records associated with the domain it pointed at, Exim crashed.
8088 35. If a DNS black list lookup returned more than one A record, Exim ignored
8089 all but the first. It now scans all returned addresses if a particular IP
8090 value is being sought. In this situation, the contents of the
8091 $dnslist_value variable are a list of all the addresses, separated by a
8094 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8095 to accept a host name and all its aliases if the forward lookup for any of
8096 them yielded the IP address of the incoming connection. Now it accepts only
8097 those names whose forward lookup yields the correct IP address. Any other
8098 names are discarded. This closes a loophole whereby a rogue DNS
8099 administrator could create reverse DNS records to break through a
8100 wildcarded host restriction in an ACL.
8102 37. If a user filter or a system filter that ran in a subprocess used any of
8103 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8104 the wrong values were passed to the pipe command ($thisaddress had the
8105 value of $0, $0 had the value of $1, etc). This bug was introduced by
8106 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8108 38. Improved the line breaking for long SMTP error messages from ACLs.
8109 Previously, if there was no break point between 40 and 75 characters, Exim
8110 left the rest of the message alone. Two changes have been made: (a) I've
8111 reduced the minimum length to 35 characters; (b) if it can't find a break
8112 point between 35 and 75 characters, it looks ahead and uses the first one
8113 that it finds. This may give the occasional overlong line, but at least the
8114 remaining text gets split now.
8116 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8117 file descriptors might be low, and that setting 1000 would always raise it.
8118 It turns out that in some environments, the limit is already over 1000 and
8119 that lowering it causes trouble. So now Exim takes care not to decrease it.
8121 40. When delivering a message, the value of $return_path is set to $sender_
8122 address at the start of routing (routers may change the value). By an
8123 oversight, this default was not being set up when an address was tested by
8124 -bt or -bv, which affected the outcome if any router or filter referred to
8127 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8128 log only when "message" or "log_message" are set. However, if one of the
8129 conditions was an address verification, or a call to a nested ACL, the
8130 messages generated by the underlying test were being passed through. This
8131 no longer happens. The underlying message is available in $acl_verify_
8132 message for both "message" and "log_message" expansions, so it can be
8133 passed through if needed.
8135 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8136 new expansion $bh_ to give the encoded byte string without charset
8137 translation. Translation happens only if iconv() is available; HAVE_ICONV
8138 indicates this at build time. HEADERS_CHARSET gives the charset to
8139 translate to; headers_charset can change it in the configuration, and
8140 "headers charset" can change it in an individual filter file.
8142 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8143 that creates RFC 2047 encoded "words" labels them as that charset instead
8144 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8145 expansion operator; (ii) when Exim creates a From: line for a local
8146 message; (iii) when a header line is rewritten to include a "phrase" part.
8148 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8149 buggy, causing it to skip the first lines of messages whose message ID
8150 ended in 'D'. This would not have bitten before Exim release 4.14, because
8151 message IDs were unlikely to end in 'D' before then. The effect was to have
8152 incorrect size information for certain domains.
8154 45. #include "config.h" was missing at the start of the crypt16.c module. This
8155 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8158 46. If there was a timeout during a "random" callout check, Exim treated it as
8159 a failure of the random address, and carried on sending RSET and the real
8160 address. If the delay was just some slowness somewhere, the response to the
8161 original RCPT would be taken as a response to RSET and so on, causing
8162 mayhem of various kinds.
8164 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8165 when I implemented it. It didn't allow for the fact that some option values
8166 may legitimately be negative (e.g. size_addition), and it didn't even do
8167 the right test for positive values.
8169 48. Domain names in DNS records are case-independent. Exim always looks them up
8170 in lower case. Some resolvers return domain names in exactly the case they
8171 appear in the zone file, that is, they may contain uppercase letters. Not
8172 all resolvers do this - some return always lower case. Exim was treating a
8173 change of case by a resolver as a change of domain, similar to a widening
8174 of a domain abbreviation. This triggered its re-routing code and so it was
8175 trying to route what was effectively the same domain again. This normally
8176 caused routing to fail (because the router wouldn't handle the domain
8177 twice). Now Exim checks for this case specially, and just changes the
8178 casing of the domain that it ultimately uses when it transmits the message
8181 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8184 50. If a filter generated a file delivery with a non-absolute name (possible if
8185 no home directory exists for the router), the forbid_file option was not
8188 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8189 the existing equality matching.
8191 52. Exim was using ints instead of ino_t variables in some places where it was
8192 dealing with inode numbers.
8194 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8195 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8196 called TMPDIR, and if it finds it is different, it changes its value.
8198 54. The smtp_printf() function is now made available to local_scan() so
8199 additional output lines can be written before returning. There is also an
8200 smtp_fflush() function to enable the detection of a dropped connection.
8201 The variables smtp_input and smtp_batched_input are exported to
8204 55. Changed the default runtime configuration: the message "Unknown user"
8205 has been removed from the ACL, and instead placed on the localuser router,
8206 using the cannot_route_message feature. This means that any verification
8207 failures that generate their own messages won't get overridden. Similarly,
8208 the "Unrouteable address" message that was in the ACL for unverifiable
8209 relay addresses has also been removed.
8211 56. Added hosts_avoid_esmtp to the smtp transport.
8213 57. The exicyclog script was not checking for the esoteric option
8214 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8215 will work only if exicyclog is run under the appropriate euid.
8217 58. Following a discussion on the list, the rules by which Exim recognises line
8218 endings on incoming messages have been changed. The -dropcr and drop_cr
8219 options are now no-ops, retained only for backwards compatibility. The
8220 following line terminators are recognized: LF CRLF CR. However, special
8221 processing applies to CR:
8223 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8224 nor a local message in the state where . is a terminator.
8226 (ii) If a bare CR is encountered in a header line, an extra space is added
8227 after the line terminator so as not to end the header. The reasoning
8228 behind this is that bare CRs in header lines are most likely either
8229 to be mistakes, or people trying to play silly games.
8231 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8232 was being incorrectly given as 18 bytes larger than it should have been.
8233 This is a VOB (very old bug).
8235 60. This may never have affected anything current, but just in case it has:
8236 When the local host is found other than at the start of a list of hosts,
8237 the local host, those with the same MX, and any that follow, are discarded.
8238 When the list in question was part of a longer list of hosts, the following
8239 hosts (not currently being processed) were also being discarded. This no
8240 longer happens. I'm not sure if this situation could ever has previously
8243 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8246 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8247 header. This is something that is recommended in a new Internet Draft, and
8248 is something that is documented as being done by Sendmail. There are two
8249 possible values. For messages generated by the autoreply transport, Exim
8252 Auto-Submitted: auto-replied
8254 whereas for all other generated messages (e.g. bounces) it adds
8256 Auto-Submitted: auto-generated
8258 63. The "personal" condition in filters now includes a test for the
8259 Auto-Submitted: header. If it contains the string "auto-" the message it
8260 not considered personal.
8262 64. Added rcpt_include_affixes as a generic transport option.
8264 65. Added queue_only_override (default true).
8266 66. Added the syslog_duplication option.
8268 67. If what should have been the first header line of a message consisted of
8269 a space followed by a colon, Exim was mis-interpreting it as a header line.
8270 It isn't of course - it is syntactically invalid and should therefore be
8271 treated as the start of the message body. The misbehaviour could have
8272 caused a number of strange effects, including loss of data in subsequent
8273 header lines, and spool format errors.
8275 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8276 client host had authenticated. This control can now be exercised by an ACL
8277 for more flexibility.
8279 69. By default, callouts do not happen when testing with -bh. There is now a
8280 variant, -bhc, which does actually run the callout code, including
8281 consulting and updating the callout cache.
8283 70. Added support for saslauthd authentication, courtesy of Alexander
8286 71. If statvfs() failed on the spool or log directories while checking their
8287 size for availability, Exim confusingly gave the error "space shortage".
8288 Furthermore, in debugging mode it crashed with a floating point exception.
8289 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8290 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8291 serious problem, Exim now writes to the main and panic logs when this
8292 happens, with details of the failure. It then refuses to accept the
8293 incoming message, giving the message "spool directory problem" or "log
8294 directory problem" with a 421 code for SMTP messages.
8296 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8297 0, 1, and 2 exist, because some OS complain for execs without them (see
8298 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8299 descriptors. However, the code omitted to check that the open succeeded,
8300 causing mysterious errors if for some reason the permissions on /dev/null
8301 got screwed. Now Exim writes a message to the main and panic logs, and
8302 bombs out if it can't open /dev/null.
8304 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8305 interact so that it is all more flexible. It is supposed to remain
8306 backwards compatible. Also added extra_local_interfaces.
8308 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8309 to bomb out with an assertion failure - to the client this appears as a
8310 connection drop. This problem occurs in the part of the code that was taken
8311 from the Samba project. Fortunately, the assertion is in a very simple
8312 function, so I have fixed this by reproducing the function inline in the
8313 one place where it is called, and arranging for authentication to fail
8314 instead of killing the process with assert().
8316 75. The SPA client code was not working when the server requested OEM rather
8317 than Unicode encoding.
8319 76. Added code to make require_files with a specific uid setting more usable in
8320 the case where statting the file as root fails - usually a non-root-mounted
8321 NFS file system. When this happens and the failure is EACCES, Exim now
8322 forks a subprocess and does the per-uid checking as the relevant uid.
8324 77. Added process_log_path.
8326 78. If log_file_path was not explicitly set, a setting of check_log_space or
8327 check_log_inodes was ignored.
8329 79. If a space check for the spool or log partitions fails, the incident is now
8330 logged. Of course, in the latter case the data may get lost...
8332 80. Added the %p formatting code to string_format() so that it can be used to
8333 print addresses in debug_print(). Adjusted all the address printing in the
8334 debugging in store.c to use %p rather than %d.
8336 81. There was a concern that a line of code in smtp_in.c could overflow a
8337 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8338 initially expressed, the concern was not well-founded, because trailing
8339 spaces are removed early. However, if the trailing spaces were followed by
8340 a NULL, they did not get removed, so the overflow was possible. Two fixes
8343 (a) I re-wrote the offending code in a cleaner fashion.
8344 (b) If an incoming SMTP command contains a NULL character, it is rejected
8347 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8348 runs initgroups(), so that if the Exim user is in any additional groups,
8349 they will be used during message reception.
8355 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob