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
84 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
85 was exactly sized compared to the log buffer, a crash occurred with the
86 misleading message "bad memory reference; pool not found".
87 Found and traced by Jasen Betts.
89 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
90 dns_again_means_nonexist included an element causing a DNS lookup which
91 iteslf returned DNS_AGAIN, unbounded recursion occurred. Possible results
92 included (though probably not limited to) a process crash from stack
93 memory limit, or from excessive open files. Replace this with a paniclog
94 whine (as this is likely a configuration error), and returning
97 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
98 this always failed, probably leading to the usual downgrade to in-clear
101 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
102 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
103 to not checking that list for these lookups.
105 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
108 JH/23 Fix crash in string expansions. Previously, if an empty variable was
109 immediately followed by an expansion operator, a null-indirection read
110 was done, killing the process.
112 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
113 include an SMTP response string which is longer than that supported
114 by the delivering transport. Alleviate by wrapping such lines before
117 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
118 chars (RFC limit). Previously a limit of 12 items was made, which with
119 a not-impossible References: in the message being bounced could still
120 be over-large and get stopped in the transport.
122 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
123 close. Previously a bare socket close was done.
125 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
128 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
129 message being considered for delivery was already being handled by
130 another process, and having an SMTP connection already open, the function
131 to close it tried to use an uninitialized variable. This would afftect
132 high-volume sites more, especially when running mailing-list-style loads.
133 Pollution of logs was the major effect, as the other process delivered
134 the message. Found and partly investigated by Graeme Fowler.
136 JH/29 Change format of the internal ID used for message identification. The old
137 version only supported 31 bits for a PID element; the new 64 (on systems
138 which can use Base-62 encoding, which is all currently supported ones
139 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
140 and must use Base-36). The new ID is 23 characters rather than 16, and is
141 visible in various places - notably logs, message headers, and spool file
142 names. Various of the ancillary utilities also have to know the format.
143 As well as the expanded PID portion, the sub-second part of the time
144 recorded in the ID is expanded to support finer precision. Theoretically
145 this permits a receive rate from a single comms channel of better than the
147 The major timestamp part of the ID is not changed; at 6 characters it is
148 usable until about year 3700.
149 Updating from previously releases is fully supported: old-format spool
150 files are still usable, and the utilities support both formats. New
151 message will use the new format. The one hints-DB file type which uses
152 message-IDs (the transport wait- DB) will be discarded if an old-format ID
153 is seen; new ones will be built with only new-format IDs.
154 Optionally, a utility can be used to convert spool files from old to new,
155 but this is only an efficiency measure not a requirement for operation
156 Downgrading from new to old requires running a provided utility, having
157 first stopped all operations. This will convert any spool files from new
158 back to old (losing time-precision and PID information) and remove any
159 wait- hints databases.
161 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
162 we treated them as item separators when parsing for a list item, but they
163 need to be protected by the doublequotes. While there, add handling for
169 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
170 after reception to before a subsequent reception. This should
171 mean slightly faster delivery, and also confirmation of reception
174 JH/02 Move from using the pcre library to pcre2. The former is no longer
175 being developed or supported (by the original developer).
177 JH/03 Constification work in the filters module required a major version
178 bump for the local-scan API. Specifically, the "headers_charset"
179 global which is visible via the API is now const and may therefore
180 not be modified by local-scan code.
182 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
183 sendfile() didi not account for the way the ClamAV driver code called it.
185 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
186 time check was being done for every character; replace that with one
189 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
190 sent was prefixed with a length byte.
192 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
193 RFC 5321. Previously Dovecot (at least) would log errors during
196 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
197 in 4.61 without a reason logged. Binaries will be bigger, which might
198 matter on diskspace-constrained systems, but debug is easier.
200 JH/09 Fix macro-definition during "-be" expansion testing. The move to
201 write-protected store for macros had not accounted for these runtime
202 additions; fix by removing this protection for "-be" mode.
204 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
205 handing out large-numbered file descriptors, violating the usual Unix
206 assumption (and required by Posix) that the lowest possible number will be
207 allocated by the kernel when a new one is needed. In the daemon, and any
208 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
209 are not useable for FD_SET() [and hence select()] and overwrite the stack.
210 Assorted crashes happen.
212 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
213 main-section options or in a connect ACL, the value from the first ever
214 connection was never replaced for subsequent connections. Found by
217 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
218 though only once PCRE2 was introduced: the memory accounting used under
219 debug offset allocations by an int, giving a hard trap in early startup.
220 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
222 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
223 with underbars is given. The write-protection of configuration introduced
224 in 4.95 trapped when normalisation was applied to an option not needing
227 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
229 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
230 failing to close the certificates directory, every hour or any time it
233 JH/16 Debugging initiated by an ACL control now continues through into routing
234 and transport processes. Previously debugging stopped any time Exim
235 re-execs, or for processing a queued message.
237 JH/17 The "expand" debug selector now gives more detail, specifically on the
238 result of expansion operators and items.
240 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
241 bad comparison between the option value and the name of the file to
242 be included was done, and a mismatch was wrongly identified.
243 4.88 to 4.95 are affected.
245 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
247 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
248 supplied as the db file-pair basename. Previously, if a directory
249 path was given, for example via the autoreply "once" option, the DB
250 file.pag and file.dir files would be created in that directory's
253 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
254 "taint" log_selector. These were previously deprecated.
256 JH/22 Fix static address-list lookups to properly return the matched item.
257 Previously only the domain part was returned.
259 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
260 the call into OpenSSL to send a TLS Close was being repeated; this
261 resulted in the library waiting for the peer's Close. If that was never
262 sent we waited forever. Fix by tracking send calls.
264 JH/24 The ${run} expansion item now expands its command string elements after
265 splitting. Previously it was before; the new ordering makes handling
266 zero-length arguments simpler. The old ordering can be obtained by
267 appending a new option "preexpand", after a comma, to the "run".
269 JH/25 Taint-check exec arguments for transport-initiated external processes.
270 Previously, tainted values could be used. This affects "pipe", "lmtp" and
271 "queryprogram" transport, transport-filter, and ETRN commands.
272 The ${run} expansion is also affected: in "preexpand" mode no part of
273 the command line may be tainted, in default mode the executable name
276 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
277 the the facility was not passed across execs, and only the first message
278 passed over a connection could use BDAT; any further ones using DATA.
280 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
281 uses $sending_ip_address and an interface is specified.
282 Previously any use of the local address in the EHLO name disabled
283 PIPECONNECT, the common case being to use the rDNS of it.
285 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
286 resumption. Previously verify failed because no certificate status is
287 passed on the wire for the restarted session. Fix by using the recorded
288 ocsp status of the stored session for the new connection.
290 JH/29 TLS resumption: the key for session lookup in the client now includes
291 more info that a server could potentially use in configuring a TLS
292 session, avoiding oferring mismatching sessions to such a server.
293 Previously only the server IP was used.
295 JH/30 Fix string_copyn() for limit greater than actual string length.
296 Previously the copied amount was the limit, which could result in a
297 overlapping memcpy for newly allocated destination soon after a
298 source string shorter than the limit. Found/investigated by KM.
300 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
301 close; it may be needed for a subsequent connection. This caused a
302 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
304 JH/32 Fix CHUNKING for a second message on a connection when the first was
305 rejected. Previously we did not reset the chunking-offered state, and
306 erroneously rejected the BDAT command. Investigation help from
309 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
310 an empty address. Previously the expansion returned an error.
312 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
313 proxy. Previously these were misparsed, leading to paniclog entries.
319 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
320 related applications. Previously an "H" was used where available info
321 says that "M" should be, so change to match.
323 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
324 as arguments, so an implementation trying to copy these into a local
325 buffer was taking a taint-enforcement trap. Fix by using dynamically
326 created buffers. Similar fix for radius expansion condition.
328 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
329 reasonable, eg. to count headers. Fix by using dynamically created
330 buffers rather than a local. Do similar fixes for ACL actions "dcc",
331 "log_reject_target", "malware" and "spam"; the arguments are expanded
332 so could be handling tainted values.
334 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
335 broken the (no-op) support for this sendmail command. Restore it
336 to doing nothing, silently, and returning good status.
338 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
339 record path was given (or the default used) without a leading directory
340 path, an error occurred on trying to open it. Use the transport's working
343 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
344 transport. Previously it was the name on the DNS A-record; use instead
345 the head of the CNAME chain leading there (if there is one). This seems
346 to align better with RFC 6125.
348 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
349 smtp_accept_max_per_host allocated resources which were not released
350 when the limit was exceeded. This eventually crashed the daemon. Fix
351 by adding a release action in that path.
353 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
354 expanded; previously using tainted values was rejected. Fix by using
355 dynamically-created buffers.
357 JH/09 Relax restrictions on ACL verify condition needing access to message
358 headers. Previously they were only permitted in data and non-smtp ACLs;
359 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
360 not_blind, header_sender and header_names_ascii verification.
362 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
363 Previously a macro used one argument twice; when called with the
364 argument as an expression having side-effects, incorrect operation
365 resulted. Use an inlineable function.
367 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
368 held open for a verify callout. Previously this wan not accounted for
369 and a corrupt onward SMTP conversation resulted.
371 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
372 Previously they were embedded naively in the constructed address; when
373 needed, strip the quoting and quote the entire local_part.
374 Also make the inbound_srs expansion condition handle quoting.
376 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
377 excluded, not matching the documentation.
379 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
380 was given for the sqlite_dbfile a trap resulted.
382 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
383 "name" argument resulted in a trap. There is no reason to disallow such;
384 this was a coding error.
386 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
387 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
388 spent suspended, ignoring the POSIX definition. Previously we assumed
389 it did and a constant offset from real time could be used as a correction.
390 Change to using the same clock source for the start-of-message and the
391 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
392 exists, just to get a clock slightly more aligned to reality.
394 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
395 RFC says it is optional some validators care. The missing char was not
396 intended but triggered by a line-wrap alignment. Discovery and fix by
397 Guillaume Outters, hacked on by JH.
399 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
400 name being quoted was tainted a trap would be taken. Fix by using
401 dynamically created buffers. The routine could have been called by a
402 rewrite with the "h" flag, by using the "-F" command-line option, or
403 by using a "name=" option on a control=submission ACL modifier.
405 JH/19 SPF: change the Authentication-Results expansion component to give
406 smtp.helo when the sender domain is empty. Previously it gave
409 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
410 not in 127.0.0.0/8 to help in spotting list domains taken over by a
411 domain-parking registrar.
413 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
414 Previously when a whitespace character was specified it was not inserted
415 after removing the newline.
417 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
418 the domain part of the recipient address. This overrides any tls_sni
419 option set, which was previously used.
421 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
424 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
425 is_tainted() had an off-by-one error in the overenthusiastic direction.
426 Find and fix by Gavan. Although NetBSD is not a supported platform for
427 4.94 this bug could affect other platforms.
429 PP/01 Fix default prime selection to be consistent.
430 One path used ike23 still, instead of exim.dev.20160529.3; now both
431 execution flows will use the same DH primes (currently
432 exim.dev.20160529.3).
434 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
435 option in smtp transport, to match the documentation. Previously
436 verification was not being done in some cases where it should have been.
438 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
439 than one server was defined and depending on the platform memory layout
440 details, an internal consistency trap could be hit while walking the list
443 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
444 files. The value is used by the authresults expansion item. Previously
445 if this was used in a router or transport, a crash could result.
447 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
448 have a sufficient incidence of EINTR returns from select that an
449 interaction with other operations done by the main daemon loop exposed
450 a bug in the error-handling. This was benign apart from the log
453 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
454 for consistency with delivered "=>" and failed "**" lines. While we're
455 there, handle PRX and TFO.
457 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
458 applied. This resulted, if any header-line rewrite rules were configured,
459 in a panic-log triggerable by sending a message with a long address in
460 a header. Fix by increasing the arbitrary limit to larger than a single
461 (dewrapped) 5322 header line maximum size.
463 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
464 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
465 RFC 6648 which deprecates X- options in protocols as a general practice.
466 Changeover between the implementations is handled by the mechanisms
469 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
470 a remote which uses callout/hold. Previously the local was queued.
472 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
475 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
476 log line. The asterisk is now consistently appended to the remote IP
477 (and port, if given), and will also be provided on defer and fail log
478 lines. Previously it could be placed on the local IP if that was being
479 logged, and was only provided on delivery lines.
481 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
483 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
484 authenticator client_send option. Previously the next char, after a pair
485 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
486 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
487 documentation. There is still no way to get a leading ^ immediately
488 after a NUL (ie. for the password of a PLAIN method authenticator.
490 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
491 files. For bad sizes read, delete the record and whine to paniclog.
493 JH/38 When logging an AUTH failure, as server, do not include sensitive
494 information. Previously, the credentials would be included if given
495 as part of the AUTH command line and an ACL denied authentication.
497 JH/39 Bug 2691: fix $local_part_data. When the matching list element
498 referred to a file, bad data was returned. This likely also affected
501 JH/40 The gsasl authenticator now supports caching of the salted password
502 generated by the client-side implementation. This required the addition
503 of a new variable: $auth4.
505 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
506 left undeleted; the attempt to re-create it then failed - resulting in
507 the usual "SIGHUP tp have daemon reload configuration" to not work.
508 This affected any platform not supporting "abstract" Unix-domain
509 sockets (i.e. not Linux).
511 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
512 recipients" response to RCPT in a later response, with a 250. The
513 previous coding assumed this would not happen, and under PIPELINING
514 would result in both lost and duplicate recipients for a message.
516 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
517 Previously the weighting was incorrectly applied. Similar fix for socks
518 proxies. Found and fixed by Heiko Schlichting.
520 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
521 not handle sub-lists included using the +namedlist syntax. While
522 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
523 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
526 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
527 Previously the permanent pool was used, so the sore could not be freed.
528 This meant a connection with many messages would use continually-growing
531 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
532 per-pool so as not to waste too much space. Previously a constant size
533 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
534 DOS attacks harder. The cost is wasted memory use in the larger blocks.
536 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
537 same space repeatedly, and a smaller process footprint.
539 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
540 for GnuTLS. Previously the name started "%s" which, while not a bug,
541 looked as if if might be one.
543 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
544 "continued connection") make the $proxy_* variables available. Previously
545 the information was not passed across the exec() call for subsequent
546 transport executions. This also mean that the log lines for the
547 messages can show the proxy information.
549 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
550 receive time. With modern systems the difference is significant.
551 The historical behaviour can be restored by disabling (a new) log_selector
552 "queue_time_exclusive".
554 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
555 content) were specifically tested for. Now, also cover expansions
556 resulting in ACL names and inline ACL content.
558 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
559 making it unusable in complex expressions.
561 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
562 fail with a taint-check on the spoolfile name, and leave the message
565 HS/01 Enforce absolute PID file path name.
567 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
569 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
571 PP/02 Bug 2643: Correct TLS DH constants.
572 A missing NUL termination in our code-generation tool had led to some
573 incorrect Diffie-Hellman constants in the Exim source.
574 Reported by kylon94, code-gen tool fix by Simon Arlott.
576 PP/03 Impose security length checks on various command-line options.
577 Fixes CVE-2020-SPRSS reported by Qualys.
579 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
580 better. Reported by Qualys.
582 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
583 providing a particularly obnoxious sender full name.
586 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
588 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
591 PP/08 Change default for recipients_max from unlimited to 50,000.
593 PP/09 Fix security issue with too many recipients on a message (to remove a
594 known security problem if someone does set recipients_max to unlimited,
595 or if local additions add to the recipient list).
596 Fixes CVE-2020-RCPTL reported by Qualys.
598 PP/10 Fix security issue in SMTP verb option parsing
599 Fixes CVE-2020-EXOPT reported by Qualys.
601 PP/11 Fix security issue in BDAT state confusion.
602 Ensure we reset known-good where we know we need to not be reading BDAT
603 data, as a general case fix, and move the places where we switch to BDAT
604 mode until after various protocol state checks.
605 Fixes CVE-2020-BDATA reported by Qualys.
607 HS/03 Die on "/../" in msglog file names
609 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
610 the Exim runtime user are allowed to create files.
612 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
615 QS/03 When reading the output from interpreted forward files we do not
616 pass the pipe between the parent and the interpreting process to
617 executed child processes (if any).
619 QS/04 Always die if requested from internal logging, even is logging is
622 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
623 the API; compilation noo longer completes with DMARC support included.
624 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
625 been reported on other platforms.
627 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
629 JH/56 Make the majority of info read from config files readonly, for defence-in-
630 depth against exploits. Suggestion by Qualys.
631 Not supported on Solaris 10.
633 JH/57 Fix control=fakreject for a custom message containing tainted data.
634 Previously this resulted in a log complaint, due to a re-expansion present
635 since fakereject was originally introduced.
637 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
638 modifier was given, a loop resulted.
640 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
641 pipelined SMTP command followed the BDAT LAST then it would be
642 incorrectly treated as part of the message body, causing a verification
645 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
646 require looser validation rules than those for 821-level addresses,
647 which only permit IP addresses.
653 JH/01 Avoid costly startup code when not strictly needed. This reduces time
654 for some exim process initialisations. It does mean that the logging
655 of TLS configuration problems is only done for the daemon startup.
657 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
659 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
660 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
663 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
664 DKIM signing is being done. Previously a transport_filter would always
665 disable CHUNKING, falling back to traditional DATA.
667 JH/05 Regard command-line recipients as tainted.
669 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
671 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
672 PAM library frees one of the arguments given to it, despite the
673 documentation. Therefore a plain malloc must be used.
675 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
676 on-stack buffers were used, resulting in a taint trap when DSN information
677 copied from a received message was written into the buffer.
679 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
680 the ordering of its ARC headers. This caused a crash.
682 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
683 a new record was being constructed with information from the peer, a trap
686 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
687 installation would get error messages from DMARC verify, when it hit the
688 nonexistent file indicated by the default. Distros wanting DMARC enabled
689 should both provide the file and set the option.
690 Also enforce no DMARC verification for command-line sourced messages.
692 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
693 could, depending on the platform, hang at the STARTTLS response.
695 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
696 message on a connection. Previously if one message had ARC headers and
697 the following one did not, a crash could result when adding an
698 Authentication-Results: header.
700 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
701 Exim main code and Exim-related utities. The introduction of taint
702 tracking also did many adjustments to string handling. Since then, eximon
703 frequently terminated with an assert failure.
705 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
706 check for 452 responses. This slightly helps the inefficieny of doing
707 a large alias-expansion into a recipient-limited target. The max_rcpt
708 transport option still applies (and at the current default, will override
709 the new feature). The check is done for either cause of synch, and forces
710 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
711 connection. The new facility is not tunable at this time.
713 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
714 library live data was being used, so the results became garbage. Make
715 copies while it is still usable.
717 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
718 on delivery deferred (==) and failed (**) lines (if a delivery was
719 attemtped). Previously it was only on completion (=>) lines.
721 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
722 for the expansion of the server_scram_iter and server_scram_salt options.
724 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
725 are now specifically given a NO_DATA response without hitting the system
726 resolver. The library goes on to do the now-standard TXT lookup.
727 Use of dnsdb lookups is not affected.
729 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
730 only retrieve the errormessage once. Previously two calls to dlerror()
731 were used, and the second one (for mainlog/paniclog) retrieved null
734 JH/20 Taint checking: disallow use of tainted data for
735 - the appendfile transport file and directory options
736 - the pipe transport command
737 - the autoreply transport file, log and once options
738 - file names used by the redirect router (including filter files)
740 - paths used by single-key lookups
741 Previously this was permitted.
743 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
744 adjusted the size of a major service buffer; this failed because the
745 buffer was in use at the time. Change to a compile-time increase in the
746 buffer size, when this authenticator is compiled into exim.
748 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
749 previous fast-mode was untenable in the face of glibs using mmap to
750 support larger malloc requests.
752 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
753 New values supported, if defined on system where compiled:
754 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
755 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
757 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
758 running a limited number of proceses in parallel, a benefit is gained. The
759 amount varies with the platform hardware and load. The use of the option
760 queue_run_in_order means we cannot do this, as ordering becomes
763 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
764 had introduced a string-copy (for ensuring NUL-termination) which was not
765 appropriate for that case, which can include embedded NUL bytes in the
766 block of data. Investigation showed the copy to actually be needless, the
767 data being length-specified.
769 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
770 done during a receiving connection, and both used TLS, global info was
771 used rather than per-connection info for tracking the state of data
772 queued for transmission. This could result in a connection hang.
774 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
775 Previously, when delivering serveral messages down a single connection
776 only the first would provide a SIZE. This was due to the size information
777 not being properly tracked.
779 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
780 TAI (at 37 seconds currently), pretend to be in UTC for time-related
781 expansion and logging. Previously, spurious values such as a future
782 minute could be seen.
784 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
785 it could crash from a null-deref. This could also affect the
786 ${addresses: } operator and ${readsock } item.
788 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
789 message following a mime one, the variable was not reset.
791 JH/30 When an pipelined-connect fails at the first response, assume incorrect
792 cached capability (perhaps the peer reneged?) and immediately retry in
795 JH/31 Fix spurious detection of timeout while writing to transport filter.
797 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
798 an attempt to copy the string was made before checking it.
800 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
801 taint of the lookup key was maintained; we now regard the presence in the
802 filesystem as sufficient validation.
804 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
805 argument is supplied.
807 JH/35 The dsearch lookup now requires that the directory is an absolute path.
808 Previously this was not checked, and nonempty relative paths made an
809 access under Exim's current working directory.
811 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
812 Previously no event was raised.
814 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
815 parameter supplied by the sender MAIL FROM command. Previously it was
816 ignored, and only the check_spool_space option value for the required
819 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
820 the size of the signing public-key. Previously it was instead giving
821 the size of the signature hash.
823 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
824 the default. See the (new) dkim_verify_min_keysizes option.
826 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
827 and an ACL use a lookup for checking either the local_part or domain,
828 stale data could be accessed. Ensure that variable references are
829 dropped between messages.
831 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
832 by the client was not checked as pointing within response data before
833 being used. A malicious client could thus cause an out-of-bounds read and
834 possibly gain authentication. Fix by adding the check.
836 JH/42 Internationalisation: change the default for downconversion in the smtp
837 transport to be "if needed". Previously it was "as previously set" for
838 the message, which usually meant "if needed" for message-submission but
839 "no" for everything else. However, MTAs have been seen using SMTPUTF8
840 even when the envelope addresses did not need it, resulting in forwarding
841 failures to non-supporting MTAs. A downconvert in such cases will be
842 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
843 transport. The change does mean that addresses needing conversion will
844 be converted when previously a delivery failure would occur.
846 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
847 response was received it would be used unchecked in a fail-DSN, violating
848 standards on line-length limits. Truncate if needed.
850 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
851 doesn't work. (Additionally add a new main config option to configure the
858 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
859 side, to decode a TLS 1.3 packet capture.
861 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
862 Previously the default library behaviour applied, sending two, each in
865 JH/03 Debug output for ACL now gives the config file name and line number for
868 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
870 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
872 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
873 buffer overrun for (non-chunking) other transports.
875 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
876 TLS1.3, means that a server rejecting a client certificate is not visible
877 to the client until the first read of encrypted data (typically the
878 response to EHLO). Add detection for that case and treat it as a failed
879 TLS connection attempt, so that the normal retry-in-clear can work (if
880 suitably configured).
882 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
883 and/or domain. Found and fixed by Jason Betts.
885 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
886 configuration). If a CNAME target was not a wellformed name pattern, a
889 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
890 the OS reports them interleaved with other addresses.
892 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
893 used both for input and for a verify callout, both encrypted, SMTP
894 responses being sent by the server could be lost. This resulted in
895 dropped connections and sometimes bounces generated by a peer sending
898 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
899 string. Previously it was possible to cause undefined behaviour in a
900 library routine (usually a crash). Found by "zerons".
902 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
905 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
906 API was removed, so update to use the newer ones.
908 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
909 any timeout set, is taking a long time. Previously we would hang on to a
910 rotated logfile "forever" if the input was arriving with long gaps
911 (a previous attempt to fix addressed lack, for a long time, of initial
914 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
915 shared (NFS) environment. The length of the tempfile name is now
916 4 + 16 ("hdr.$message_exim_id") which might break on file
917 systems which restrict the file name length to lower values.
920 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
921 shared (NFS) environment.
923 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
924 did for all versions <4.90). Notably -M, -m, --invert, -I may be
927 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
928 on some platforms for bit 31.
930 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
931 to changes apparently associated with TLS1.3 handling some of the APIs
932 previously used were either nonfunctional or inappropriate. Strings
933 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
934 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
935 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
936 This affects log line X= elements, the $tls_{in,out}_cipher variables,
937 and the use of specific cipher names in the encrypted= ACL condition.
939 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
941 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
942 verification result was not updated unless hosts_require_ocsp applied.
944 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
945 queue_list_requires_admin set to false, non-admin users were denied the
948 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
949 directory-of-certs mode. Previously they were advertised despite the
952 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
953 A single TCP connection by a client will now hold a TLS connection open
954 for multiple message deliveries, by default. Previously the default was to
957 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
958 default. If built with the facility, DANE will be used. The facility
959 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
961 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
962 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
963 must be defined and you must still, unless you define DISABLE_TLS, manage
964 the the include-dir and library-file requirements that go with that
965 choice. Non-TLS builds are still supported.
967 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
970 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
971 default. If the platform supports and has the facility enabled, it will
972 be requested on all coneections.
974 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
975 controlled by the build-time option SUPPORT_PIPE_CONNECT.
977 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
979 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
980 success-DSN messages. Previously the From: header was always the default
981 one for these; the option was ignored.
983 JH/28 Fix the timeout on smtp response to apply to the whole response.
984 Previously it was reset for every read, so a teergrubing peer sending
985 single bytes within the time limit could extend the connection for a
986 long time. Credit to Qualsys Security Advisory Team for the discovery.
988 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
989 delivery address, which leaked information of the results of local
990 forwarding. Change to the original envelope recipient address, per
993 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
994 requested. Previously not bounce was generated and a log entry of
995 error ignored was made.
997 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
999 JH/32 Introduce a general tainting mechanism for values read from the input
1000 channel, and values derived from them. Refuse to expand any tainted
1001 values, to catch one form of exploit.
1003 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1004 was unused and the unexpanded text used for the test. Found and
1005 fixed by Ruben Jenster.
1007 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1008 an attempt to use a TLS library read routine dereffed a nul pointer,
1011 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1012 them in our smtp response.
1014 JH/36 Have the general router option retry_use_local_part default to true when
1015 any of the restrictive preconditions are set (to anything). Previously it
1016 was only for check_local user. The change removes one item of manual
1017 configuration which is required for proper retries when a remote router
1018 handles a subset of addresses for a domain.
1020 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1021 link count into consideration.
1023 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1024 caused the extension of big_buffer, the following lines were ignored.
1026 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1027 accordance with RFC 2308. Previously there was no expiry, so a longlived
1028 receive process (eg. due to ACL delays) versus a short SOA value could
1031 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1033 JH/39 Promote DMARC support to mainline.
1035 JH/40 Bug 2452: Add a References: header to DSNs.
1037 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1038 parameters. The relevant library call is documented as "Deprecated: This
1039 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1040 3.6.0, DH parameters are negotiated following RFC7919."
1042 HS/06 Change the default of dnssec_request_domains to "*"
1044 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1045 carried on and emitted a BDAT command, even when PIPELINING was not
1048 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1049 buffer was used for the filename, resulting in a trap when tainted
1050 arguments (eg. $domain) were used.
1052 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1053 recommended to avoid a possible server-load attack. The feature can be
1054 re-enabled via the openssl_options main cofiguration option.
1056 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1057 for version 4.90 - adding a "more data" boolean to the arguments.
1058 Bumped the ABI version number also, this having been missed previously;
1059 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1060 respect of usage of smtp_printf() by either local_scan code or libraries
1061 accessed via the ${dlfunc } expansion item. Both will need coding
1062 adjustment for any calls to smtp_printf() to match the new function
1063 signature; a FALSE value for the new argument is always safe.
1065 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1066 the file-offset (which the Linux syscall does, and exim expects); this
1067 resulted in an indefinite loop.
1069 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1070 to do ARC verification. The Authentication-Results: header line added
1071 by the configuration then had no ARC item.
1073 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1074 an incorrect mode was used for reading the result, resulting in it being
1081 JH/01 Remove code calling the customisable local_scan function, unless a new
1082 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1084 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1085 non-signal-safe functions being used.
1087 JH/03 Bug 2269: When presented with a received message having a stupidly large
1088 number of DKIM-Signature headers, disable DKIM verification to avoid
1089 a resource-consumption attack. The limit is set at twenty.
1091 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1092 report of oldest_pass in ${authres } in consequence, and separate out
1093 some descriptions of reasons for verification fail.
1095 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1096 files in the spool were present and unlocked. A queue-runner could spot
1097 them, resulting in a duplicate delivery. Fix that by doing the unlock
1098 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1099 add more error-checking on spoolfile handling while that code is being
1102 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1103 No known attacks, no CVE, this is defensive hardening.
1105 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1106 a queue-runner could start a delivery while other operations were ongoing.
1107 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1108 Found and investigated by Tim Stewart. Fix by using the open message data
1109 file handle rather than opening another, and not locally closing it (which
1110 releases a lock) for that case, while creating the temporary .eml format
1111 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1113 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1114 $sender_verify_failure/$recipient_verify_failure to "random".
1116 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1119 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1120 Previously this would segfault.
1122 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1125 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1126 like zero, since the resolver should be doing this for us, But we need one
1127 as a CNAME but no MX presence gets the CNAME returned; we need to check
1128 that doesn't point to an MX to declare it "no MX returned" rather than
1129 "error, loop". A new main option is added so the older capability of
1130 following some limited number of chain links is maintained.
1132 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1134 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1135 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1136 now-deprecated earlier definitions used only specified the range up to TLS
1137 1.2 (in the older-version library docs).
1139 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1141 JH/15 Rework TLS client-side context management. Stop using a global, and
1142 explicitly pass a context around. This enables future use of TLS for
1143 connections to service-daemons (eg. malware scanning) while a client smtp
1144 connection is using TLS; with cutthrough connections this is quite likely.
1146 JH/16 Fix ARC verification to do AS checks in reverse order.
1148 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1150 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1151 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1152 Previously the "utf8" would be re-prepended for every additional message.
1154 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1155 Previously thery were accepted, resulting in issues when attempting to
1156 forward messages to a non-supporting MTA.
1158 PP/02 Let -n work with printing macros too, not just options.
1160 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1161 one parent address was copied, and bogus data was used at delivery-logging
1162 time. Either a crash (after delivery) or bogus log data could result.
1163 Discovery and analysis by Tim Stewart.
1165 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1166 Previously if the string ended mid-character, we did not insert the
1167 promised '?' replacement.
1169 PP/04 Documentation: current string operators work on bytes, not codepoints.
1171 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1172 should pack well giving a smaller memory footprint so better caching and
1173 therefore performance. Group the declarations where this can't be done so
1174 that the byte-sized flag variables are not interspersed among pointer
1175 variables, giving a better chance of good packing by the compiler.
1177 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1178 non-null, to avoid issues with sites running BATV. Previously reports were
1179 sent with an empty envelope sender so looked like bounces.
1181 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1182 The ignore_error flag wasn't being returned from the filter subprocess so
1183 was not set for later routers. Investigation and fix by Matthias Kurz.
1185 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1186 and a msg:complete for the whole, when a message is manually removed using
1187 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1189 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1190 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1191 hope that is portable enough.
1193 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1194 requiring a known-CA anchor certificate; make it now rely entirely on the
1195 TLSA as an anchor. Checking the name on the leaf cert against the name
1196 on the A-record for the host is still done for TA (but not for EE mode).
1198 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1199 included in delivery lines for non-proxied connections, when compiled with
1200 SUPPORT_SOCKS and running with proxy logging enabled.
1202 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1203 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1204 move the existing event to fire before the normal logging of message
1205 failure so that custom logging is bracketed by normal logging.
1207 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1208 msg:fail:internal event. Developement by Matthias Kurz.
1210 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1211 far too small for todays use of crypto signatures stored there. Go all
1212 the way to the max DNS message size of 64kB, even though this might be
1213 overmuch for IOT constrained device use.
1215 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1216 copy a string over itself. The library routine is documented as not
1217 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1219 JH/32 For main options check_spool_space and check_inode_space, where the
1220 platform supports 64b integers, support more than the previous 2^31 kB
1221 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1222 the previous G, M, k.
1224 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1225 $authenticated_fail_id variable on authentication failure. Previously
1228 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1229 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1230 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1231 GNUTLS_SEC_PARAM_MEDIUM.
1233 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1234 side. Previously we would continue as if no SNI had been received.
1236 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1237 "<" character, which should be a list-separator specification, we walked
1238 off past the nul-terimation.
1240 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1241 causes) even when the retry time is not yet met. Previously they were
1242 not, meaning that when (say) an account was over-quota and temp-rejecting,
1243 and multiple senders' messages were queued, only one sender would get
1244 notified on each configured delay_warning cycle.
1246 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1248 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1249 cipher-suites, an error can be left on the stack even for a succeeding
1250 accept; this results in impossible error messages when a later operation
1253 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1254 return error codes indicating retry. Under TLS1.3 this becomes required.
1256 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1257 it only wrote the new authenticators, resulting in a lack of tracking of
1258 peer changes of ESMTP extensions until the next cache flush.
1260 JH/41 Fix the loop reading a message header line to check for integer overflow,
1261 and more-often against header_maxsize. Previously a crafted message could
1262 induce a crash of the recive process; now the message is cleanly rejected.
1264 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1265 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1271 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1272 When redis_servers is set to a list of > 1 element, and the Redis servers
1273 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1274 case of MOVED into a DEFER case instead, thus moving the query onto the
1275 next server in the list. For a cluster of N elements, all N servers must
1276 be defined in redis_servers.
1278 GF/02 Catch and remove uninitialized value warning in exiqsumm
1279 Check for existence of @ARGV before looking at $ARGV[0]
1281 JH/01 Replace the store_release() internal interface with store_newblock(),
1282 which internalises the check required to safely use the old one, plus
1283 the allocate and data copy operations duplicated in both (!) of the
1284 extant use locations.
1286 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1287 modifier. This matches the restriction on the commandline.
1289 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1290 Previously only the last row was returned.
1292 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1293 we assumed that tags in the header were well-formed, and parsed the
1294 element content after inspecting only the first char of the tag.
1295 Assumptions at that stage could crash the receive process on malformed
1298 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1299 While running the DKIM ACL we operate on the Permanent memory pool so that
1300 variables created with "set" persist to the DATA ACL. Also (at any time)
1301 DNS lookups that fail create cache records using the Permanent pool. But
1302 expansions release any allocations made on the current pool - so a dnsdb
1303 lookup expansion done in the DKIM ACL releases the memory used for the
1304 DNS negative-cache, and bad things result. Solution is to switch to the
1305 Main pool for expansions.
1306 While we're in that code, add checks on the DNS cache during store_reset,
1307 active in the testsuite.
1308 Problem spotted, and debugging aided, by Wolfgang Breyha.
1310 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1311 When none of the hosts presented to a transport match an already-open
1312 connection, close it and proceed with the list. Previously we would
1313 queue the message. Spotted by Lena with Yahoo, probably involving
1316 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1317 Previously a spurious "250 OK id=" response was appended to the proper
1320 JH/08 The "support for" informational output now, which built with Content
1321 Scanning support, has a line for the malware scanner interfaces compiled
1322 in. Interface can be individually included or not at build time.
1324 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1325 by the template makefile "src/EDITME". The "STREAM" support for an older
1326 ClamAV interface method is removed.
1328 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1329 rows affected is given instead).
1331 JH/11 The runtime Berkeley DB library version is now additionally output by
1332 "exim -d -bV". Previously only the compile-time version was shown.
1334 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1335 SMTP connection. Previously, when one had more recipients than the
1336 first, an abortive onward connection was made. Move to full support for
1337 multiple onward connections in sequence, handling cutthrough connection
1338 for all multi-message initiating connections.
1340 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1341 routers. Previously, a multi-recipient message would fail to match the
1342 onward-connection opened for the first recipient, and cause its closure.
1344 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1345 a timeout on read on a GnuTLS initiating connection, resulting in the
1346 initiating connection being dropped. This mattered most when the callout
1347 was marked defer_ok. Fix to keep the two timeout-detection methods
1350 JH/15 Relax results from ACL control request to enable cutthrough, in
1351 unsupported situations, from error to silently (except under debug)
1352 ignoring. This covers use with PRDR, frozen messages, queue-only and
1355 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1357 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1358 metadata, resulting in a crash in free().
1360 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1361 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1362 Broken also in d185889f4, with init system revamp.
1364 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1365 Previously we abruptly closed the connection after reading a malware-
1366 found indication; now we go on to read the "scan ok" response line,
1369 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1370 ACL. Previously, a crash would result.
1372 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1373 macro text after a replacement (previously it was only once per line) and
1374 by skipping builtin macros when searching for an uppercase lead character.
1376 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1377 for the build is renamed.
1379 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1380 was allocated for every new TLS startup, meaning one per message. Fix
1381 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1383 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1384 reported the original. Fix to report (as far as possible) the ACL
1385 result replacing the original.
1387 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1388 OpenSSL. Certificate information is loaded for every new TLS startup,
1389 and the resources needed to be freed.
1391 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1393 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1396 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1397 DATA response info to the (existing) per-recipient response info for
1398 the "C=" log element. It can have useful tracking info from the
1399 destination system. Patch from Simon Arlott.
1401 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1402 length value. Previously this would segfault.
1404 HS/02 Support Avast multiline protoocol, this allows passing flags to
1405 newer versions of the scanner.
1407 JH/28 Ensure that variables possibly set during message acceptance are marked
1408 dead before release of memory in the daemon loop. This stops complaints
1409 about them when the debug_store option is enabled. Discovered specifically
1410 for sender_rate_period, but applies to a whole set of variables.
1411 Do the same for the queue-runner and queue-list loops, for variables set
1412 from spool message files. Do the same for the SMTP per-message loop, for
1413 certain variables indirectly set in ACL operations.
1415 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1416 as a multi-recipient message from a mailinglist manager). The coding had
1417 an arbitrary cutoff number of characters while checking for more input;
1418 enforced by writing a NUL into the buffer. This corrupted long / fast
1419 input. The problem was exposed more widely when more pipelineing of SMTP
1420 responses was introduced, and one Exim system was feeding another.
1421 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1422 receiving system, and refused recipients seen by the sending system
1423 (propating to people being dropped from mailing lists).
1424 Discovered and pinpointed by David Carter.
1426 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1427 replaced by the ${authresults } expansion.
1429 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1431 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1432 allows proper process termination in container environments.
1434 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1435 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1437 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1438 and "err_perm", deprecated since 4.83 when the RFC-defined words
1439 "temperror" and "permerror" were introduced.
1441 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1442 transport-filters or DKIM-signing. The restriction was lost in the
1443 consolidation of verify-callout and delivery SMTP handling.
1444 Extend the restriction to also cover ARC-signing.
1446 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1447 in defer=pass mode supply a 450 to the initiator. Previously the message
1450 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1451 tls_require_ciphers is used as before.
1453 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1454 "pass_unscanned". Only tmpfails from the scanner are written to
1455 the paniclog, as they may require admin intervention (permission
1456 denied, license issues). Other scanner errors (like decompression
1457 bombs) do not cause a paniclog entry.
1459 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1460 Previously it was possible to log spurious information in receive log
1463 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1464 triggered odd behaviour from Outlook Express clients.
1466 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1469 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1470 since the IETF WG has not yet settled on that versus the original
1471 "bare" representation.
1473 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1474 Previously the millisecond value corrupted the output.
1475 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1476 corrupted the output.
1482 JH/01 Rework error string handling in TLS interface so that the caller in
1483 more cases is responsible for logging. This permits library-sourced
1484 string to be attached to addresses during delivery, and collapses
1485 pairs of long lines into single ones.
1487 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1488 during configuration. Wildcards are allowed and expanded.
1490 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1491 This permits better logging.
1493 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1494 do not close it down and have the child transport start it up again on
1495 the passed-on TCP connection. Instead, proxy the child (and any
1496 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1497 affected: the continued delivery log lines do not have any DNSSEC, TLS
1498 Certificate or OCSP information. TLS cipher information is still logged.
1500 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1501 identical IP addresses on different listening ports. Will also affect
1504 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1505 add noisy ifdef guards to special-case this sillyness.
1506 Patch from Bernd Kuhls.
1508 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1509 than 255 are no longer allowed.
1511 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1512 Disable the session-cache too, which might reduce our load. Since we
1513 currrectly use a new context for every connection, both as server and
1514 client, there is no benefit for these.
1515 GnuTLS appears to not support tickets server-side by default (we don't
1516 call gnutls_session_ticket_enable_server()) but client side is enabled
1517 by default on recent versions (3.1.3 +) unless the PFS priority string
1520 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1521 <https://reproducible-builds.org/specs/source-date-epoch/>.
1523 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1524 the check for any unsuccessful recipients did not notice the limit, and
1525 erroneously found still-pending ones.
1527 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1528 MSG_MORE. Only in-clear (not on TLS connections).
1530 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1531 transport-filter is involved we can buffer the headers in memory for
1532 creating the signature, and read the spool data file once for the
1533 signature and again for transmission.
1535 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1536 4.77 as the kernel support then wasn't solid, having issues in 64bit
1537 mode. Now, it's been long enough. Add support for FreeBSD also.
1539 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1540 case where the routing stage had gathered several addresses to send to
1541 a host before calling the transport for the first, we previously failed
1542 to close down TLS in the old transport process before passing the TCP
1543 connection to the new process. The new one sent a STARTTLS command
1544 which naturally failed, giving a failed delivery and bloating the retry
1545 database. Investigation and fix prototype from Wolfgang Breyha.
1547 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1548 false-negatives in the check that the sender had not preempted a response
1549 or prompt from Exim (running as a server), due to that code's lack of
1550 awareness of the SMTP input buffering.
1552 PP/04 Add commandline_checks_require_admin option.
1553 Exim drops privileges sanely, various checks such as -be aren't a
1554 security problem, as long as you trust local users with access to their
1555 own account. When invoked by services which pass untrusted data to
1556 Exim, this might be an issue. Set this option in main configuration
1557 AND make fixes to the calling application, such as using `--` to stop
1560 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1561 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1562 responses to those, into a single TLS record each way (this usually means
1563 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1566 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1567 affects you only if you're dancing at the edge of the param size limits.
1568 If you are, and this message makes sense to you, then: raise the
1569 configured limit or use OpenSSL 1.1. Nothing we can do for older
1572 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1573 cmdline element to get the documented default one. Previously it was
1576 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1577 are unsuable for non-message-specific reasons - eg. connection timeout,
1578 banner-time rejection.
1580 JH/16 Fix logging of delivery remote port, when specified by router, under
1583 PP/06 Repair manualroute's ability to take options in any order, even if one
1584 is the name of a transport.
1587 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1589 JH/17 Change the list-building routines interface to use the expanding-string
1590 triplet model, for better allocation and copying behaviour.
1592 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1593 Previously it was constructed the first time a possibly-matching string
1594 was met in the configuration file input during startup; now it is done
1597 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1598 compatible one, to avoid the (poorly documented) possibility of a config
1599 file in the working directory redirecting the DB files, possibly correpting
1600 some existing file. CVE-2017-10140 assigned for BDB.
1602 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1603 cache-hot. Previously, although the result was properly cached, the
1604 initial verify call returned a defer.
1606 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1607 the main verify for receipient in uncached-mode.
1609 JH/22 Retire historical build files to an "unsupported" subdir. These are
1610 defined as "ones for which we have no current evidence of testing".
1612 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1613 if present. Previously it was ignored.
1615 JH/24 Start using specified-initialisers in C structure init coding. This is
1616 a C99 feature (it's 2017, so now considered safe).
1618 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1619 if was a fixed-sized field and bitmask ops via macros; it is now more
1622 PP/07 GitHub PR 56: Apply MariaDB build fix.
1623 Patch provided by Jaroslav Škarvada.
1625 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1626 during Coverity cleanups [4.87 JH/47]
1627 Diagnosis and fix provided by Michael Fischer v. Mollard.
1629 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1630 the right size to place the terminating semicolon on its own folded
1631 line, the header hash was calculated to an incorrect value thanks to
1632 the (relaxed) space the fold became.
1634 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1635 and confused the parent.
1637 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1638 which could crash as a result. This could lead to undeliverable messages.
1640 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1643 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1644 matters on fast-turnover and PID-randomising systems, which were getting
1645 out-of-order delivery.
1647 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1648 a possibly-overlapping copy. The symptom was that "Remote host closed
1649 connection in response to HELO" was logged instead of the actual 4xx
1652 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1653 Previously only that bufferd was discarded, resulting in SYMTP command
1656 JH/32 DKIM: when a message has multiple signatures matching an identity given
1657 in dkim_verify_signers, run the dkim acl once for each. Previously only
1658 one run was done. Bug 2189.
1660 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1661 "panic the current process" to "deliberately defer". The panic log is
1662 still written with the problem list name; the mail and reject logs now
1663 get a temp-reject line for the message that was being handled, saying
1664 something like "domains check lookup or other defer". The SMTP 451
1665 message is still "Temporary local problem".
1667 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1668 A crafted sequence of BDAT commands could result in in-use memory beeing
1669 freed. CVE-2017-16943.
1671 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1672 from SMTP input. Previously it was always done; now only done for DATA
1673 and not BDAT commands. CVE-2017-16944.
1675 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1676 to the message (such as an overlong header line). Previously this was
1677 not done and we did not exit BDAT mode. Followon from the previous item
1678 though a different problem.
1684 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1685 than -2003 did; needs libidn2 in addition to libidn.
1687 JH/02 The path option on a pipe transport is now expanded before use.
1689 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1690 Patch provided by "Björn", documentation fix added too.
1692 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1693 missing a wire-to-host endian conversion.
1695 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1696 close after a BDAT command line could be taken as a following command,
1697 giving a synch failure. Fix by only checking for synch immediately
1698 before acknowledging the chunk.
1700 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1701 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1702 macro. Patches provided by Josh Soref.
1704 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1705 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1706 by IANA as a service extension. However, John Klensin suggests that we
1709 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1710 the dkim code may be unix-mode line endings rather than smtp wire-format
1711 CRLF, so prepend a CR to any bare LF.
1713 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1714 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1716 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1717 the first were themselves being wrongly included in the feed into dkim
1718 processing; with most chunk sizes in use this resulted in an incorrect
1719 body hash calculated value.
1721 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1722 DKIM signature block, for verification. Although advised against by
1723 standards it is specifically not ruled illegal.
1725 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1727 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1728 missing a body hash (the bh= tag).
1730 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1731 It seems that HAProxy sends the Proxy Protocol information in clear and
1732 only then does a TLS startup, so do the same.
1734 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1735 TCP connections (such as for Spamd) unless the daemon successfully set
1736 Fast Open mode on its listening sockets. This fixes breakage seen on
1737 too-old kernels or those not configured for Fast Open, at the cost of
1738 requiring both directions being enabled for TFO, and TFO never being used
1739 by non-daemon-related Exim processes.
1741 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1742 endings, at least on the first header line. Try to canonify any that get
1743 past that check, despite the cost.
1745 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1746 now limited to an arbitrary five deep, while parsing addresses with the
1747 strip_excess_angle_brackets option enabled.
1749 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1750 instead leave the unprompted TLS handshake in socket buffer for the
1751 TLS library to consume.
1753 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1755 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1757 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1758 region in one operation, for speed, and this leaves assigned pointers
1759 dangling. Add checks run only under the testsuite which checks all
1760 variables at a store-reset and panics on a dangling pointer; add code
1761 explicitly nulling out all the variables discovered. Fixes one known
1762 bug: a transport crash, where a dangling pointer for $sending_ip_address
1763 originally assigned in a verify callout, is re-used.
1765 PP/06 Drop '.' from @INC in various Perl scripts.
1767 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1769 PP/08 Reduce a number of compilation warnings under clang; building with
1770 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1771 should be warning-free.
1773 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1775 HS/01 Fix portability problems introduced by PP/08 for platforms where
1776 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1778 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1779 chunk. This allows us to accept broken chunked messages. We need a more
1780 general solution here.
1782 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1783 already-broken messages in the queue.
1785 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1787 JH/19 Fix reference counting bug in routing-generated-address tracking.
1793 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1794 supports it and a size is available (ie. the sending peer gave us one).
1796 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1797 years of being deprecated). The replacements are the ACLs
1798 acl_smtp_mime and acl_not_smtp_mime.
1800 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1801 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1802 or even an in-clear connection were permitted. Now, if the host lookup
1803 was dnssec and dane was requested then the host is only used if the
1804 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1805 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1806 if one fails this test.
1807 This means that a poorly-configured remote DNS will make it incommunicado;
1808 but it protects against a DNS-interception attack on it.
1810 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1811 non-noisy when a race steals the message being considered.
1813 JH/05 If main configuration option tls_certificate is unset, generate a
1814 self-signed certificate for inbound TLS connections.
1816 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1817 in rewrites and routers.
1819 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1820 and logged a warning sing 4.83; now they are a configuration file error.
1822 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1823 (lacking @domain). Apply the same qualification processing as RCPT.
1825 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1827 JH/10 Support ${sha256:} applied to a string (as well as the previous
1830 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1831 a cutthrough deliver is pending, as we always want to make a connection.
1832 This also avoids re-routing the message when later placing the cutthrough
1833 connection after a verify cache hit.
1834 Do not update it with the verify result either.
1836 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1837 when routing results in more than one destination address.
1839 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1840 signing (which inhibits the cutthrough capability). Previously only
1841 the presence of an option was tested; now an expansion evaluating as
1842 empty is permissible (obviously it should depend only on data available
1843 when the cutthrough connection is made).
1845 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1846 the relevant preceding SMTP command did not note the pipelining mode.
1848 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1849 Previously they were not counted.
1851 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1852 as one having no matching records. Previously we deferred the message
1853 that needed the lookup.
1855 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1856 distinguished as "(=".
1858 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1859 for missing MX records. Previously it only worked for missing A records.
1861 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1863 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1864 after the data-go-ahead and data-ack. Patch from Jason Betts.
1866 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1867 even for a "none" policy. Patch from Tony Meyer.
1869 JH/22 Fix continued use of a connection for further deliveries. If a port was
1870 specified by a router, it must also match for the delivery to be
1873 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1874 When one of the recipients of a message was unsuitable for the connection
1875 (has no matching addresses), we lost track of needing to mark it
1876 deferred. As a result mail would be lost.
1878 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1880 JH/25 Decoding ACL controls is now done using a binary search; the source code
1881 takes up less space and should be simpler to maintain. Merge the ACL
1882 condition decode tables also, with similar effect.
1884 JH/26 Fix problem with one_time used on a redirect router which returned the
1885 parent address unchanged. A retry would see the parent address marked as
1886 delivered, so not attempt the (identical) child. As a result mail would
1889 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1890 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1891 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1894 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1895 Main config options check_{log,spool}_{inodes,space} are now
1896 100 inodes, 10MB unless set otherwise in the configuration.
1898 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1899 Previously it only applied to the main-section connection policy
1902 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1904 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1905 by me. Added RFC7919 DH primes as an alternative.
1907 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1908 are not in the system include path.
1910 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1911 GnuTLS, when a session startup failed (eg because the client disconnected)
1912 Exim did stdio operations after fclose. This was exposed by a recent
1913 change which nulled out the file handle after the fclose.
1915 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1916 signed directly by the cert-signing cert, rather than an intermediate
1917 OCSP-signing cert. This is the model used by LetsEncrypt.
1919 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1921 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1922 an incoming connection.
1924 HS/02 Bug 1802: Do not half-close the connection after sending a request
1927 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1928 fallback to "prime256v1".
1930 JH/34 SECURITY: Use proper copy of DATA command in error message.
1931 Could leak key material. Remotely exploitable. CVE-2016-9963.
1937 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1938 and 3.4.4 - once the server is enabled to respond to an OCSP request
1939 it does even when not requested, resulting in a stapling non-aware
1940 client dropping the TLS connection.
1942 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1943 support variable-length bit vectors. No functional change.
1945 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1946 The I= interface field on outgoing lines is now after the H= remote
1947 host field, same as incoming lines. There is a separate
1948 outgoing_interface log selector which allows you to disable the
1951 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1952 If not running log_selector +smtp_connection the mainlog would be held
1953 open indefinitely after a "too many connections" event, including to a
1954 deleted file after a log rotate. Leave the per net connection logging
1955 leaving it open for efficiency as that will be quickly detected by the
1956 check on the next write.
1958 HS/01 Bug 1671: Fix post transport crash.
1959 Processing the wait-<transport> messages could crash the delivery
1960 process if the message IDs didn't exist for some reason. When
1961 using 'split_spool_directory=yes' the construction of the spool
1962 file name failed already, exposing the same netto behaviour.
1964 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1965 mime_regex ACL conditions.
1967 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1968 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1969 to HELO, local diagnostic string.
1971 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1972 log line to debug. Even when configured with a tls authenticator many
1973 client connections are expected to not authenticate in this way, so
1974 an authenticate fail is not an error.
1976 HS/02 Add the Exim version string to the process info. This way exiwhat
1977 gives some more detail about the running daemon.
1979 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1980 matter for fast-change records such as DNSBLs.
1982 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
1983 retry record, even if constant. There may be multiple transports with
1984 different interface settings and the retry behaviour needs to be kept
1987 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
1989 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
1991 JH/10 Bug 840: fix log_defer_output option of pipe transport
1993 JH/11 Bug 830: use same host for all RCPTS of a message, even under
1994 hosts_randomize. This matters a lot when combined with mua_wrapper.
1996 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
1997 ${quote_pgsql:<string>} operator.
1999 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2001 JH/14 Change header file name for freeradius-client. Relevant if compiling
2002 with Radius support; from the Gentoo tree and checked under Fedora.
2004 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2006 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2007 option transport_filter, meaning no filtering.
2009 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2011 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2012 defaults to "*" (all hosts). The variable is now available when not built
2013 with TLS, default unset, mainly to enable keeping the testsuite sane.
2014 If a server certificate is not supplied (via tls_certificate) an error is
2015 logged, and clients will find TLS connections fail on startup. Presumably
2016 they will retry in-clear.
2017 Packagers of Exim are strongly encouraged to create a server certificate
2018 at installation time.
2020 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2021 with the $config_file variable.
2023 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2024 in transport context, after the attempt, and per-recipient. The latter type
2025 is per host attempted. The event data is the error message, and the errno
2026 information encodes the lookup type (A vs. MX) used for the (first) host,
2027 and the trailing two digits of the smtp 4xx response.
2029 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2030 to write to mainlog (or rejectlog, paniclog) in the window between file
2031 creation and permissions/ownership being changed. Particularly affects
2032 installations where exicyclog is run as root, rather than exim user;
2033 result is that the running daemon panics and dies.
2035 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2037 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2038 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2039 "pri" and "weight". Note that the previous implicit priority given by the
2040 list order is no longer honoured.
2042 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2043 for DKIM processing.
2045 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2046 by defining SUPPORT_SOCKS.
2048 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2049 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2050 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2051 variables are renamed to proxy_{local,external}_{address,port}.
2053 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2054 for a build by defining SUPPORT_I18N
2056 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2057 of the query string, and make ${quote_redis:} do that quoting.
2059 JH/29 Move Events support from Experimental to mainline, enabled by default
2060 and removable for a build by defining DISABLE_EVENT.
2062 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2064 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2065 cached by the daemon.
2067 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2068 by defining LOOKUP_REDIS. The libhiredis library is required.
2070 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2071 keys are given for lookup.
2073 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2074 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2075 only supported when built with TLS support. The PolarSSL SHA routines
2076 are still used when the TLS library is too old for convenient support.
2078 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2079 openssl_options), for security. OpenSSL forces this from version 1.1.0
2080 server-side so match that on older versions.
2082 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2083 allocation for $value could be released as the expansion processing
2084 concluded, but leaving the global pointer active for it.
2086 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2087 and to use the domains and local_parts ACL conditions.
2089 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2090 incorrectly not doubled on cutthrough transmission, hence seen as a
2091 body-termination at the receiving system - resulting in truncated mails.
2092 Commonly the sender saw a TCP-level error, and retransmitted the message
2093 via the normal store-and-forward channel. This could result in duplicates
2094 received - but deduplicating mailstores were liable to retain only the
2095 initial truncated version.
2097 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2099 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2101 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2102 we're in there, support oversigning also; bug 1309.
2104 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2106 HS/04 Add support for keep_environment and add_environment options.
2108 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2109 either intentional arithmetic overflow during PRNG, or testing config-
2112 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2113 delivery resulted in actual delivery. Cancel cutthrough before DATA
2116 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2117 rejects a recipient: pass the reject to the originator.
2119 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2120 Many were false-positives and ignorable, but it's worth fixing the
2123 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2124 for the new environment-manipulation done at startup. Move the routines
2125 from being local to tls.c to being global via the os.c file.
2127 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2128 an extract embedded as result-arg for a map, the first arg for extract
2129 is unavailable so we cannot tell if this is a numbered or keyed
2130 extraction. Accept either.
2136 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2139 JH/02 The smtp transport option "multi_domain" is now expanded.
2141 JH/03 The smtp transport now requests PRDR by default, if the server offers
2144 JH/04 Certificate name checking on server certificates, when exim is a client,
2145 is now done by default. The transport option tls_verify_cert_hostnames
2146 can be used to disable this per-host. The build option
2147 EXPERIMENTAL_CERTNAMES is withdrawn.
2149 JH/05 The value of the tls_verify_certificates smtp transport and main options
2150 default to the word "system" to access the system default CA bundle.
2151 For GnuTLS, only version 3.0.20 or later.
2153 JH/06 Verification of the server certificate for a TLS connection is now tried
2154 (but not required) by default. The verification status is now logged by
2155 default, for both outbound TLS and client-certificate supplying inbound
2158 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2161 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2162 Status Notification (bounce) messages are now MIME format per RFC 3464.
2163 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2164 under the control of the dsn_advertise_hosts option, and routers may
2165 have a dsn_lasthop option.
2167 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2168 default, modifiable by a malware= option. The list separator for
2169 the options can now be changed in the usual way. Bug 68.
2171 JH/10 The smtp_receive_timeout main option is now expanded before use.
2173 JH/11 The incoming_interface log option now also enables logging of the
2174 local interface on delivery outgoing connections.
2176 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2177 if the interface and destination host and port all match.
2179 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2182 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2183 Patch from Andrew Lewis.
2185 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2186 now supports optional time-restrictions, weighting, and priority
2187 modifiers per server. Patch originally by <rommer@active.by>.
2189 JH/16 The spamd_address main option now supports a mixed list of local
2190 and remote servers. Remote servers can be IPv6 addresses, and
2191 specify a port-range.
2193 JH/17 Bug 68: The spamd_address main option now supports an optional
2194 timeout value per server.
2196 JH/18 Bug 1581: Router and transport options headers_add/remove can
2197 now have the list separator specified.
2199 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2202 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2205 JH/21 Support for the A6 type of dns record is withdrawn.
2207 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2208 rather than the verbs used.
2210 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2211 from 255 to 1024 chars.
2213 JH/24 Verification callouts now attempt to use TLS by default.
2215 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2216 are generic router options now. The defaults didn't change.
2218 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2219 Original patch from Alexander Shikoff, worked over by JH.
2221 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2222 files if ZCAT_COMMAND is not executable.
2224 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2226 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2228 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2229 Normally benign, it bites when the pair was led to by a CNAME;
2230 modern usage is to not canonicalize the domain to a CNAME target
2231 (and we were inconsistent anyway for A-only vs AAAA+A).
2233 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2235 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2236 when evaluating $sender_host_dnssec.
2238 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2239 $sender_helo_dnssec variable.
2241 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2243 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2245 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2247 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2248 documented as working, but never had. Support all but $spam_report.
2250 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2251 added for tls authenticator.
2253 HS/03 Add perl_taintmode main config option
2259 TL/01 When running the test suite, the README says that variables such as
2260 no_msglog_check are global and can be placed anywhere in a specific
2261 test's script, however it was observed that placement needed to be near
2262 the beginning for it to behave that way. Changed the runtest perl
2263 script to read through the entire script once to detect and set these
2264 variables, reset to the beginning of the script, and then run through
2265 the script parsing/test process like normal.
2267 TL/02 The BSD's have an arc4random API. One of the functions to induce
2268 adding randomness was arc4random_stir(), but it has been removed in
2269 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2270 function when detected.
2272 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2273 cause callback expansion.
2275 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2276 syntax errors in an expansion can be treated as a string instead of
2277 logging or causing an error, due to the internal use of bool_lax
2278 instead of bool when processing it.
2280 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2281 server certificates when making smtp deliveries.
2283 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2285 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2287 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2289 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2290 Merged patch from Sebastian Wiedenroth.
2292 JH/05 Fix results-pipe from transport process. Several recipients, combined
2293 with certificate use, exposed issues where response data items split
2294 over buffer boundaries were not parsed properly. This eventually
2295 resulted in duplicates being sent. This issue only became common enough
2296 to notice due to the introduction of connection certificate information,
2297 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2299 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2300 size buffer was used, resulting in syntax errors when an expansion
2303 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2304 version 3.3.6 or later.
2306 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2307 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2308 both become "event_action", the variables become $event_name, $event_data
2309 and $event_defer_errno. There is a new variable $verify_mode, usable in
2310 routers, transports and related events. The tls:cert event is now also
2311 raised for inbound connections, if the main configuration event_action
2314 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2315 early OCSP support, but no stapling (appears to be less than 1.0.0).
2317 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2318 server certificate names available under the smtp transport option
2319 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2322 JH/10 Time-related extraction expansions from certificates now use the main
2323 option "timezone" setting for output formatting, and are consistent
2324 between OpenSSL and GnuTLS compilations. Bug 1541.
2326 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2327 encoded parameter in the incoming message. Bug 1558.
2329 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2330 include certificate info, eximon was claiming there were spoolfile
2333 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2335 JH/14 Log delivery-related information more consistently, using the sequence
2336 "H=<name> [<ip>]" wherever possible.
2338 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2339 are problematic for Debian distribution, omit them from the release
2342 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2344 JH/16 Fix string representation of time values on 64bit time_t architectures.
2347 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2348 output list separator was used.
2353 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2354 checkers that were complaining about end of non-void function with no
2357 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2358 This was a regression introduced in 4.83 by another bugfix.
2360 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2362 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2363 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2369 TF/01 Correctly close the server side of TLS when forking for delivery.
2371 When a message was received over SMTP with TLS, Exim failed to clear up
2372 the incoming connection properly after forking off the child process to
2373 deliver the message. In some situations the subsequent outgoing
2374 delivery connection happened to have the same fd number as the incoming
2375 connection previously had. Exim would try to use TLS and fail, logging
2376 a "Bad file descriptor" error.
2378 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2379 utilities have not been installed.
2381 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2382 temporary space as the ACL may create new global variables.
2384 TL/01 LDAP support uses per connection or global context settings, depending
2385 upon the detected version of the libraries at build time.
2387 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2388 to extract and use the src ip:port in logging and expansions as if it
2389 were a direct connection from the outside internet. PPv2 support was
2390 updated based on HAProxy spec change in May 2014.
2392 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2394 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2395 Properly escape header and check for NULL return.
2397 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2400 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2402 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2403 characters in header names, implemented as a verify condition.
2404 Contributed by Michael Fischer v. Mollard.
2406 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2407 results permerror and temperror. Previous values are deprecated but
2408 still accepted. In a future release, err_perm and err_temp will be
2409 completely removed, which will be a backward incompatibility if the
2410 ACL tests for either of these two old results. Patch contributed by
2411 user bes-internal on the mailing list.
2413 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2415 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2416 selectors, in both main and reject logs.
2418 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2421 JH/07 Add malware type "sock" for talking to simple daemon.
2423 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2425 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2426 routers/transports under cutthrough routing.
2428 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2429 numbers. Touch up "bool" conditional to keep the same definition.
2431 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2433 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2435 JH/12 Expand items in router/transport headers_add or headers_remove lists
2436 individually rather than the list as a whole. Bug 1452.
2438 Required for reasonable handling of multiple headers_ options when
2439 they may be empty; requires that headers_remove items with embedded
2440 colons must have them doubled (or the list-separator changed).
2442 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2443 view the policy declared in the DMARC record. Currently, $dmarc_status
2444 is a combined value of both the record presence and the result of the
2447 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2449 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2450 dnslookup router and the smtp transport (applying to the forward
2453 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2454 of ldap servers used for a specific lookup. Patch provided by Heiko
2457 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2458 New variable $lookup_dnssec_authenticated for observability.
2460 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2461 Patch submitted by Lars Timman.
2463 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2465 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2466 Requires trusted mode and valid format message id, aborts otherwise.
2467 Patch contributed by Heiko Schlichting.
2469 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2470 certextract with support for various fields. Bug 1358.
2472 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2473 is requested by default, modifiable by smtp transport option
2476 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2477 operate on certificate variables to give certificate fingerprints
2478 Also new ${sha256:cert_variable}.
2480 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2482 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2483 Christian Aistleitner.
2485 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2487 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2488 file. Patch from Wolfgang Breyha.
2490 JH/25 Expand the coverage of the delivery $host and $host_address to
2491 client authenticators run in verify callout. Bug 1476.
2493 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2494 align with daemon_smtp_ports. Bug 72.
2496 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2497 support and error reporting did not work properly.
2499 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2500 and is readable. Patch from Andrew Colin Kissa.
2502 TL/14 Enhance documentation of ${run expansion and how it parses the
2503 commandline after expansion, particularly in the case when an
2504 unquoted variable expansion results in an empty value.
2506 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2508 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2509 by renaming away from it. Addresses GH issue 15, reported by
2512 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2514 TL/15 SECURITY: prevent double expansion in math comparison functions
2515 (can expand unsanitized data). Not remotely exploitable.
2522 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2524 PP/02 Make -n do something, by making it not do something.
2525 When combined with -bP, the name of an option is not output.
2527 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2530 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2531 $sender_host_name and config options to manage this, and basic check
2534 PP/05 DSCP support for outbound connections and control modifier for inbound.
2536 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2537 (Only plugin which currently uses this is kerberos4, which nobody should
2538 be using, but we should make it available and other future plugins might
2539 conceivably use it, even though it would break NAT; stuff *should* be
2540 using channel bindings instead).
2542 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2543 name; added for Sendmail compatibility; requires admin caller.
2544 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2545 just ignore it); requires trusted caller.
2546 Also parse but ignore: -Ac -Am -X<logfile>
2549 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2551 TL/02 Add +smtp_confirmation as a default logging option.
2553 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2554 Patch by Magnus Holmgren from 2007-02-20.
2556 TL/04 Bugzilla 1281 - Spec typo.
2557 Bugzilla 1283 - Spec typo.
2558 Bugzilla 1290 - Spec grammar fixes.
2560 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2562 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2564 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2565 mailing list by by Dmitry Isaikin.
2567 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2569 JH/02 Support "G" suffix to numbers in ${if comparisons.
2571 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2573 NM/01 Bugzilla 1197 - Spec typo
2574 Bugzilla 1196 - Spec examples corrections
2576 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2578 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2579 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2582 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2583 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2585 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2586 "acl {{name}{arg}...}", and optional args on acl condition
2589 JH/05 Permit multiple router/transport headers_add/remove lines.
2591 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2593 JH/07 Avoid using a waiting database for a single-message-only transport.
2594 Performance patch from Paul Fisher. Bugzilla 1262.
2596 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2599 JH/09 Add $headers_added variable, with content from use of ACL modifier
2600 add_header (but not yet added to the message). Bugzilla 199.
2602 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2603 Pulled from Bugzilla 817 by Wolfgang Breyha.
2605 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2607 (nb: this is the same fix as in Exim 4.80.1)
2609 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2612 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2615 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2617 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2618 advertise SMTP AUTH mechanism to us, instead of a generic
2619 protocol violation error. Also, make Exim more robust to bad
2620 data from the Dovecot auth socket.
2622 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2624 When a queue runner is handling a message, Exim first routes the
2625 recipient addresses, during which it prunes them based on the retry
2626 hints database. After that it attempts to deliver the message to
2627 any remaining recipients. It then updates the hints database using
2630 So if a recipient address works intermittently, it can get repeatedly
2631 deferred at routing time. The retry hints record remains fresh so the
2632 address never reaches the final cutoff time.
2634 This is a fairly common occurrence when a user is bumping up against
2635 their storage quota. Exim had some logic in its local delivery code
2636 to deal with this. However it did not apply to per-recipient defers
2637 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2639 This change adds a proper retry rule check during routing so that the
2640 final cutoff time is checked against the message's age. We only do
2641 this check if there is an address retry record and there is not a
2642 domain retry record; this implies that previous attempts to handle
2643 the address had the retry_use_local_parts option turned on. We use
2644 this as an approximation for the destination being like a local
2645 delivery, as in LMTP.
2647 I suspect this new check makes the old local delivery cutoff check
2648 redundant, but I have not verified this so I left the code in place.
2650 TF/02 Correct gecos expansion when From: is a prefix of the username.
2652 Test 0254 submits a message to Exim with the header
2656 When I ran the test suite under the user fanf2, Exim expanded
2657 the header to contain my full name, whereas it should have added
2658 a Resent-Sender: header. It erroneously treats any prefix of the
2659 username as equal to the username.
2661 This change corrects that bug.
2663 GF/01 DCC debug and logging tidyup
2664 Error conditions log to paniclog rather than rejectlog.
2665 Debug lines prefixed by "DCC: " to remove any ambiguity.
2667 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2669 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2670 Bug spotted by Jeremy Harris; was flawed since initial commit.
2671 Would have resulted in OCSP responses post-SNI triggering an Exim
2672 NULL dereference and crash.
2674 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2676 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2677 Bug detection, analysis and fix by Samuel Thibault.
2678 Bugzilla 1331, Debian bug #698092.
2680 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2682 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2683 Server implementation by Todd Lyons, client by JH.
2684 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2685 config variable "prdr_enable" controls whether the server
2686 advertises the facility. If the client requests PRDR a new
2687 acl_data_smtp_prdr ACL is called once for each recipient, after
2688 the body content is received and before the acl_smtp_data ACL.
2689 The client is controlled by both of: a hosts_try_prdr option
2690 on the smtp transport, and the server advertisement.
2691 Default client logging of deliveries and rejections involving
2692 PRDR are flagged with the string "PRDR".
2694 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2695 fclose(). Diagnosis by Todd Lyons.
2697 PP/17 Update configure.default to handle IPv6 localhost better.
2698 Patch by Alain Williams (plus minor tweaks).
2701 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2702 This is now consistent with GnuTLS, and is now documented: the
2703 previous undocumented portable approach to treating the option as
2704 unset was to force an expansion failure. That still works, and
2705 an empty string is now equivalent.
2707 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2708 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2709 not performing validation itself.
2711 PP/20 Added force_command boolean option to pipe transport.
2712 Patch from Nick Koston, of cPanel Inc.
2714 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2717 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2719 PP/21 Fix eximon continuous updating with timestamped log-files.
2720 Broken in a format-string cleanup in 4.80, missed when I repaired the
2721 other false fix of the same issue.
2722 Report and fix from Heiko Schlichting.
2725 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2726 Report from Prashanth Katuri.
2728 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2729 It's SecureTransport, so affects any MacOS clients which use the
2730 system-integrated TLS libraries, including email clients.
2732 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2733 using a MIME ACL for non-SMTP local injection.
2734 Report and assistance in diagnosis by Warren Baker.
2736 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2738 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2740 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2741 last id that failed so it may be referenced in subsequent ACL's.
2743 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2746 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2747 ldap library initialization, allowing self-signed CA's to be
2748 used. Also properly sets require_cert option later in code by
2749 using NULL (global ldap config) instead of ldap handle (per
2750 session). Bug diagnosis and testing by alxgomz.
2752 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2753 the src/util/ subdirectory.
2755 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2756 renamed to Transport Post Delivery Action by Jeremy Harris, as
2759 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2760 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2761 redis_servers = needs to be configured which will be used by the redis
2762 lookup. Patch from Warren Baker, of The Packet Hub.
2764 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2766 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2767 hostname or reverse DNS when processing a host list. Used suggestions
2768 from multiple comments on this bug.
2770 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2772 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2773 Missed a few lines, added it to make the runtest require no keyboard
2776 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2777 contains upper case chars. Make router use caseful_local_part.
2779 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2780 support when GnuTLS has been built with p11-kit.
2786 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2788 This, or similar/improved, will also be change PP/11 of 4.82.
2794 PP/01 Handle short writes when writing local log-files.
2795 In practice, only affects FreeBSD (8 onwards).
2796 Bugzilla 1053, with thanks to Dmitry Isaikin.
2798 NM/01 Bugzilla 949 - Documentation tweak
2800 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2803 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2805 PP/02 Implemented gsasl authenticator.
2807 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2809 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2810 `pkg-config foo` for cflags/libs.
2812 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2813 with rest of GSASL and with heimdal_gssapi.
2815 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2816 `pkg-config foo` for cflags/libs for the TLS implementation.
2818 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2819 properties get this fed in as external SSF. A number of robustness
2820 and debugging improvements to the cyrus_sasl authenticator.
2822 PP/08 cyrus_sasl server now expands the server_realm option.
2824 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2825 Patch by Jeremy Harris.
2827 PP/10 Added dbmjz lookup type.
2829 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2831 PP/12 MAIL args handles TAB as well as SP, for better interop with
2832 non-compliant senders.
2833 Analysis and variant patch by Todd Lyons.
2835 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2836 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2837 Patch from Dirk Mueller <dmueller@suse.com>
2839 PP/13 tls_peerdn now print-escaped for spool files.
2840 Observed some $tls_peerdn in wild which contained \n, which resulted
2841 in spool file corruption.
2843 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2844 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2845 or write after TLS renegotiation, which otherwise led to messages
2848 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2849 as a tracking header (ie: a signed header comes before the signature).
2850 Patch from Wolfgang Breyha.
2852 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2853 comma-sep list; embedded commas doubled.
2855 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2857 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2859 Report and patch from Dmitry Banschikov.
2861 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2862 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2863 failures. We appear to now support TLS1.1+ with Exim.
2865 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2866 lets Exim select keys and certificates based upon TLS SNI from client.
2867 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2868 before an outbound SMTP session. New log_selector, +tls_sni.
2870 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2871 NULL dereference. Report and patch from Alun Jones.
2873 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2874 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2875 Not seeing resolver debug output on NetBSD, but suspect this is a
2876 resolver implementation change.
2878 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2879 Left warnings. Added "eximon gdb" invocation mode.
2881 PP/21 Defaulting "accept_8bitmime" to true, not false.
2883 PP/22 Added -bw for inetd wait mode support.
2885 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2886 locate the relevant includes and libraries. Made this the default.
2888 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2889 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2891 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2892 This may cause build issues on older platforms.
2894 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2895 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2896 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2897 Added SNI support via GnuTLS too.
2898 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2900 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2902 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2903 Applied second patch from Janne, implementing suggestion to default
2904 multiple-strings-in-record handling to match SPF spec.
2906 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2908 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2909 read-only, out of scope).
2910 Patch from Wolfgang Breyha, report from Stuart Northfield.
2912 PP/29 Fix three issues highlighted by clang analyser static analysis.
2913 Only crash-plausible issue would require the Cambridge-specific
2914 iplookup router and a misconfiguration.
2915 Report from Marcin Mirosław.
2917 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2919 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2920 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2921 As part of this, removing so much warning spew let me fix some minor
2922 real issues in debug logging.
2924 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2925 assignment on my part. Fixed.
2927 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2928 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2929 Janne Snabb (who went above and beyond: thank you).
2931 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2932 string otherwise requires a connection and a bunch more work and it's
2933 relatively easy to get wrong. Should also expose TLS library linkage
2936 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2937 64-bit ${eval} (JH/03).
2939 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2940 GNU libc to support some of the 64-bit stuff, should not lead to
2941 conflicts. Defined before os.h is pulled in, so if a given platform
2942 needs to override this, it can.
2944 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2945 protection layer was required, which is not implemented.
2946 Bugzilla 1254, patch from Wolfgang Breyha.
2948 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2949 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2950 tls_dhparam take prime identifiers. Also unbreak combination of
2951 OpenSSL+DH_params+TLSSNI.
2953 PP/39 Disable SSLv2 by default in OpenSSL support.
2959 PP/01 Solaris build fix for Oracle's LDAP libraries.
2960 Bugzilla 1109, patch from Stephen Usher.
2962 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2964 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2967 TF/02 Fix a couple more cases where we did not log the error message
2968 when unlink() failed. See also change 4.74-TF/03.
2970 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2971 lock up or crash if it happened to be inside a call to libc when it
2972 got a SIGUSR1 from exiwhat.
2974 The SIGUSR1 handler appends the current process status to the process
2975 log which is later printed by exiwhat. It used to use the general
2976 purpose logging code to do this, but several functions it calls are
2977 not safe for signals.
2979 The new output code in the SIGUSR1 handler is specific to the process
2980 log, and simple enough that it's easy to inspect for signal safety.
2981 Removing some special cases also simplifies the general logging code.
2982 Removing the spurious timestamps from the process log simplifies
2985 TF/04 Improved ratelimit ACL condition.
2987 The /noupdate option has been deprecated in favour of /readonly which
2988 has clearer semantics. The /leaky, /strict, and /readonly update modes
2989 are mutually exclusive. The update mode is no longer included in the
2990 database key; it just determines when the database is updated. (This
2991 means that when you upgrade Exim will forget old rate measurements.)
2993 Exim now checks that the per_* options are used with an update mode that
2994 makes sense for the current ACL. For example, when Exim is processing a
2995 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
2996 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
2997 must specify per_mail/readonly. If you omit the update mode it defaults to
2998 /leaky where that makes sense (as before) or /readonly where required.
3000 The /noupdate option is now undocumented but still supported for
3001 backwards compatibility. It is equivalent to /readonly except that in
3002 ACLs where /readonly is required you may specify /leaky/noupdate or
3003 /strict/noupdate which are treated the same as /readonly.
3005 A useful new feature is the /count= option. This is a generalization
3006 of the per_byte option, so that you can measure the throughput of other
3007 aggregate values. For example, the per_byte option is now equivalent
3008 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3010 The per_rcpt option has been generalized using the /count= mechanism
3011 (though it's more complicated than the per_byte equivalence). When it is
3012 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3013 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3014 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3015 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3016 Note that using per_rcpt with a non-readonly update mode in more than
3017 one ACL will cause the recipients to be double-counted. (The per_mail
3018 and per_byte options don't have this problem.)
3020 The handling of very low rates has changed slightly. If the computed rate
3021 is less than the event's count (usually one) then this event is the first
3022 after a long gap. In this case the rate is set to the same as this event's
3023 count, so that the first message of a spam run is counted properly.
3025 The major new feature is a mechanism for counting the rate of unique
3026 events. The new per_addr option counts the number of different
3027 recipients that someone has sent messages to in the last time period. It
3028 behaves like per_rcpt if all the recipient addresses are different, but
3029 duplicate recipient addresses do not increase the measured rate. Like
3030 the /count= option this is a general mechanism, so the per_addr option
3031 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3032 example, measure the rate that a client uses different sender addresses
3033 with the options per_mail/unique=$sender_address. There are further
3034 details in the main documentation.
3036 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3038 TF/06 Removed a few PCRE remnants.
3040 TF/07 Automatically extract Exim's version number from tags in the git
3041 repository when doing development or release builds.
3043 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3044 Bugzilla 879. Patch from Paul Fisher.
3046 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3047 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3050 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3052 PP/05 Variable $av_failed, true if the AV scanner deferred.
3053 Bugzilla 1078. Patch from John Horne.
3055 PP/06 Stop make process more reliably on build failure.
3056 Bugzilla 1087. Patch from Heiko Schlittermann.
3058 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3059 Bugzilla 1089. Patch from Heiko Schlittermann.
3061 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3062 Bugzilla 1131. Patch from Holger Weiß.
3064 PP/09 Handle IPv6 addresses with SPF.
3065 Bugzilla 860. Patch from Wolfgang Breyha.
3067 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3069 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3072 PP/11 match_* no longer expand right-hand-side by default.
3073 New compile-time build option, EXPAND_LISTMATCH_RHS.
3074 New expansion conditions, "inlist", "inlisti".
3076 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3078 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3080 PP/14 fix log_write() format string regression from TF/03.
3081 Bugzilla 1152. Patch from Dmitry Isaikin.
3087 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3089 PP/02 Harmonised TLS library version reporting; only show if debugging.
3090 Layout now matches that introduced for other libraries in 4.74 PP/03.
3092 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3094 PP/04 New "dns_use_edns0" global option.
3096 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3099 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3100 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3102 TK/01 Updated PolarSSL code to 0.14.2.
3103 Bugzilla 1097. Patch from Andreas Metzler.
3105 PP/07 Catch divide-by-zero in ${eval:...}.
3106 Fixes bugzilla 1102.
3108 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3111 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3112 format-string attack -- SECURITY: remote arbitrary code execution.
3114 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3115 time unintentionally subject to list matching rules, letting the header
3116 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3117 arbitrary string expansion). This allowed for information disclosure.
3119 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3120 INT_MIN/-1 -- value coerced to INT_MAX.
3126 NM/01 Workaround for PCRE version dependency in version reporting
3129 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3130 This fixes portability to compilers other than gcc, notably
3131 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3133 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3134 makefiles for portability to HP-UX and POSIX correctness.
3136 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3137 Also via indented variable definition in the Makefile.
3138 (Debugging by Oliver Heesakkers).
3140 PP/02 Restore caching of spamd results with expanded spamd_address.
3141 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3143 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3144 Improves build reliability. Fix from: Frank Elsner
3146 NM/02 Fix wide character breakage in the rfc2047 coding
3147 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3149 NM/03 Allow underscore in dnslist lookups
3150 Fixes bug 1026. Patch from Graeme Fowler
3152 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3153 Code patches from Adam Ciarcinski of NetBSD.
3155 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3158 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3159 is logged, to avoid truncation. Patch from John Horne.
3161 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3162 Patch from Jakob Hirsch.
3164 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3165 SQL string expansion failure details.
3166 Patch from Andrey Oktyabrski.
3168 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3169 Patch from Simon Arlott.
3171 PP/09 New lookups functionality failed to compile on old gcc which rejects
3172 extern declarations in function scope.
3173 Patch from Oliver Fleischmann
3175 PP/10 Use sig_atomic_t for flags set from signal handlers.
3176 Check getgroups() return and improve debugging.
3177 Fixed developed for diagnosis in bug 927 (which turned out to be
3180 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3181 Patch from Mark Zealey.
3183 PP/12 Bugzilla 1056: Improved spamd server selection.
3184 Patch from Mark Zealey.
3186 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3187 Based on patch from Heiko Schlittermann.
3189 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3190 Patch from Uwe Doering, sign-off by Michael Haardt.
3192 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3193 variable declaration deep within a block. Bug and patch from
3196 PP/15 lookups-Makefile IRIX compatibility coercion.
3198 PP/16 Make DISABLE_DKIM build knob functional.
3200 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3201 Patch by Simon Arlott
3203 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3204 variable argument macros. Our copy now differs from upstream.
3210 TF/01 Failure to get a lock on a hints database can have serious
3211 consequences so log it to the panic log.
3213 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3214 controlled using the smtp_confirmation log selector.
3216 TF/03 Include the error message when we fail to unlink a spool file.
3218 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3219 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3220 for maintaining out-of-tree patches for some time.
3222 PP/01 Bugzilla 139: Documentation and portability issues.
3223 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3224 Handle per-OS dynamic-module compilation flags.
3226 PP/02 Let /dev/null have normal permissions.
3227 The 4.73 fixes were a little too stringent and complained about the
3228 permissions on /dev/null. Exempt it from some checks.
3229 Reported by Andreas M. Kirchwitz.
3231 PP/03 Report version information for many libraries, including
3232 Exim version information for dynamically loaded libraries. Created
3233 version.h, now support a version extension string for distributors
3234 who patch heavily. Dynamic module ABI change.
3236 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3237 privilege escalation vulnerability whereby the Exim run-time user
3238 can cause root to append content of the attacker's choosing to
3241 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3244 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3245 If dropping privileges for untrusted macros, we disabled normal logging
3246 on the basis that it would fail; for the Exim run-time user, this is not
3247 the case, and it resulted in successful deliveries going unlogged.
3248 Fixed. Reported by Andreas Metzler.
3254 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3255 only prepend for the Resent-* case. Fixes regression introduced in
3256 Exim 4.70 by NM/22 for Bugzilla 607.
3258 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3259 increasing numbers of administrators be bitten by this.
3261 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3263 PP/03 Bugzilla 994: added openssl_options main configuration option.
3265 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3267 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3269 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3271 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3272 without a peer certificate, leading to a segfault because of an
3273 assumption that peers always have certificates. Be a little more
3274 paranoid. Problem reported by Martin Tscholak.
3276 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3277 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3278 NB: ClamAV planning to remove STREAM in "middle of 2010".
3279 CL also introduces -bmalware, various -d+acl logging additions and
3280 more caution in buffer sizes.
3282 PP/09 Implemented reverse_ip expansion operator.
3284 PP/10 Bugzilla 937: provide a "debug" ACL control.
3286 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3288 PP/12 Bugzilla 973: Implement --version.
3290 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3292 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3294 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3296 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3297 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3298 ignore trailing whitespace.
3300 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3302 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3305 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3306 Notification from Dr Andrew Aitchison.
3308 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3309 ExtendedDetectionInfo response format.
3310 Notification from John Horne.
3312 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3315 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3316 XSL and documented dependency on system catalogs, with examples of how
3319 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3322 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3323 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3324 configuration file which is writeable by the Exim user or group.
3326 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3327 of configuration files to cover files specified with the -C option if
3328 they are going to be used with root privileges, not just the default
3331 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3332 option (effectively making it always true).
3334 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3335 files to be used while preserving root privileges.
3337 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3338 that rogue child processes cannot use them.
3340 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3341 run-time user, instead of root.
3343 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3344 Exim run-time user without dropping privileges.
3346 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3347 result string, instead of calling string_vformat() twice with the same
3350 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3351 for other users. Others should always drop root privileges if they use
3352 -C on the command line, even for a whitelisted configure file.
3354 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3356 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3362 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3363 $data_path, and $header_path variables; fixed documentation bugs and
3366 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3367 exipick to access non-standard spools, including the "frozen" queue
3370 NM/01 Bugzilla 965: Support mysql stored procedures.
3371 Patch from Alain Williams
3373 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3375 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3376 Patch from Andreas Metzler
3378 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3379 Patch from Kirill Miazine
3381 NM/05 Bugzilla 671: Added umask to procmail example.
3383 JJ/03 installed exipick 20100323.0, fixing doc bug
3385 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3386 directory. Notification and patch from Dan Rosenberg.
3388 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3390 TK/02 Improve log output when DKIM signing operation fails.
3392 MH/01 Treat the transport option dkim_domain as a colon separated
3393 list, not as a single string, and sign the message with each element,
3394 omitting multiple occurences of the same signer.
3396 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3397 Bugzilla 985, 986. Patch by Simon Arlott
3399 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3400 Patch by Simon Arlott
3402 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3403 MBX locking. Notification from Dan Rosenberg.
3409 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3411 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3413 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3415 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3417 NM/04 Bugzilla 915: Flag broken perl installation during build.
3423 TK/01 Added patch by Johannes Berg that expands the main option
3424 "spamd_address" if it starts with a dollar sign.
3426 TK/02 Write list of recipients to X-Envelope-Sender header when building
3427 the mbox-format spool file for content scanning (suggested by Jakob
3430 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3431 (http://www.dcc-servers.net/) support via dccifd. Activated by
3432 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3434 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3435 by Mark Daniel Reidel <mr@df.eu>.
3437 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3438 When building exim an external PCRE library is now needed -
3439 PCRE is a system library on the majority of modern systems.
3440 See entry on PCRE_LIBS in EDITME file.
3442 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3443 conversation. Added nologin parameter to request.
3444 Patch contributed by Kirill Miazine.
3446 TF/01 Do not log submission mode rewrites if they do not change the address.
3448 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3450 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3451 log files in place. Contributed by Roberto Lima.
3453 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3455 TF/03 Bugzilla 615: When checking the local_parts router precondition
3456 after a local_part_suffix or local_part_prefix option, Exim now
3457 does not use the address's named list lookup cache, since this
3458 contains cached lookups for the whole local part.
3460 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3461 Robert Millan. Documentation is in experimental-spec.txt.
3463 TF/04 Bugzilla 668: Fix parallel build (make -j).
3465 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3467 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3468 Patch provided by Jan Srzednicki.
3470 TF/05 Leading white space used to be stripped from $spam_report which
3471 wrecked the formatting. Now it is preserved.
3473 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3474 that they are available at delivery time.
3476 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3478 TF/08 TLS error reporting now respects the incoming_interface and
3479 incoming_port log selectors.
3481 TF/09 Produce a more useful error message if an SMTP transport's hosts
3482 setting expands to an empty string.
3484 NM/06 Bugzilla 744: EXPN did not work under TLS.
3485 Patch provided by Phil Pennock.
3487 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3488 Patch provided by Richard Godbee.
3490 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3491 acl_smtp_notquit, added index entry.
3493 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3494 Patch provided by Eugene Bujak.
3496 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3497 accept(). Patch provided by Maxim Dounin.
3499 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3500 Patch provided by Phil Pennock.
3502 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3504 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3505 Patch provided by Brad "anomie" Jorsch.
3507 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3508 Patch provided by Dean Brooks.
3510 TK/05 Add native DKIM support (does not depend on external libraries).
3512 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3513 Patch provided by Graeme Fowler.
3515 NM/16 Bugzilla 851: Documentation example syntax fix.
3517 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3519 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3522 NM/19 Bugzilla 745: TLS version reporting.
3523 Patch provided by Phil Pennock.
3525 NM/20 Bugzilla 167: bool: condition support.
3526 Patch provided by Phil Pennock.
3528 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3529 clients. Patch provided by Phil Pennock.
3531 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3532 Patch provided by Brad "anomie" Jorsch.
3534 NM/23 Bugzilla 687: Fix misparses in eximstats.
3535 Patch provided by Heiko Schlittermann.
3537 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3538 Patch provided by Heiko Schlittermann.
3540 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3541 plus update to original patch.
3543 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3545 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3546 Patch provided by David Brownlee.
3548 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3550 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3552 NM/30 Bugzilla 888: TLS documentation bugfixes.
3554 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3556 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3557 Unlike the original bugzilla I have changed all shell scripts in src tree.
3559 NM/33 Bugzilla 898: Transport filter timeout fix.
3560 Patch by Todd Rinaldo.
3562 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3563 Patch by Serge Demonchaux.
3565 NM/35 Bugzilla 39: Base64 decode bug fixes.
3566 Patch by Jakob Hirsch.
3568 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3570 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3572 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3574 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3580 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3581 ALT-N's libdkim that I have put here:
3582 http://duncanthrax.net/exim-experimental/
3584 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3585 were called 'true' and it seems that C99 defines that as a reserved
3586 keyword to be used with 'bool' variable types. That means you could
3587 not include C99-style headers which use bools without triggering
3588 build errors in sieve.c.
3590 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3591 as mailq or other aliases. Changed the --help handling significantly
3592 to do whats expected. exim_usage() emits usage/help information.
3594 SC/01 Added the -bylocaldomain option to eximstats.
3596 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3598 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3600 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3606 PH/01 Another patch from the Sieve maintainer.
3608 PH/02 When an IPv6 address is converted to a string for single-key lookup
3609 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3610 dots are used instead of colons so that keys in lsearch files need not
3611 contain colons. This was done some time before quoting was made available
3612 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3613 (notated using the quote facility) so as to distinguish them from IPv4
3614 keys. This meant that lookups for IP addresses in host lists did not work
3615 for iplsearch lookups.
3617 This has been fixed by arranging for IPv6 addresses to be expressed with
3618 colons if the lookup type is iplsearch. This is not incompatible, because
3619 previously such lookups could never work.
3621 The situation is now rather anomalous, since one *can* have colons in
3622 ordinary lsearch keys. However, making the change in all cases is
3623 incompatible and would probably break a number of configurations.
3625 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3628 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3629 conversion specification without a maximum field width, thereby enabling
3630 a rogue spamd server to cause a buffer overflow. While nobody in their
3631 right mind would setup Exim to query an untrusted spamd server, an
3632 attacker that gains access to a server running spamd could potentially
3633 exploit this vulnerability to run arbitrary code as the Exim user.
3635 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3636 $primary_hostname instead of what libspf2 thinks the hosts name is.
3638 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3639 a directory entry by the name of the lookup key. Previously, if a
3640 symlink pointed to a non-existing file or a file in a directory that
3641 Exim lacked permissions to read, a lookup for a key matching that
3642 symlink would fail. Now it is enough that a matching directory entry
3643 exists, symlink or not. (Bugzilla 503.)
3645 PH/03 The body_linecount and body_zerocount variables are now exported in the
3648 PH/04 Added the $dnslist_matched variable.
3650 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3651 This means they are set thereafter only if the connection becomes
3654 PH/06 Added the client_condition to authenticators so that some can be skipped
3655 by clients under certain conditions.
3657 PH/07 The error message for a badly-placed control=no_multiline_responses left
3658 "_responses" off the end of the name.
3660 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3662 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3663 (without spaces) instead of just copying the configuration text.
3665 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3667 PH/11 Added $max_received_linelength.
3669 PH/12 Added +ignore_defer and +include_defer to host lists.
3671 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3672 way in which PCRE > 7.0 is built.
3674 PH/14 Implemented queue_only_load_latch.
3676 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3677 MAIL command. The effect was to mangle the value on 64-bit systems.
3679 PH/16 Another patch from the Sieve maintainer.
3681 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3683 PH/18 If a system quota error occurred while trying to create the file for
3684 a maildir delivery, the message "Mailbox is full" was not appended to the
3685 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3686 only to a quota excession during the actual writing of the file.
3688 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3689 characters?) which causes problems in log lines. The DN values are now
3690 passed through string_printing() before being added to log lines.
3692 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3693 and InterBase are left for another time.)
3695 PH/21 Added message_body_newlines option.
3697 PH/22 Guard against possible overflow in moan_check_errorcopy().
3699 PH/23 POSIX allows open() to be a macro; guard against that.
3701 PH/24 If the recipient of an error message contained an @ in the local part
3702 (suitably quoted, of course), incorrect values were put in $domain and
3703 $local_part during the evaluation of errors_copy.
3709 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3710 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3713 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3714 issue a MAIL command.
3716 PH/02 In an ACL statement such as
3718 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3720 if a client was not listed at all, or was listed with a value other than
3721 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3722 the condition was not true (as it should be), so access was not denied.
3723 The bug was that the ! inversion was incorrectly passed on to the second
3724 item. This has been fixed.
3726 PH/03 Added additional dnslists conditions == and =& which are different from
3727 = and & when the dns lookup returns more than one IP address.
3729 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3730 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3732 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3733 FSYNC, which compiles an option called disable_fsync that allows for
3734 bypassing fsync(). The documentation is heavily laced with warnings.
3736 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3738 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3739 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3740 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3741 including adding "make clean"; (3) Added -fPIC when compiling the test
3742 dynamically loaded module, to get rid of a warning.
3744 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3745 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3746 = 0s. The bug is otherwise harmless.
3748 PH/07 There was a bug in the dovecot authenticator such that the value of
3749 $auth1 could be overwritten, and so not correctly preserved, after a
3750 successful authentication. This usually meant that the value preserved by
3751 the server_setid option was incorrect.
3753 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3755 PH/09 Installed PCRE release 7.0.
3757 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3758 run for batched SMTP input. It is now run at the start of every message
3759 in the batch. While fixing this I discovered that the process information
3760 (output by running exiwhat) was not always getting set for -bs and -bS
3761 input. This is fixed, and it now also says "batched" for BSMTP.
3763 PH/11 Added control=no_pipelining.
3765 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3766 patch, slightly modified), and move the expansion of helo_data till after
3767 the connection is made in the smtp transport (so it can use these
3770 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3772 PH/14 Added log_selector = +pid.
3774 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3776 PH/16 Add ${if forany and ${if forall.
3778 PH/17 Added dsn_from option to vary the From: line in DSNs.
3780 PH/18 Flush SMTP output before performing a callout, unless control =
3781 no_callout_flush is set.
3783 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3784 was true (the default) a successful delivery failed to delete the retry
3785 item, thus causing premature timeout of the address. The bug is now
3788 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3790 PH/21 Long custom messages for fakedefer and fakereject are now split up
3791 into multiline responses in the same way that messages for "deny" and
3792 other ACL rejections are.
3794 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3795 with slight modification.
3797 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3798 draft, changing the syntax and factoring some duplicate code".
3800 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3801 for deliveries of the second and subsequent messages over the same SMTP
3804 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3805 ${reduce, with only minor "tidies".
3807 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3809 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3810 expansion side effects.
3812 PH/27 When a message times out after an over-quota error from an Exim-imposed
3813 quota, the bounce message says "mailbox is full". This message was not
3814 being given when it was a system quota that was exceeded. It now should
3817 MH/03 Made $recipients available in local_scan(). local_scan() already has
3818 better access to the recipient list through recipients_list[], but
3819 $recipients can be useful in postmaster-provided expansion strings.
3821 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3822 in the case of a MAIL command with additional options following the
3823 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3824 were accidentally chopped off.
3826 PH/29 SMTP synchronization checks are implemented when a command is read -
3827 there is a check that no more input is waiting when there shouldn't be
3828 any. However, for some commands, a delay in an ACL can mean that it is
3829 some time before the response is written. In this time, more input might
3830 arrive, invalidly. So now there are extra checks after an ACL has run for
3831 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3832 pipelining has not been advertised.
3834 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3836 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3837 returned in the SMTP response when smtp_return_error_details was false.
3838 This has been fixed.
3840 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3841 instead of the C I/O that was originally supplied, because problems were
3842 reported on Solaris.
3844 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3845 Exim which did not show up earlier: it was assuming that a call to
3846 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3847 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3848 was a macro that became an assignment, so it seemed to work. This has
3849 changed to a proper function call with a void return, hence the compile
3850 error. Exim's code has been fixed.
3852 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3855 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3857 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3859 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3860 because it turns out to be too big or there is a timeout) there is no
3861 "Completed" line in the log. When some messages of this type were
3862 selected by exigrep, they were listed as "not completed". Others were
3863 picked up by some special patterns. I have improved the selection
3864 criteria to be more general.
3866 PH/38 The host_find_failed option in the manualroute router can now be set
3867 to "ignore", to completely ignore a host whose IP address cannot be
3868 found. If all hosts are ignored, the behaviour is controlled by the new
3869 host_all_ignored option.
3871 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3872 homing or because of multiple MX records with /mx) generated more than
3873 one IP address, and the following item turned out to be the local host,
3874 all the secondary addresses of the first item were incorrectly removed
3875 from the list, along with the local host and any following hosts (which
3876 is what is supposed to happen).
3878 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3879 whoever called Exim into the -H file. In the case of the daemon it was
3880 behaving confusingly. When first started, it used values for whoever
3881 started the daemon, but after a SIGHUP it used the Exim user (because it
3882 calls itself on a restart). I have changed the code so that it now always
3885 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3886 message are rejected with the same error (e.g. no authentication or bad
3887 sender address), and a DATA command is nevertheless sent (as can happen
3888 with PIPELINING or a stupid MUA), the error message that was given to the
3889 RCPT commands is included in the rejection of the DATA command. This is
3890 intended to be helpful for MUAs that show only the final error to their
3893 PH/42 Another patch from the Sieve maintainer.
3895 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3896 Eximstats - Fixed some broken HTML links and added missing column headers
3898 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3899 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3901 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3903 PH/43 Yet another patch from the Sieve maintainer.
3905 PH/44 I found a way to check for a TCP/IP connection going away before sending
3906 the response to the final '.' that terminates a message, but only in the
3907 case where the client has not sent further data following the '.'
3908 (unfortunately, this is allowed). However, in many cases there won't be
3909 any further data because there won't be any more messages to send. A call
3910 to select() can be used: if it shows that the input is "ready", there is
3911 either input waiting, or the socket has been closed. An attempt to read
3912 the next input character can distinguish the two cases. Previously, Exim
3913 would have sent an OK response which the client would never have see.
3914 This could lead to message repetition. This fix should cure that, at
3915 least in a lot of common cases.
3917 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3918 advertised in response to EHLO.
3924 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3925 fixed by 4.65/MH/01 (is this a record?) are fixed:
3927 (i) An empty string was always treated as zero by the numeric comparison
3928 operators. This behaviour has been restored.
3930 (ii) It is documented that the numeric comparison operators always treat
3931 their arguments as decimal numbers. This was broken in that numbers
3932 starting with 0 were being interpreted as octal.
3934 While fixing these problems I realized that there was another issue that
3935 hadn't been noticed. Values of message_size_limit (both the global option
3936 and the transport option) were treated as octal if they started with 0.
3937 The documentation was vague. These values are now always treated as
3938 decimal, and I will make that clear in the documentation.
3944 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3945 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3948 MH/01 Don't check that the operands of numeric comparison operators are
3949 integers when their expansion is in "skipping" mode (fixes bug
3950 introduced by 4.64-PH/07).
3952 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3953 child addresses, Exim now panics and dies. Previously, because the count
3954 is held in a short int, deliveries were likely to be lost. As such a
3955 large number of recipients for a single message is ridiculous
3956 (performance will be very, very poor), I have chosen to impose a limit
3957 rather than extend the field.
3963 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3964 leftover -K file (the existence of which was triggered by #402).
3965 While we were at it, introduced process PID as part of the -K
3966 filename. This should rule out race conditions when creating
3969 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3970 processing considerably. Previous code took too long for large mails,
3971 triggering a timeout which in turn triggers #401.
3973 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3974 in the DK code in transports.c. sendfile() is not really portable,
3975 hence the _LINUX specificness.
3977 TF/01 In the add_headers option to the mail command in an Exim filter,
3978 there was a bug that Exim would claim a syntax error in any
3979 header after the first one which had an odd number of characters
3982 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
3983 callout verification, Exim cached a "reject" for the entire domain. This
3984 is correct for most verifications, but it is not correct for a recipient
3985 verification with use_sender or use_postmaster set, because in that case
3986 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
3987 case of MAIL FROM:<> rejection from other early rejections (e.g.
3988 rejection of HELO). When verifying a recipient using a non-null MAIL
3989 address, the cache is ignored if it shows MAIL FROM:<> rejection.
3990 Whatever the result of the callout, the value of the domain cache is
3991 left unchanged (for any other kind of callout, getting as far as trying
3992 RCPT means that the domain itself is ok).
3994 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
3997 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
3998 manifest itself as EPIPE rather than ECONNECT. When tidying away a
3999 session, the daemon ignores ECONNECT errors and logs others; it now
4000 ignores EPIPE as well.
4002 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4003 (quoted-printable decoding).
4005 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4006 later the small subsequent patch to fix an introduced bug.
4008 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4010 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4012 PH/08 An error is now given if message_size_limit is specified negative.
4014 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4015 to be given (somewhat) arbitrary names.
4017 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4020 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4021 miscellaneous code fixes
4023 PH/10 Added the log_reject_target ACL modifier to specify where to log
4026 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4027 hostname. This is wrong, because it relates to the incoming message (and
4028 probably the interface on which it is arriving) and not to the outgoing
4029 callout (which could be using a different interface). This has been
4030 changed to use the value of the helo_data option from the smtp transport
4031 instead - this is what is used when a message is actually being sent. If
4032 there is no remote transport (possible with a router that sets up host
4033 addresses), $smtp_active_hostname is used.
4035 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4036 tweaks were necessary in order to get it to work (see also 21 below):
4037 (a) The code assumed that strncpy() returns a negative number on buffer
4038 overflow, which isn't the case. Replaced with Exim's string_format()
4040 (b) There were several signed/unsigned issues. I just did the minimum
4041 hacking in of casts. There is scope for a larger refactoring.
4042 (c) The code used strcasecmp() which is not a standard C function.
4043 Replaced with Exim's strcmpic() function.
4044 (d) The code set only $1; it now sets $auth1 as well.
4045 (e) A simple test gave the error "authentication client didn't specify
4046 service in request". It would seem that Dovecot has changed its
4047 interface. Fortunately there's a specification; I followed it and
4048 changed what the client sends and it appears to be working now.
4050 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4053 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4054 address A is aliased to B and C, where B exists and C does not. Without
4055 -v the output is "A verified" because verification stops after a
4056 successful redirection if more than one address is generated. However,
4057 with -v the child addresses are also verified. Exim was outputting "A
4058 failed to verify" and then showing the successful verification for C,
4059 with its parentage. It now outputs "B failed to verify", showing B's
4060 parentage before showing the successful verification of C.
4062 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4063 look up a TXT record in a specific list after matching in a combined
4066 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4067 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4068 they consult the DNS. I had assumed they would set it the way they
4069 wanted; and indeed my experiments on Linux seem to show that in some
4070 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4071 To be on the safe side, however, I have now made the interface to
4072 host_find_byname() similar to host_find_bydns(), with an argument
4073 containing the DNS resolver options. The host_find_byname() function now
4074 sets these options at its start, just as host_find_bydns() does. The smtp
4075 transport options dns_qualify_single and dns_search_parents are passed to
4076 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4077 of host_find_byname() use the default settings of RES_DEFNAMES
4078 (qualify_single) but not RES_DNSRCH (search_parents).
4080 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4081 spool_read_header() do less string testing, by means of a preliminary
4082 switch on the second character of optional "-foo" lines. (This is
4083 overdue, caused by the large number of possibilities that now exist.
4084 Originally there were few.) While I was there, I also converted the
4085 str(n)cmp tests so they don't re-test the leading "-" and the first
4086 character, in the hope this might squeeze out yet more improvement.
4088 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4089 flag allowing group syntax was set by the header_syntax check but not
4090 turned off, possible causing trouble later; (2) The flag was not being
4091 set at all for the header_verify test, causing "group"-style headers to
4092 be rejected. I have now set it in this case, and also caused header_
4093 verify to ignore an empty address taken from a group. While doing this, I
4094 came across some other cases where the code for allowing group syntax
4095 while scanning a header line wasn't quite right (mostly, not resetting
4096 the flag correctly in the right place). These bugs could have caused
4097 trouble for malformed header lines. I hope it is now all correct.
4099 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4100 with the "reply" argument non-NULL. The code, however (which originally
4101 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4102 but it didn't always do it. This confused somebody who was copying the
4103 code for some other use. I have removed all the tests.
4105 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4106 feature that was used to support insecure browsers during the U.S. crypto
4107 embargo. It requires special client support, and Exim is probably the
4108 only MTA that supported it -- and would never use it because real RSA is
4109 always available. This code has been removed, because it had the bad
4110 effect of slowing Exim down by computing (never used) parameters for the
4111 RSA_EXPORT functionality.
4113 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4114 authenticator to fail if there's a tab character in the incoming data
4115 (there should never be unless someone is messing about, as it's supposed
4116 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4117 if the connection is using TLS or if the remote IP is the same as the
4118 local IP, and the "valid-client-cert option" if a client certificate has
4121 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4122 authenticators. This can be used for authorization after authentication
4123 succeeds. (In the case of plaintext, it servers for both authentication
4126 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4127 if any retry times were supplied.
4129 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4130 connection, where there is no client IP address to check. In this
4131 situation, the verify now always succeeds.
4133 PH/25 Applied John Jetmore's -Mset patch.
4135 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4137 PH/27 In a string expansion for a processed (not raw) header when multiple
4138 headers of the same name were present, leading whitespace was being
4139 removed from all of them, but trailing whitespace was being removed only
4140 from the last one. Now trailing whitespace is removed from each header
4141 before concatenation. Completely empty headers in a concatenation (as
4142 before) are ignored.
4144 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4145 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4147 PH/29 [Removed. This was a change that I later backed out, and forgot to
4148 correct the ChangeLog entry (that I had efficiently created) before
4149 committing the later change.]
4151 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4152 address errors (4xx response to RCPT) over the same connection as other
4153 messages routed to the same hosts. Such deliveries are always "forced",
4154 so retry times are not inspected. This resulted in far too many retries
4155 for the affected addresses. The effect occurred only when there were more
4156 hosts than the hosts_max_try setting in the smtp transport when it had
4157 the 4xx errors. Those hosts that it had tried were not added to the list
4158 of hosts for which the message was waiting, so if all were tried, there
4159 was no problem. Two fixes have been applied:
4161 (i) If there are any address or message errors in an SMTP delivery, none
4162 of the hosts (tried or untried) are now added to the list of hosts
4163 for which the message is waiting, so the message should not be a
4164 candidate for sending over the same connection that was used for a
4165 successful delivery of some other message. This seems entirely
4166 reasonable: after all the message is NOT "waiting for some host".
4167 This is so "obvious" that I'm not sure why it wasn't done
4168 previously. Hope I haven't missed anything, but it can't do any
4169 harm, as the worst effect is to miss an optimization.
4171 (ii) If, despite (i), such a delivery is accidentally attempted, the
4172 routing retry time is respected, so at least it doesn't keep
4173 hammering the server.
4175 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4176 in ${readsocket because some servers need this prod.
4178 PH/32 Added some extra debug output when updating a wait-xxx database.
4180 PH/33 The hint "could be header name not terminated by colon", which has been
4181 given for certain expansion errors for a long time, was not being given
4182 for the ${if def:h_colon_omitted{... case.
4184 PH/34 The spec says: "With one important exception, whenever a domain list is
4185 being scanned, $domain contains the subject domain." There was at least
4186 one case where this was not true.
4188 PH/35 The error "getsockname() failed: connection reset by peer" was being
4189 written to the panic log as well as the main log, but it isn't really
4190 panic-worthy as it just means the connection died rather early on. I have
4191 removed the panic log writing for the ECONNRESET error when getsockname()
4194 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4195 runs only) independently of the message's sender address. This meant
4196 that, if the 4xx error was in fact related to the sender, a different
4197 message to the same recipient with a different sender could confuse
4198 things. In particular, this can happen when sending to a greylisting
4199 server, but other circumstances could also provoke similar problems.
4200 I have changed the default so that the retry time for these errors is now
4201 based a combination of the sender and recipient addresses. This change
4202 can be overridden by setting address_retry_include_sender=false in the
4205 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4206 remote server are returned as part of bounce messages. This was not
4207 happening for LMTP over a pipe (the lmtp transport), but now it is the
4208 same for both kinds of LMTP.
4210 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4211 in header lines that were in fact CNAMEs. This is no longer the case.
4213 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4214 and queue runs started by the daemon processed all messages. This has
4215 been fixed so that -R and -S can now usefully be given with -q<time>.
4217 PH/40 Import PCRE release 6.7 (fixes some bugs).
4219 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4221 PH/42 Give an error if -q is specified more than once.
4223 PH/43 Renamed the variables $interface_address and $interface_port as
4224 $received_ip_address and $received_port, to make it clear that these
4225 values apply to message reception, and not to the outgoing interface when
4226 a message is delivered. (The old names remain recognized, of course.)
4228 PH/44 There was no timeout on the connect() call when using a Unix domain
4229 socket in the ${readsocket expansion. There now is.
4231 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4232 be meaningful with "accept".
4234 SC/01 Eximstats V1.43
4235 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4237 SC/02 Eximstats V1.44
4238 Use a glob alias rather than an array ref in the generated
4239 parser. This improves both readability and performance.
4241 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4242 Collect SpamAssassin and rejection statistics.
4243 Don't display local sender or destination tables unless
4244 there is data to show.
4245 Added average volumes into the top table text output.
4247 SC/04 Eximstats V1.46
4248 Collect data on the number of addresses (recipients)
4249 as well as the number of messages.
4251 SC/05 Eximstats V1.47
4252 Added 'Message too big' to the list of mail rejection
4253 reasons (thanks to Marco Gaiarin).
4255 SC/06 Eximstats V1.48
4256 Mainlog lines which have GMT offsets and are too short to
4257 have a flag are now skipped.
4259 SC/07 Eximstats V1.49 (Alain Williams)
4260 Added the -emptyok flag.
4262 SC/08 Eximstats V1.50
4263 Fixes for obtaining the IP address from reject messages.
4265 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4266 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4267 whitespace changes from 4.64-PH/27
4269 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4272 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4273 are found, allow negative numbers in numeric criteria)
4275 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4277 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4278 to match changes made in 4.64-PH/43
4280 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4282 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4283 "domain" is actually an IP address, and force a failure. This locks out
4284 those revolvers/nameservers that support "A-for-A" lookups, in
4285 contravention of the specifications.
4287 PH/48 When a host name was looked up from an IP address, and the subsequent
4288 forward lookup of the name timed out, the host name was left in
4289 $sender_host_name, contrary to the specification.
4291 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4292 restricted to single-key lookups, Exim was not diagnosing an error if
4293 * or *@ was used with a query-style lookup.
4295 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4297 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4298 long ago, but noone interested enough thought of it. Let's just say that
4299 the "1.1" means that there are some new functions that weren't there at
4300 some point in the past.
4302 PH/51 Error processing for expansion failure of helo_data from an smtp
4303 transport during callout processing was broken.
4305 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4306 tested/used via the -bh/-bhc/-bs options.
4308 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4309 bug, fixed in subsequent PCRE releases).
4311 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4312 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4314 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4320 SC/01 Use a glob alias rather than an array ref in eximstats generated
4321 parser. This improves both readability and performance.
4323 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4324 Don't display local sender or destination tables in eximstats unless
4325 there is data to show.
4326 Added average volumes into the eximstats top table text output.
4328 SC/03 Collect data on the number of addresses (recipients) as well
4329 as the number of messages in eximstats.
4331 TF/01 Correct an error in the documentation for the redirect router. Exim
4332 does (usually) call initgroups() when daemonizing.
4334 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4335 with consistent privilege compared to when running as a daemon.
4337 TF/03 Note in the spec that $authenticated_id is not set for local
4338 submissions from trusted users.
4340 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4341 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4343 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4344 by adding some example configuration directives to the default
4345 configuration file. A little bit of work is required to uncomment the
4346 directives and define how usernames and passwords are checked, but
4347 there is now a framework to start from.
4349 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4350 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4351 without this. I don't know how relevant this is to other LDAP libraries.
4353 PH/02 Add the verb name to the "unknown ACL verb" error.
4355 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4357 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4359 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4360 directory not expanded when it should be if an expanded home directory
4361 was set for the address (which is overridden by the transport).
4363 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4366 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4367 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4368 because it is too late at that time, and has no effect.
4370 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4371 security issue with \' (bugzilla #107). I could not use the
4372 PQescapeStringConn() function, because it needs a PGconn value as one of
4375 PH/08 When testing addresses using -bt, indicate those final addresses that
4376 are duplicates that would not cause an additional delivery. At least one
4377 person was confused, thinking that -bt output corresponded to deliveries.
4378 (Suppressing duplicates isn't a good idea as you lose the information
4379 about possibly different redirections that led to the duplicates.)
4381 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4382 systems where poll() doesn't work, in particular OS X.
4384 PH/10 Added more information to debugging output for retry time not reached.
4386 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4387 operations in malware.c.
4389 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4392 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4393 syslog_duplication set false, log lines that would normally be written
4394 both the the main log and to the reject log were not written to syslog at
4397 PH/14 In the default configuration, change the use of "message" in ACL warn
4398 statements to "add_header".
4400 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4401 not followed by a command (e.g. "seen endif").
4403 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4404 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4407 PH/17 Added extra conditions to the default value of delay_warning_condition
4411 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4412 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4413 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4416 The Auto-Submitted: and various List- headers are standardised, whereas I
4417 don't think Precedence: ever was.
4419 PH/18 Refactored debugging code in route_finduser() to show more information,
4420 in particular, the error code if getpwnam() issues one.
4422 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4423 This is apparently needed in addition to the PH/07 change above to avoid
4424 any possible encoding problems.
4426 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4427 but not after initializing Perl.
4429 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4430 output them only if debugging. By default they are written stderr,
4431 apparently, which is not desirable.
4433 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4436 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4439 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4441 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4442 authenticated or an ident call has been made. Suppress the default
4443 values for $authenticated_id and $authenticated_sender (but permit -oMai
4444 and -oMas) when testing with -bh.
4446 PH/24 Re-jigged the order of the tests in the default configuration so that the
4447 tests for valid domains and recipients precede the DNS black list and CSA
4448 tests, on the grounds that those ones are more expensive.
4450 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4451 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4452 This bug exists in every version of Exim that I still have, right back to
4455 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4456 However, an attempt to turn on case-sensitivity in a regex key by
4457 including (?-i) didn't work because the subject string was already
4458 lowercased, and the effects were non-intuitive. It turns out that a
4459 one-line patch can be used to allow (?-i) to work as expected.
4465 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4466 other effects) broke the use of negated acl sub-conditions.
4468 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4471 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4472 "Deny" causes Exim to reject the incoming connection with a 554 error.
4473 Unfortunately, if there is a major crisis, such as a disk failure,
4474 tcp-wrappers gives "deny", whereas what one would like would be some
4475 kind of temporary error. A kludge has been added to help with this.
4476 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4477 554 error is used if errno is still zero or contains ENOENT (which occurs
4478 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4481 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4483 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4484 errors. Otherwise a message that provokes a temporary error (when other
4485 messages do not) can cause a whole host to time out.
4487 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4488 addresses were routed directly to files or pipes from a redirect router.
4489 File deliveries just didn't batch; pipe deliveries might have suffered
4492 PH/06 A failure to get a lock for a hints database would erroneously always say
4493 "Failed to get write lock", even when it was really a read lock.
4495 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4496 of 0600. This has been changed to use the value of the lockfile_mode
4497 option (which defaults to 0600).
4499 PH/08 Applied small patch from the Sieve maintainer.
4501 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4502 folder from quota calculations, a direct delivery into this folder messed
4503 up the contents of the maildirsize file. This was because the regex was
4504 used only to exclude .Trash (or whatever) when the size of the mailbox
4505 was calculated. There was no check that a delivery was happening into an
4506 excluded directory. This bug has been fixed by ignoring all quota
4507 processing for deliveries into excluded directories.
4509 PH/10 Added the maildirfolder_create_regex option to appendfile.
4515 PH/01 The code for finding all the local interface addresses on a FreeBSD
4516 system running IPv6 was broken. This may well have applied to all BSD
4517 systems, as well as to others that have similar system calls. The broken
4518 code found IPv4 interfaces correctly, but gave incorrect values for the
4519 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4520 that it would not match correctly against @[] and not recognize the IPv6
4523 PH/02 The ipliteral router was not recognizing addresses of the form user@
4524 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4526 PH/03 Added disable_ipv6.
4528 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4529 decoded form, because it is most often used to construct To: headers
4530 lines in autoreplies, and the decoded form may well be syntactically
4531 invalid. However, $reply_address has leading white space removed, and all
4532 newlines turned into spaces so that the autoreply transport does not
4535 PH/05 If group was specified without a user on a router, and no group or user
4536 was specified on a transport, the group from the router was ignored.
4538 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4539 for visible compile-time settings that can be used to change these
4540 numbers, for those that want even more. Backwards compatibility with old
4541 spool files has been maintained. However, going back to a previous Exim
4542 release will lost any variables that are in spool files.
4544 PH/07 Two small changes when running in the test harness: increase delay when
4545 passing a TCP/IP connection to a new process, in case the original
4546 process has to generate a bounce, and remove special handling of
4547 127.0.0.2 (sic), which is no longer necessary.
4549 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4550 be the same on different OS.
4552 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4555 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4556 whether --show-vars was specified or not
4558 JJ/02 exipick: Added support for new ACL variable spool format introduced
4561 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4562 syntactically invalid From: or Reply-to: line, and a filter used this to
4563 generate an autoreply, and therefore failed to obtain an address for the
4564 autoreply, Exim could try to deliver to a non-existent relative file
4565 name, causing unrelated and misleading errors. What now happens is that
4566 it logs this as a hard delivery error, but does not attempt to create a
4569 PH/11 The exinext utility has a -C option for testing purposes, but although
4570 the given file was scanned by exinext itself; it wasn't being passed on
4571 when Exim was called.
4573 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4574 an end-of-file indication when reading a command response.
4576 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4577 compiled. In many other places in Exim, IPv6 addresses are always
4578 recognized, so I have changed this. It also means that IPv4 domain
4579 literals of the form [IPV4:n.n.n.n] are now always recognized.
4581 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4582 used if the router is not running as root, for example, when verifying at
4583 ACL time, or when using -bh. The debugging output from this situation was
4584 non-existent - all you got was a failure to exec. I have made two
4587 (a) Failures to set uid/gid, the current directory, or a process leader
4588 in a subprocess such as that created by queryprogram now generate
4589 suitable debugging output when -d is set.
4591 (b) The queryprogram router detects when it is not running as root,
4592 outputs suitable debugging information if -d is set, and then runs
4593 the subprocess without attempting to change uid/gid.
4595 PH/15 Minor change to Makefile for building test_host (undocumented testing
4598 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4599 additional section of a DNS packet that returns MX or SRV records.
4600 Instead, it always explicitly searches for A/AAAA records. This avoids
4601 major problems that occur when a DNS server includes only records of one
4602 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4603 fixed another bug: if SRV records were looked up and the corresponding
4604 address records were *not* found in the additional section, the port
4605 values from the SRV records were lost.
4607 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4608 using the correct key (the original address) when searching the retry
4609 rules in order to find which one to use for generating the retry hint.
4611 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4612 adding the default one. Similarly, if it contains a Reply-To: header, the
4613 errors_reply_to option, if set, is not used.
4615 PH/19 When calculating a retry time, Exim used to measure the "time since
4616 failure" by looking at the "first failed" field in the retry record. Now
4617 it does not use this if it is later than than the arrival time of the
4618 message. Instead it uses the arrival time. This makes for better
4619 behaviour in cases where some deliveries succeed, thus re-setting the
4620 "first failed" field. An example is a quota failure for a huge message
4621 when small messages continue to be delivered. Without this change, the
4622 "time since failure" will always be short, possible causing more frequent
4623 delivery attempts for the huge message than are intended.
4624 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4626 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4627 $1, $2, $3) because the numerical variables can be reset during some
4628 expansion items (e.g. "match"), thereby losing the authentication data.
4630 PH/21 Make -bV show the size of off_t variables so that the test suite can
4631 decide whether to run tests for quotas > 2G.
4633 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4634 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4635 filecount value is greater than 2G or if a quota value is greater than 2G
4636 on a system where the size of off_t is not greater than 4, a panic error
4639 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4640 never match. The debug and -bh output now contains an explicit error
4641 message indicating a malformed IPv4 address or mask.
4643 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4644 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4645 PH/23 above applies.
4647 PH/25 Do not write to syslog when running in the test harness. The only
4648 occasion when this arises is a failure to open the main or panic logs
4649 (for which there is an explicit test).
4651 PH/26 Added the /no_tell option to "control=freeze".
4653 PH/27 If a host name lookup failed very early in a connection, for example, if
4654 the IP address matched host_lookup and the reverse lookup yielded a name
4655 that did not have a forward lookup, an error message of the form "no IP
4656 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4657 could be logged. Now it outputs the IP address instead of "NULL".
4659 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4660 allows the sender and the authenticated sender to be set when
4661 submitting a message from within Exim. Since child_open_exim() is
4662 documented for local_scan(), the new function should be too.
4664 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4665 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4666 results in an empty string is now treated as unset.
4668 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4670 PH/31 Added sender_verify_fail logging option.
4672 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4673 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4674 tidied the source and removed it altogether.
4676 PH/33 When a queue run was abandoned because the load average was too high, a
4677 log line was always written; now it is written only if the queue_run log
4678 selector is set. In addition, the log line for abandonment now contains
4679 information about the queue run such as the pid. This is always present
4680 in "start" and "stop" lines but was omitted from the "abandon" line.
4682 PH/34 Omit spaces between a header name and the colon in the error message that
4683 is given when verify = headers_syntax fails (if there are lots of them,
4684 the message gets confusing).
4686 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4687 names, as there are now some PTR records that contain slashes. This check
4688 is only to protect against broken name servers that fall over on strange
4689 characters, so the fact that it applies to all lookups doesn't matter.
4691 PH/36 Now that the new test suite is complete, we can remove some of the
4692 special code in Exim that was needed for the old test suite. For example,
4693 sorting DNS records because real resolvers return them in an arbitrary
4694 order. The new test suite's fake resolver always returns records in the
4697 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4698 bounces) except when queue_only is set, to avoid logging races between
4699 the different processes.
4701 PH/38 Panic-die if .include specifies a non-absolute path.
4703 PH/39 A tweak to the "H" retry rule from its user.
4705 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4706 a label. They prevented compilation on older perls.
4708 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4709 a warning to be raised on newish perls.
4711 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4712 on queue. Changes to match documented behaviour of showing count of
4713 messages matching specified criteria.
4715 PH/40 Changed the default ident timeout from 30s to 5s.
4717 PH/41 Added support for the use of login_cap features, on those BSD systems
4718 that have them, for controlling the resources used by pipe deliveries.
4720 PH/42 The content-scanning code uses fopen() to create files in which to put
4721 message data. Previously it was not paying any attention to the mode of
4722 the files. Exim runs with umask(0) because the rest of the code creates
4723 files with open(), and sets the required mode explicitly. Thus, these
4724 files were ending up world-writeable. This was not a big issue, because,
4725 being within the spool directory, they were not world-accessible. I have
4726 created a function called modefopen, which takes an additional mode
4727 argument. It sets umask(777), creates the file, chmods it to the required
4728 mode, then resets the umask. All the relevant calls to fopen() in the
4729 content scanning code have been changed to use this function.
4731 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4732 to 24 hours. This avoids potential overflow problems when processing G
4733 and H retry rules. I suspect nobody ever tinkers with this value.
4735 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4737 PH/45 When the plaintext authenticator is running as a client, the server's
4738 challenges are checked to ensure they are valid base64 strings. By
4739 default, the authentication attempt is cancelled if an invalid string is
4740 received. Setting client_ignore_invalid_base64 true ignores these errors.
4741 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4742 they are received. Thus, the responses can be made to depend on the
4743 challenges. If an invalid string is ignored, an empty string is placed in
4746 PH/46 Messages that are created by the autoreply transport now contains a
4747 References: header, in accordance with RFCs 2822 and 3834.
4749 PH/47 Added authenticated_sender_force to the smtp transport.
4751 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4753 PH/49 Installed latest patch from the Sieve maintainer.
4755 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4756 was also set, the appendfile transport was unnecessarily scanning a
4757 directory of message files (e.g. for maildir delivery) to find the count
4758 of files (along with the size), even though it did not need this
4759 information. It now does the scan only if it needs to find either the
4760 size of the count of files.
4762 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4764 PH/52 Two bugs concerned with error handling when the smtp transport is
4767 (i) Exim was not creating retry information for temporary errors given
4768 for individual recipients after the DATA command when the smtp transport
4769 was used in LMTP mode. This meant that they could be retried too
4770 frequently, and not timed out correctly.
4772 (ii) Exim was setting the flag that allows error details to be returned
4773 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4774 recipients that were returned after the DATA command.
4776 PH/53 This is related to PH/52, but is more general: for any failing address,
4777 when detailed error information was permitted to be returned to the
4778 sender, but the error was temporary, then after the final timeout, only
4779 "retry timeout exceeded" was returned. Now it returns the full error as
4780 well as "retry timeout exceeded".
4782 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4783 do this, and (what is worse) MTAs that accept it.
4785 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4786 will now be deprecated.
4788 PH/56 New os.c-cygwin from the Cygwin maintainer.
4790 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4791 formats (previously only available in exim formats via -bpr, -bpru,
4792 and -bpra. Now also available in native and exiqgrep formats)
4794 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4795 with very large, slow to parse queues
4797 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4799 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4801 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4802 responses to authentication challenges, though it was showing the
4803 challenges; (ii) I've removed the CR characters from the debug output for
4806 PH/58 Allow for the insertion of a newline as well as a space when a string
4807 is turned into more than one encoded-word during RFC 2047 encoding. The
4808 Sieve code now uses this.
4810 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4811 data_4xx, lost_connection, tls_required.
4813 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4814 message was being sent as an SMTP response.
4816 PH/61 Add -l and -k options to exicyclog.
4818 PH/62 When verifying, if an address was redirected to one new address, so that
4819 verification continued, and the new address failed or deferred after
4820 having set something in $address_data, the value of $address_data was not
4821 passed back to the ACL. This was different to the case when no
4822 redirection occurred. The value is now passed back in both cases.
4824 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4825 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4826 use login_cap.h, so on its own it isn't the distinguishing feature. The
4827 new name refers directly to the setclassresources() function.
4829 PH/65 Added configuration files for NetBSD3.
4831 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4833 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4834 is preferred over IPv4.
4836 PH/68 The bounce_return_message and bounce_return_body options were not being
4837 honoured for bounces generated during the reception of non-SMTP messages.
4838 In particular, this applied to messages rejected by the ACL. This bug has
4839 been fixed. However, if bounce_return_message is true and bounce_return_
4840 body is false, the headers that are returned for a non-SMTP message
4841 include only those that have been read before the error was detected.
4842 (In the case of an ACL rejection, they have all been read.)
4844 PH/69 The HTML version of the specification is now built in a directory called
4845 spec_html instead of spec.html, because the latter looks like a path with
4846 a MIME-type, and this confuses some software.
4848 PH/70 Catch two compiler warnings in sieve.c.
4850 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4851 function verify_get_ident() calls ip_connect() to connect a socket, but
4852 if the "connect()" function timed out, ip_connect() used to close the
4853 socket. However, verify_get_ident() also closes the socket later, and in
4854 between Exim writes to the log, which may get opened at this point. When
4855 the socket was closed in ip_connect(), the log could get the same file
4856 descriptor number as the socket. This naturally causes chaos. The fix is
4857 not to close the socket in ip_connect(); the socket should be closed by
4858 the function that creates it. There was only one place in the code where
4859 this was missing, in the iplookup router, which I don't think anybody now
4860 uses, but I've fixed it anyway.
4862 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4863 well as to direct DNS lookups. Otherwise the handling of names in host
4864 lists is inconsistent and therefore confusing.
4870 PH/01 Two changes to the default runtime configuration:
4872 (1) Move the checks for relay_from_hosts and authenticated clients from
4873 after to before the (commented out) DNS black list checks.
4875 (2) Add control=submission to the relay_from_hosts and authenticated
4876 clients checks, on the grounds that messages accepted by these
4877 statements are most likely to be submissions.
4879 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4881 (1) Generate an error if the third argument for the ${prvs expansion is
4884 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4887 (3) Reset the variables that are obtained from the first argument of
4888 ${prvscheck and used in the second argument before leaving the code,
4889 because their memory is reclaimed, so using them afterwards may do
4892 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4893 one (it's much easier than Tom thought :-).
4895 (5) Because of (4), we can now allow for the use of $prvscheck_result
4896 inside the third argument.
4898 PH/03 For some reason, the default setting of PATH when running a command from
4899 a pipe transport was just "/usr/bin". I have changed it to
4902 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4903 anything to be listed in the output from -bV.
4905 PH/05 When a filter generated an autoreply, the entire To: header line was
4906 quoted in the delivery log line, like this:
4908 => >A.N.Other <ano@some.domain> <original@ddress> ...
4910 This has been changed so that it extracts the operative address. There
4911 may be more than one such address. If so, they are comma-separated, like
4914 => >ano@some.domain,ona@other.domain <original@ddress> ...
4916 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4917 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4918 not being looked up in the rDNS to get a host name, Exim was showing the
4919 IP address twice in Received: lines, even though the IP addresses were
4920 identical. For example:
4922 Received: from [1.2.3.4] (helo=[1.2.3.4])
4924 However, if the real host name was known, it was omitting the HELO data
4925 if it matched the actual IP address. This has been tidied up so that it
4926 doesn't show the same IP address twice.
4928 PH/07 When both +timestamp and +memory debugging was on, the value given by
4929 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4930 called by the debug printing, thereby overwriting the timestamp buffer.
4931 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4933 PH/08 When the original message was included in an autoreply transport, it
4934 always said "this is a copy of the message, including all the headers",
4935 even if body_only or headers_only was set. It now gives an appropriate
4938 PH/09 Applied a patch from the Sieve maintainer which:
4940 o fixes some comments
4941 o adds the (disabled) notify extension core
4942 o adds some debug output for the result of if/elsif tests
4943 o points to the current vacation draft in the documentation
4944 and documents the missing references header update
4948 o fixes a bug in processing the envelope test (when testing
4949 multiple envelope elements, the last element determined the
4952 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4953 Electronic Mail") by including:
4955 Auto-submitted: auto-generated
4957 in the messages that it generates (bounce messages and others, such as
4958 warnings). In the case of bounce messages for non-SMTP messages, there was
4959 also a typo: it was using "Auto_submitted" (underscore instead of
4960 hyphen). Since every message generated by Exim is necessarily in response
4961 to another message, thes have all been changed to:
4963 Auto-Submitted: auto-replied
4965 in accordance with these statements in the RFC:
4967 The auto-replied keyword:
4969 - SHOULD be used on messages sent in direct response to another
4970 message by an automatic process,
4972 - MUST NOT be used on manually-generated messages,
4974 - MAY be used on Delivery Status Notifications (DSNs) and Message
4975 Disposition Notifications (MDNs),
4977 - MUST NOT be used on messages generated by automatic or periodic
4978 processes, except for messages which are automatic responses to
4981 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
4982 to the default Received: header definition.
4984 PH/12 Added log selector acl_warn_skipped (default on).
4986 PH/13 After a successful wildlsearch lookup, discard the values of numeric
4987 variables because (a) they are in the wrong storage pool and (b) even if
4988 they were copied, it wouldn't work properly because of the caching.
4990 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
4991 checking when decoding. Apparently there are clients that generate
4992 overlong encoded strings. Why am I not surprised?
4994 PH/15 If the first argument of "${if match_address" was not empty, but did not
4995 contain an "@" character, Exim crashed. Now it writes a panic log message
4996 and treats the condition as false.
4998 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5000 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5001 extension "envelope-auth". The code is finished and in agreement with
5002 other implementations, but there is no documentation so far and in fact,
5003 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5004 not changing the active code.
5006 Print executed "if" and "elsif" statements when debugging is used. This
5007 helps a great deal to understand what a filter does.
5009 Document more things not specified clearly in RFC3028. I had all this
5010 sorted out, when out of a sudden new issues came to my mind. Oops."
5012 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5015 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5016 canonical form (as documented). However, after a host name lookup from
5017 the IP address, check_host() was doing a simple string comparison with
5018 addresses acquired from the DNS when checking that the found name did
5019 have the original IP as one of its addresses. Since any found IPv6
5020 addresses are likely to be in abbreviated form, the comparison could
5021 fail. Luckily, there already exists a function for doing the comparison
5022 by converting both addresses to binary, so now that is used instead of
5023 the text comparison.
5025 PH/20 There was another similar case to PH/19, when a complete host name was
5026 given in a host list; looking up its IP address could give an abbreviated
5027 form, whereas the current host's name might or might not be abbreviated.
5028 The same fix has been applied.
5034 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5035 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5038 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5039 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5041 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5043 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5044 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5045 Also, for "Auto-Submitted", treat anything other than "no" as
5046 non-personal, in accordance with RFC 3834. (Previously it treated
5047 anything starting "auto-" as non-personal.)
5049 TF/01 The control=submission/name=... option had a problem with syntax
5050 errors if the name included a slash character. The /name= option
5051 now slurps the rest of the string, so it can include any characters
5052 but it must come last in the list of options (after /sender_retain
5055 PH/05 Some modifications to the interface to the fake nameserver for the new
5063 TK/01 Added the "success_on_redirect" address verification option. See
5064 NewStuff for rationale and an example.
5066 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5068 PH/02 Patch to exigrep to allow it to work on syslog lines.
5070 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5071 fread() to skip over the body file's header line, because in Cygwin the
5072 header line is locked and is inaccessible.
5074 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5075 co-exist for some time) to make it clear that it is the Exim ID that is
5076 referenced, not the Message-ID: header line.
5078 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5079 string_format() function, because snprintf() does not exist on all
5082 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5083 ${stat: expansion item.
5085 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5086 protocol synchronization error", to keep the pedants happy.
5088 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5089 well as for IRIX systems, when gcc is being used. See the host.c source
5092 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5094 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5097 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5098 transport and to the smtp transport in LMTP mode.
5100 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5102 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5103 run a filter in a subprocess. This could lead to confusion in subsequent
5104 lookups in the parent process. There should also be a search_tidyup() at
5105 the end of the subprocess.
5107 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5108 only if the host matched helo_try_verify_hosts, which caused the
5109 verification to occur when the EHLO/HELO command was issued. The ACL just
5110 tested the remembered result. Now, if a previous verification attempt has
5111 not happened, "verify = helo" does it there and then.
5113 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5115 TK/03 Fix log output including CR from clamd.
5117 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5118 exist provoked a memory error which could cause a segfault.
5120 PH/15 Installed PCRE 6.2
5122 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5124 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5125 of the problem. Specifically, suggested +O2 rather than +O1 for the
5128 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5130 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5131 record, the port was not correctly logged when the outgoing_port log
5132 selector was set (it logged the transort's default port).
5134 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5135 fallback_hosts, and "hosts" in the smtp transport.
5137 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5138 host errors such as "Connection refused".
5140 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5141 authentication with radiusclient 0.4.9:
5143 - Error returned from rc_read_config was caught wrongly
5144 - Username/password not passed on to radius server due to wrong length.
5146 The presumption is that some radiusclient API changes for 4.51/PH/17
5147 were not taken care of correctly. The code is still untested by me (my
5148 Linux distribution still has 0.3.2 of radiusclient), but it was
5149 contributed by a Radius user.
5151 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5152 expanding the "port" option of the smtp transport.
5154 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5155 while reading a MIME header. Thanks to Tom Hughes for a patch.
5157 PH/24 Include config.h inside local_scan.h so that configuration settings are
5160 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5161 that in an ACL for RCPT (for example), you can examine exactly what was
5164 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5165 commands, but it was not correctly comparing the address with the actual
5166 client host address. Thus, it would show the EHLO address in Received:
5167 header lines when this was not necessary.
5169 PH/27 Added the % operator to ${eval:}.
5171 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5172 it should be ignoring failures (because with -C, for example, it has lost
5173 privilege). It wasn't ignoring creation failures other than "already
5176 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5179 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5180 because an input line was too long, either on its own, or by virtue of
5181 too many continuations, the temporary file was not being removed, and the
5182 return code was incorrect.
5184 PH/31 Missing "BOOL" in function definition in filtertest.c.
5186 PH/32 Applied Sieve patches from the maintainer.
5188 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5190 PH/33 Added "verify = not_blind".
5192 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5193 Local/Makefile (with some defaults set). These are used in built scripts
5194 such as exicyclog, but they have never been used in the exim_install
5195 script (though there are many overriding facilities there). I have
5196 arranged that the exim_install script now takes note of these two
5199 PH/35 Installed configuration files for Dragonfly.
5201 PH/36 When a locally submitted message by a trusted user did not contain a
5202 From: header, and the sender address was obtained from -f or from an SMTP
5203 MAIL command, and the trusted user did not use -F to supply a sender
5204 name, $originator_name was incorrectly used when constructing a From:
5205 header. Furthermore, $originator_name was used for submission mode
5206 messages from external hosts without From: headers in a similar way,
5207 which is clearly wrong.
5209 PH/37 Added control=suppress_local_fixups.
5211 PH/38 When log_selector = +received_sender was set, and the addition of the
5212 sender made the log line's construction buffer exactly full, or one byte
5213 less than full, an overflow happened when the terminating "\n" was
5216 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5217 when the result of a list match is failure because a DNS lookup failed.
5219 PH/40 RM_COMMAND is now used in the building process.
5221 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5222 the "build-* directories that it finds.
5224 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5225 address in a domain literal was a prefix of an interface address.
5227 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5228 when verifying a sender address, unless rewrite_headers is false.
5230 PH/44 Wrote a long comment about why errors_to addresses are verified as
5231 recipients, not senders.
5233 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5234 the ratelimit ACL was added.
5236 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5238 PH/46 Added extra information about PostgreSQL errors to the error string.
5240 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5241 avoiding the need to install special zones in a real server. This is
5242 backwards compatible; if it can't find the fake resolver, it drops back.
5243 Thus, both old and new test suites can be run.
5245 TF/02 Added util/ratelimit.pl
5247 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5248 clock is set back in time.
5250 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5251 Candler <B.Candler@pobox.com>.
5253 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5254 OK for addresses that are the result of redirections.
5256 PH/48 A number of further additions for the benefit of the new test suite,
5257 including a fake gethostbyname() that interfaces to the fake DNS resolver
5260 TF/06 The fix for widen_domains has also been applied to qualify_single and
5261 search_parents which are the other dnslookup options that can cause
5264 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5267 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5269 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5270 service" if it encountered an MX record with an empty target hostname.
5271 The message is now "an MX or SRV record indicated no SMTP service".
5273 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5274 if the DNS of the sending site is misconfigured. This is quite a
5275 common situation. This change restores the behaviour of treating a
5276 helo verification defer as a failure.
5278 PH/51 If self=fail was set on a router, the bounce message did not include the
5279 actual error message.
5285 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5287 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5288 command itself ended in error, the underlying message about the transport
5289 filter timeout was being overwritten with the pipe command error. Now the
5290 underlying error message should be appended to the second error message.
5292 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5294 PH/02 Reduce the amount of output that "make" produces by default. Full output
5295 can still be requested.
5297 PH/03 The warning log line about a condition test deferring for a "warn" verb
5298 was being output only once per connection, rather than after each
5299 occurrence (because it was using the same function as for successful
5300 "warn" verbs). This seems wrong, so I have changed it.
5302 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5303 it should not have, which might have caused a crash in the right
5304 circumstances, but probably never did.
5306 PH/04 Installed a modified version of Tony Finch's patch to make submission
5307 mode fix the return path as well as the Sender: header line, and to
5308 add a /name= option so that you can make the user's friendly name appear
5311 TF/03 Added the control = fakedefer ACL modifier.
5313 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5314 Mark Lowes for thorough testing.
5316 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5318 TK/03 Merged latest SRS patch from Miles Wilton.
5320 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5321 with the definition in sysexits.h (which is #included earlier).
5322 Fortunately, Exim does not actually use EX_OK. The code used to try to
5323 preserve the sysexits.h value, by assuming that macro definitions were
5324 scanned for macro replacements. I have been disabused of this notion,
5325 so now the code just undefines EX_OK before #including unistd.h.
5327 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5328 in the smtp transport. When a block could not be written in a single
5329 write() function, the timeout was being re-applied to each part-write.
5330 This seems wrong - if the receiver was accepting one byte at a time it
5331 would take for ever. The timeout is now adjusted when this happens. It
5332 doesn't have to be particularly precise.
5334 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5335 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5337 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5338 without a domain if the check to <postmaster@domain> fails.
5340 SC/01 Eximstats: added -xls and the ability to specify output files
5341 (patch written by Frank Heydlauf).
5343 SC/02 Eximstats: use FileHandles for outputting results.
5345 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5347 SC/04 Eximstats: fixed display of large numbers with -nvr option
5349 SC/05 Eximstats: fixed merging of reports with empty tables.
5351 SC/06 Eximstats: added the -include_original_destination flag
5353 SC/07 Eximstats: removed tabs and trailing whitespace.
5355 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5357 TK/06 MBOX spool code: Add real "From " MBOX separator line
5358 so the .eml file is really in mbox format (even though
5359 most programs do not really care). Patch from Alex Miller.
5361 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5362 The latter is generated from $received_to and is only set if the
5363 message has one envelope recipient. SA can use these headers,
5364 obviously out-of-the-box. Patch from Alex Miller.
5366 PH/08 The ${def test on a variable was returning false if the variable's
5367 value was "0", contrary to what the specification has always said!
5368 The result should be true unless the variable is empty.
5370 PH/09 The syntax error of a character other than { following "${if
5371 def:variable_name" (after optional whitespace) was not being diagnosed.
5372 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5373 accidental colon was present, for example, could give incorrect results.
5375 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5376 result is used (not including appendfile, where other changes are about
5379 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5380 This involved changing a lot of size variables from int to off_t. It
5381 should work with maildirs and everything.
5383 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5384 spamd dying while we are connected to it.
5386 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5389 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5390 function for BDB 4.3.
5392 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5394 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5395 the sources for PCRE have been re-arranged and more files are now
5398 PH/15 The code I had for printing potentially long long variables in PH/11
5399 above was not the best (it lost precision). The length of off_t variables
5400 is now inspected at build time, and an appropriate printing format (%ld
5401 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5402 to be "long long int" or "long int". This is needed for the internal
5403 formatting function string_vformat().
5405 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5406 the configuration file to be ":syslog", then the script "guesses" where
5407 the logs files are, rather than using the compiled in default. In our
5408 case the guess is not the same as the compiled default, so the script
5409 suddenly stopped working when I started to use syslog. The patch checks
5410 to see if log_file_path is "". If so, it attempts to read it from exim
5411 with no configuration file to get the compiled in version, before it
5412 falls back to the previous guessing code."
5414 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5415 implementing BATV in an Exim configuration. See NewStuff for the gory
5418 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5419 Makefile that are specific to HP-UX.
5421 PH/18 If the "use_postmaster" option was set for a recipient callout together
5422 with the "random" option, the postmaster address was used as the MAIL
5423 FROM address for the random test, but not for the subsequent recipient
5424 test. It is now used for both.
5426 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5427 patch removes a few documentation additions to RFC 3028, because the
5428 latest draft now contains them. It adds the new en;ascii-case comparator
5429 and a new error check for 8bit text in MIME parts. Comparator and
5430 require names are now matched exactly. I enabled the subaddress
5431 extension, but it is not well tested yet (read: it works for me)."
5433 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5434 rework some of the code of TK/09 above to avoid the hardwired use of
5435 "%lld" and "long long". Replaced the call to snprintf() with a call to
5438 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5439 records point to non-existent hosts", "retry timeout exceeded", and
5440 "retry time not reached for any host after a long failure period".
5442 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5443 experimental DomainKeys support:
5445 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5446 (2) On an error such as an illegally used "control", the wrong name for
5447 the control was given.
5449 These problems did NOT occur unless DomainKeys support was compiled.
5451 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5453 PH/24 Added ${if match_ip condition.
5455 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5456 databases so that it will be absolutely obvious if a crash occurs in the
5457 DB library. This is a regular occurrence (often caused by mis-matched
5460 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5461 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5462 picked up on a user's system that detects such things. There doesn't seem
5463 to be a gcc warning option for this - only an attribute that has to be
5464 put on the function's prototype. It seems that in Fedora Core 4 they have
5465 set this on a number of new functions. No doubt there will be more in due
5468 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5469 specify a transport. However, if an address that was verified by such a
5470 router was the subject of a callout, Exim crashed because it tried to
5471 read the rcpt_include_affixes from the non-existent transport. Now it
5472 just assumes that the setting of that option is false. This bug was
5473 introduced by 4.51/PH/31.
5475 PH/28 Changed -d+all to exclude +memory, because that information is very
5476 rarely of interest, but it makes the output a lot bigger. People tend to
5477 do -d+all out of habit.
5479 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5480 code in os-type was giving problems when libc.so lives in lib64, like on
5483 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5484 aren't the modern standard, and it seems that some systems' include files
5485 don't always have them. Exim was already checking for some of the newer
5486 ones like T_AAAA, and defining it itself. I've added checks for all the
5487 record types that Exim uses.
5489 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5490 not automatically generating a new one, as it is supposed to. This
5491 prevented TLS from working. If the file did exist, but contained invalid
5492 data, a new version was generated, as expected. It was only the case of a
5493 non-existent file that was broken.
5495 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5496 with a change in libdomainkeys > 0.64.
5498 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5499 from DNS. If the selector record carries the flag, it now has
5500 precedence over the domain-wide flag.
5502 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5504 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5505 the use of a port name that isn't defined in /etc/services) occurred, the
5506 message was deferred as in a normal delivery, and thus remained on the
5507 spool, instead of being failed because of the mua_wrapper setting. This
5508 is now fixed, and I tidied up some of the mua_wrapper messages at the
5511 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5512 the messages in a hash of arrays rather than using individual hashes.
5513 This is a bit cleaner and results in dramatic memory savings, albeit
5514 at a slight CPU cost.
5516 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5517 as requested by Marc Sherman.
5519 SC/10 Eximstats: added histograms for user specified patterns as requested
5522 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5524 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5525 fopen() in the content-scanning modules that did not already have it.
5531 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5532 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5534 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5536 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5538 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5539 to test Sieve filters that use "vacation".
5541 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5542 that changes the way the GnuTLS parameters are stored in the cache file.
5543 The new format can be generated externally. For backward compatibility,
5544 if the data in the cache doesn't make sense, Exim assumes it has read an
5545 old-format file, and it generates new data and writes a new file. This
5546 means that you can't go back to an older release without removing the
5549 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5550 work if there are any delivery delays because "one_time" forces the
5551 parent to be marked "delivered", so its unseen clone is never tried
5552 again. For this reason, Exim now forbids the simultaneous setting of
5555 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5556 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5557 entry if you want to know the details. The fix, however, neglected to
5558 consider the case when local delivery batching is involved. The test for
5559 "previously delivered" was not happening when checking to see if an
5560 address could be batched with a previous (undelivered) one; under
5561 certain circumstances this could lead to multiple deliveries to the same
5564 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5565 in its include files, and this causes problems building Exim.
5567 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5568 header_syntax) but Exim was just ignoring anything given after a slash.
5569 In particular, this caused confusion with an attempt to use "verify =
5570 reverse_host_lookup/defer_ok". An error is now given when options are
5571 supplied for verify items that do not have them. (Maybe reverse_host_
5572 lookup should have a defer_ok option, but that's a different point.)
5574 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5575 defined by RFC 821) to 2048, because there were problems with some AUTH
5576 commands, and RFC 1869 says the size should be increased for extended
5577 SMTP commands that take arguments.
5579 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5582 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5583 "unknown" error; now it says that the functionality isn't in the binary.
5585 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5586 an address' error message when a string expansion fails (syntax or
5587 whatever). Otherwise the password may appear in the log. Following change
5588 PH/42 below, there is no longer a chance of it appearing in a bounce
5591 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5593 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5594 of its addresses was ever tried. (Bugzilla bug #2.)
5596 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5597 the result incorrectly in the debug output. (It correctly added a newline
5598 to what was transported.)
5600 TF/01 Added $received_time.
5602 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5603 commented out examples of how to interface to a virus scanner and to
5604 SpamAssassin. Also added commented examples of av_scanner and
5605 spamd_address settings.
5607 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5608 and controls are allowed in which ACLs. There were a couple of minor
5609 errors. Some of the entries in the conditions table (which is a table of
5610 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5611 negation of where the condition IS allowed.
5613 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5615 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5616 header file does not have a version number, so I've had to invent a new
5617 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5618 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5619 radiusclient), but it was contributed by a Radius user.
5621 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5622 files or directories, for OpenSSL.
5624 PH/19 When an Exim process that is running as root has to create an Exim log
5625 file, it does so in a subprocess that runs as exim:exim so as to get the
5626 ownership right at creation (otherwise, other Exim processes might see
5627 the file with the wrong ownership). There was no test for failure of this
5628 fork() call, which would lead to the process getting stuck as it waited
5629 for a non-existent subprocess. Forks do occasionally fail when resources
5630 run out. I reviewed all the other calls to fork(); they all seem to check
5633 PH/20 When checking for unexpected SMTP input at connect time (before writing
5634 the banner), Exim was not dealing correctly with a non-positive return
5635 from the read() function. If the client had disconnected by this time,
5636 the result was a log entry for a synchronization error with an empty
5637 string after "input=" when read() returned zero. If read() returned -1
5638 (an event I could not check), uninitialized data bytes were printed.
5639 There were reports of junk text (parts of files, etc) appearing after
5642 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5644 PH/22 Added support for macro redefinition, and (re)definition in between
5645 driver and ACL definitions.
5647 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5648 forgetting to use the resulting value; it was using the unexpanded value.
5650 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5651 hadn't been configured. The fix is from Juergen Kreileder, who
5652 understands it better than I do:
5654 "Here's what I see happening with three configured cyrus_sasl
5655 authenticators configured (plain, login, cram-md5):
5657 On startup auth_cyrus_sasl_init() gets called for each of these.
5658 This means three calls to sasl_listmech() without a specified mech_list.
5659 => SASL tests which mechs of all available mechs actually work
5660 => three warnings about OTP not working
5661 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5663 With the patch, sasl_listmech() also gets called three times. But now
5664 SASL's mech_list option is set to the server_mech specified in the the
5665 authenticator. Or in other words, the answer from sasl_listmech()
5666 gets limited to just the mech you're testing for (which is different
5668 => the return list contains just 'plain' or 'login', 'cram-md5' or
5669 nothing depending on the value of ob->server_mech.
5671 I've just tested the patch: Authentication still works fine,
5672 unavailable mechs specified in the exim configuration are still
5673 caught, and the auth.log warnings about OTP are gone."
5675 PH/25 When debugging is enabled, the contents of the command line are added
5676 to the debugging output, even when log_selector=+arguments is not
5679 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5680 answer is "GNU", and only if the return is "GNU/something" is the answer
5683 PH/27 $acl_verify_message is now set immediately after the failure of a
5684 verification in an ACL, and so is available in subsequent modifiers. In
5685 particular, the message can be preserved by coding like this:
5687 warn !verify = sender
5688 set acl_m0 = $acl_verify_message
5690 Previously, $acl_verify_message was set only while expanding "message"
5691 and "log_message" when a very denied access.
5693 PH/28 Modified OS/os.c-Linux with
5695 -#ifndef OS_LOAD_AVERAGE
5696 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5698 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5699 nomenclature these days.)
5701 PH/29 Installed patch from the Sieve maintainer that adds the options
5702 sieve_useraddress and sieve_subaddress to the redirect router.
5704 PH/30 In these circumstances:
5705 . Two addresses routed to the same list of hosts;
5706 . First host does not offer TLS;
5707 . First host accepts first address;
5708 . First host gives temporary error to second address;
5709 . Second host offers TLS and a TLS session is established;
5710 . Second host accepts second address.
5711 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5712 and peerdn, if requested) that were in fact used only for the second
5715 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5716 attention to any local part prefix or suffix that was matched by the
5717 router that accepted the address. It now behaves in the same way as it
5718 does for delivery: the affixes are removed from the local part unless
5719 rcpt_include_affixes is set on the transport.
5721 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5722 timeout during the DATA phase of an incoming message.
5724 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5725 applied a patch sanctioned by the Sieve maintainer.
5727 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5728 the uid or gid is negative. A case of a negative gid caused this to be
5729 noticed. The fix allows for either to be negative.
5731 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5732 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5735 PH/36 The change PH/12 above was broken. Fixed it.
5737 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5738 the grounds that routing the same address twice would always produce the
5739 same answer. This might have been true once, but it is certainly no
5740 longer true now. Routing a child address may depend on the previous
5741 routing that produced that child. Some complicated redirection strategies
5742 went wrong when messages had multiple recipients, and made Exim's
5743 behaviour dependent on the order in which the addresses were given.
5745 I have moved the duplicate checking until after the routing is complete.
5746 Exim scans the addresses that are assigned to local and remote
5747 transports, and removes any duplicates. This means that more work will be
5748 done, as duplicates will always all be routed, but duplicates are
5749 presumably rare, so I don't expect this is of any significance.
5751 For deliveries to pipes, files, and autoreplies, the duplicate checking
5752 still happens during the routing process, since they are not going to be
5755 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5756 It corrects a timeout issue with spamd. This is Ian's comment: "The
5757 background is that sometimes spamd either never reads data from a
5758 connection it has accepted, or it never writes response data. The exiscan
5759 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5760 blindly assumes that writes won't block so it may never time out."
5762 PH/39 Allow G after quota size as well as K and M.
5764 PH/40 The value set for $authenticated_id in an authenticator may not contain
5765 binary zeroes or newlines because the value is written to log lines and
5766 to spool files. There was no check on this. Now the value is run through
5767 the string_printing() function so that such characters are converted to
5768 printable escape sequences.
5770 PH/41 $message_linecount is a new variable that contains the total number of
5771 lines in the message. Compare $body_linecount, which is the count for the
5774 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5775 bounce and delay warning messages, except in certain special cases, which
5778 (a) An SMTP error message from a remote host;
5779 (b) A message specified in a :fail: redirection;
5780 (c) A message specified in a "fail" command in a system filter;
5781 (d) A message specified in a FAIL return from the queryprogram router;
5782 (e) A message specified by the cannot_route_message router option.
5784 In these cases only, Exim does include the error details in bounce and
5785 warning messages. There are also a few cases where bland messages such
5786 as "unrouteable address" or "local delivery error" are given.
5788 PH/43 $value is now also set for the "else" part of a ${run expansion.
5790 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5791 being worked on, but at least Exim now implements the latest version to
5794 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5795 process to complete was treated as a delivery failure, a timeout while
5796 writing the message to the pipe was logged, but erroneously treated as a
5797 successful delivery. Such timeouts include transport filter timeouts. For
5798 consistency with the overall process timeout, these timeouts are now
5799 treated as errors, giving rise to delivery failures by default. However,
5800 there is now a new Boolean option for the pipe transport called
5801 timeout_defer, which, if set TRUE, converts the failures into defers for
5802 both kinds of timeout. A transport filter timeout is now identified in
5805 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5806 systems where "make" and "gmake" are different, calling "gmake" at top
5807 level broke things. I've arranged for the value of $(MAKE) to be passed
5808 from the Makefile to this script so that it can call the same version of
5812 A note about Exim versions 4.44 and 4.50
5813 ----------------------------------------
5815 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5816 changes of various kinds. As a consequence, a big documentation update was
5817 needed. This delayed the release for rather longer than seemed good, especially
5818 in the light of a couple of (minor) security issues. Therefore, the changes
5819 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5820 release. So 4.44 and 4.50 are in effect two different branches that both start
5823 I have left the 4.50 change log unchanged; it contains all the changes since
5824 4.43. The change log for 4.44 is below; many of its items are identical to
5825 those for 4.50. This seems to be the most sensible way to preserve the
5826 historical information.
5832 1. Minor wording change to the doc/README.SIEVE file.
5834 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5835 computation of the current number of files was incorrect.
5837 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5838 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5841 4. Give more explanation in the error message when the command for a transport
5842 filter fails to execute.
5844 5. There are several places where Exim runs a non-Exim command in a
5845 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5846 was being done only for the command run by the queryprogram router. It is
5847 now done for all such subprocesses. The other cases are: ${run, transport
5848 filters, and the commands run by the lmtp and pipe transports.
5850 6. Added CONFIGURE_GROUP build-time option.
5852 7. Some older OS have a limit of 256 on the maximum number of file
5853 descriptors. Exim was using setrlimit() to set 1000 as a large value
5854 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5855 systems. I've change it so that if it can't get 1000, it tries for 256.
5857 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5858 was an oversight, and furthermore, ever since the addition of extra
5859 controls (e.g. 4.43/32), the checks on when to allow different forms of
5860 "control" were broken. There should now be diagnostics for all cases when a
5861 control that does not make sense is encountered.
5863 9. Added the /retain_sender option to "control=submission".
5865 10. $recipients is now available in the predata ACL (oversight).
5867 11. Tidy the search cache before the fork to do a delivery from a message
5868 received from the command line. Otherwise the child will trigger a lookup
5869 failure and thereby defer the delivery if it tries to use (for example) a
5870 cached ldap connection that the parent has called unbind on.
5872 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5873 of $address_data from the recipient verification was clobbered by the
5874 sender verification.
5876 13. The value of address_data from a sender verification is now available in
5877 $sender_address_data in subsequent conditions in the ACL statement.
5879 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5881 15. Added a new option "connect=<time>" to callout options, to set a different
5884 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5885 was its contents. (It was OK if the option was not defined at all.)
5887 17. A "Completed" log line is now written for messages that are removed from
5888 the spool by the -Mrm option.
5890 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5891 information about exactly what failed.
5893 19. Added -dd to debug only the daemon process.
5895 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5896 handles timeouts, both on the server side and network timeouts. Renamed the
5897 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5899 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5900 It is now set to "smtps".
5902 22. $host_address is now set to the target address during the checking of
5903 ignore_target_hosts.
5905 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5906 being passed; this would have caused $sender_host_name to have been used if
5907 matching the list had actually called for a host name (not very likely,
5908 since this list is usually IP addresses). A host name is now passed as
5911 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5912 code that specifies a non-restarting handler (typically sigaction() in
5913 modern systems) in an attempt to fix a rare and obscure crash bug.
5915 25. Narrowed the window for a race in the daemon that could cause it to ignore
5916 SIGCHLD signals. This is not a major problem, because they are used only to
5917 wake it up if nothing else does.
5919 26. A malformed maildirsize file could cause Exim to calculate negative values
5920 for the mailbox size or file count. Odd effects could occur as a result.
5921 The maildirsize information is now recalculated if the size or filecount
5924 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5925 support for a long time. Removed HAVE_SYS_VFS_H.
5927 28. Installed the latest version of exipick from John Jetmore.
5929 29. In an address list, if the pattern was not a regular expression, an empty
5930 subject address (from a bounce message) matched only if the pattern was an
5931 empty string. Non-empty patterns were not even tested. This was the wrong
5932 because it is perfectly reasonable to use an empty address as part of a
5933 database query. An empty address is now tested by patterns that are
5934 lookups. However, all the other forms of pattern expect the subject to
5935 contain a local part and a domain, and therefore, for them, an empty
5936 address still always fails if the pattern is not itself empty.
5938 30. Exim went into a mad DNS loop when attempting to do a callout where the
5939 host was specified on an smtp transport, and looking it up yielded more
5940 than one IP address.
5942 31. Re-factored the code for checking spool and log partition space into a
5943 function that finds that data and another that does the check. The former
5944 is then used to implement four new variables: $spool_space, $log_space,
5945 $spool_inodes, and $log_inodes.
5947 32. The RFC2047 encoding function was originally intended for short strings
5948 such as real names; it was not keeping to the 75-character limit for
5949 encoded words that the RFC imposes. It now respects the limit, and
5950 generates multiple encoded words if necessary. To be on the safe side, I
5951 have increased the buffer size for the ${rfc2047: expansion operator from
5954 33. It is now permitted to omit both strings after an "if" condition; if the
5955 condition is true, the result is "true". As before, when the second string
5956 is omitted, a false condition yields an empty string. This makes it less
5957 cumbersome to write custom ACL and router conditions.
5959 34. Failure to deliver a bounce message always caused it to be frozen, even if
5960 there was an errors_to setting on the router. The errors_to setting is now
5963 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5964 canonical form (fully expanded) before being placed in
5965 $sender_host_address.
5967 36. The table in the code that translates DNS record types into text (T_A to
5968 "A" for instance) was missing entries for NS and CNAME. It is just possible
5969 that this could have caused confusion if both these types were looked up
5970 for the same domain, because the text type is used as part of Exim's
5971 per-process caching. But the chance of anyone hitting this buglet seems
5974 37. The dnsdb lookup has been extended in a number of ways.
5976 (1) There is a new type, "zns", which walks up the domain tree until it
5977 finds some nameserver records. It should be used with care.
5979 (2) There is a new type, "mxh", which is like "mx" except that it returns
5980 just the host names, not the priorities.
5982 (3) It is now possible to give a list of domains (or IP addresses) to be
5983 looked up. The behaviour when one of the lookups defers can be
5984 controlled by a keyword.
5986 (4) It is now possible to specify the separator character for use when
5987 multiple records are returned.
5989 38. The dnslists ACL condition has been extended: it is now possible to supply
5990 a list of IP addresses and/or domains to be looked up in a particular DNS
5993 39. Added log_selector=+queue_time_overall.
5995 40. When running the queue in the test harness, wait just a tad after forking a
5996 delivery process, to get repeatability of debugging output.
5998 41. Include certificate and key file names in error message when GnuTLS fails
5999 to set them up, because the GnuTLS error message doesn't include the name
6000 of the failing file when there is a problem reading it.
6002 42. Allow both -bf and -bF in the same test run.
6004 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6006 44. The "Exiscan patch" is now merged into the mainline Exim source.
6008 45. Sometimes the final signoff response after QUIT could fail to get
6009 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6010 < 0 before doing a fflush(). This bug looks as though it goes back to the
6011 introduction of TLS in release 3.20, but "sometimes" must have been rare
6012 because the tests only now provoked it.
6014 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6015 (this can affect the format of dates).
6017 47. exim_tidydb, when checking for the continued existence of a message for
6018 which it has found a message-specific retry record, was not finding
6019 messages that were in split spool directories. Consequently, it was
6020 deleting retry records that should have stayed in existence.
6022 48. Steve fixed some bugs in eximstats.
6024 49. The SPA authentication driver was not abandoning authentication and moving
6025 on to the next authenticator when an expansion was forced to fail,
6026 contradicting the general specification for all authenticators. Instead it
6027 was generating a temporary error. It now behaves as specified.
6029 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6030 (the order specifies the preference for clients). The order is now AES256,
6031 AES128, 3DES, ARCFOUR128.
6033 51. Small patch to Sieve code - explicitly set From: when generating an
6036 52. Exim crashed if a remote delivery caused a very long error message to be
6037 recorded - for instance if somebody sent an entire SpamAssassin report back
6038 as a large number of 550 error lines. This bug was coincidentally fixed by
6039 increasing the size of one of Exim's internal buffers (big_buffer) that
6040 happened as part of the Exiscan merge. However, to be on the safe side, I
6041 have made the code more robust (and fixed the comments that describe what
6044 53. Now that there can be additional text after "Completed" in log lines (if
6045 the queue_time_overall log selector is set), a one-byte patch to exigrep
6046 was needed to allow it to recognize "Completed" as not the last thing in
6049 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6050 patch that reportedly fixes this has been added. I am not expert enough to
6051 create a test for it. This is what the patch creator wrote:
6053 "I found a little strange behaviour of ldap code when working with
6054 Windows 2003 AD Domain, where users was placed in more than one
6055 Organization Units. When I tried to give exim partial DN, the exit code
6056 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6057 But simultaneously result of request was absolutely normal ldap result,
6058 so I produce this patch..."
6060 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6061 REFERENCE, so I have modified the code to exclude the patch when that macro
6064 55. Some experimental protocols are using DNS PTR records for new purposes. The
6065 keys for these records are domain names, not reversed IP addresses. The
6066 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6067 leaves it alone. Component reversal etc. now happens only for IP addresses.
6070 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6072 57. Double the size of the debug message buffer (to 2048) so that more of very
6073 long debug lines gets shown.
6075 58. The exicyclog utility now does better if the number of log files to keep
6076 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6078 59. Two changes related to the smtp_active_hostname option:
6080 (1) $smtp_active_hostname is now available as a variable.
6081 (2) The default for smtp_banner uses $smtp_active_hostname instead
6082 of $primary_hostname.
6084 60. The host_aton() function is supposed to be passed a string that is known
6085 to be a valid IP address. However, in the case of IPv6 addresses, it was
6086 not checking this. This is a hostage to fortune. Exim now panics and dies
6087 if the condition is not met. A case was found where this could be provoked
6088 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6089 components; fortuitously, this particular loophole had already been fixed
6090 by change 4.50/55 above.
6092 If there are any other similar loopholes, the new check in host_aton()
6093 itself should stop them being exploited. The report I received stated that
6094 data on the command line could provoke the exploit when Exim was running as
6095 exim, but did not say which command line option was involved. All I could
6096 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6097 running as the user.
6100 61. There was a buffer overflow vulnerability in the SPA authentication code
6101 (which came originally from the Samba project). I have added a test to the
6102 spa_base64_to_bits() function which I hope fixes it.
6105 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6106 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6108 63. The daemon start-up calls getloadavg() while still root for those OS that
6109 need the first call to be done as root, but it missed one case: when
6110 deliver_queue_load_max is set with deliver_drop_privilege. This is
6111 necessary for the benefit of the queue runner, because there is no re-exec
6112 when deliver_drop_privilege is set.
6114 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6115 This has been fixed.
6117 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6118 was in use, was not putting the data itself into the right store pool;
6119 consequently, it could be overwritten for a subsequent message in the same
6120 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6123 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6125 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6126 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6127 the function were treating the return as a boolean value, which happened to
6128 work because 0=false and not-0=true, but is not correct code.
6130 68. The host_aton() function was not handling scoped IPv6 addresses (those
6131 with, for example, "%eth0" on the end) correctly.
6133 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6134 negated items (that is, ~something) in unsigned ints. Some compilers
6135 apparently mutter when there is no cast.
6137 70. If an address verification called from an ACL failed, and did not produce a
6138 user-specific message (i.e. there was only a "system" message), nothing was
6139 put in $acl_verify_message. In this situation, it now puts the system
6142 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6143 session; change 4.31/43 added the unwanted input to the log line - except
6144 that it did not do this in the start of session case. It now does.
6146 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6147 This is wrong and can cause the other end to generate a synchronization
6148 error if it is another Exim or anything else that does the synchronization
6149 check. A QUIT command is no longer sent after a timeout.
6151 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6152 during host lookups.
6154 74. The defer_ok option of callout verification was not working if it was used
6155 when verifying addresses in header lines, that is, for this case:
6157 verify = header_sender/callout=defer_ok
6159 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6160 those file descriptors could be used for SMTP connections. If anything
6161 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6162 could be sent to the SMTP client, causing chaos. The daemon now opens
6163 stdin, stdout, and stderr to /dev/null when it puts itself into the
6166 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6167 log by default. The user can override this with suitable Perl magic.
6169 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6170 the log message when discard triggers, was not working for the DATA ACL or
6171 for the non-SMTP ACL.
6173 78. Error message wording change in sieve.c.
6175 79. If smtp_accept_max_per_host was set, the number of connections could be
6176 restricted to fewer than expected, because the daemon was trying to set up
6177 a new connection before checking whether the processes handling previous
6178 connections had finished. The check for completed processes is now done
6179 earlier. On busy systems, this bug wouldn't be noticed because something
6180 else would have woken the daemon, and it would have reaped the completed
6183 80. If a message was submitted locally by a user whose login name contained one
6184 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6185 It caused a spool format error. I have fixed the spool reading code. A
6186 related problem was that the "from" clause in the Received: line became
6187 illegal because of the space(s). It is now covered by ${quote_local_part.
6189 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6192 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6194 83. Patch from Sieve maintainer to fix unterminated string problem in
6195 "vacation" handling.
6197 84. Some minor changes to the Linux configuration files to help with other
6198 OS variants using glibc.
6200 85. One more patch for Sieve to update vacation handling to latest spec.
6203 ----------------------------------------------------
6204 See the note above about the 4.44 and 4.50 releases.
6205 ----------------------------------------------------
6211 1. Change 4.43/35 introduced a bug that caused file counts to be
6212 incorrectly computed when quota_filecount was set in an appendfile
6215 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6216 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6219 3. Give more explanation in the error message when the command for a transport
6220 filter fails to execute.
6222 4. There are several places where Exim runs a non-Exim command in a
6223 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6224 was being done only for the command run by the queryprogram router. It is
6225 now done for all such subprocesses. The other cases are: ${run, transport
6226 filters, and the commands run by the lmtp and pipe transports.
6228 5. Some older OS have a limit of 256 on the maximum number of file
6229 descriptors. Exim was using setrlimit() to set 1000 as a large value
6230 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6231 systems. I've change it so that if it can't get 1000, it tries for 256.
6233 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6234 was an oversight, and furthermore, ever since the addition of extra
6235 controls (e.g. 4.43/32), the checks on when to allow different forms of
6236 "control" were broken. There should now be diagnostics for all cases when a
6237 control that does not make sense is encountered.
6239 7. $recipients is now available in the predata ACL (oversight).
6241 8. Tidy the search cache before the fork to do a delivery from a message
6242 received from the command line. Otherwise the child will trigger a lookup
6243 failure and thereby defer the delivery if it tries to use (for example) a
6244 cached ldap connection that the parent has called unbind on.
6246 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6247 of $address_data from the recipient verification was clobbered by the
6248 sender verification.
6250 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6251 was its contents. (It was OK if the option was not defined at all.)
6253 11. A "Completed" log line is now written for messages that are removed from
6254 the spool by the -Mrm option.
6256 12. $host_address is now set to the target address during the checking of
6257 ignore_target_hosts.
6259 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6260 being passed; this would have caused $sender_host_name to have been used if
6261 matching the list had actually called for a host name (not very likely,
6262 since this list is usually IP addresses). A host name is now passed as
6265 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6266 code that specifies a non-restarting handler (typically sigaction() in
6267 modern systems) in an attempt to fix a rare and obscure crash bug.
6269 15. Narrowed the window for a race in the daemon that could cause it to ignore
6270 SIGCHLD signals. This is not a major problem, because they are used only to
6271 wake it up if nothing else does.
6273 16. A malformed maildirsize file could cause Exim to calculate negative values
6274 for the mailbox size or file count. Odd effects could occur as a result.
6275 The maildirsize information is now recalculated if the size or filecount
6278 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6279 support for a long time. Removed HAVE_SYS_VFS_H.
6281 18. Updated exipick to current release from John Jetmore.
6283 19. Allow an empty sender to be matched against a lookup in an address list.
6284 Previously the only cases considered were a regular expression, or an
6287 20. Exim went into a mad DNS lookup loop when doing a callout where the
6288 host was specified on the transport, if the DNS lookup yielded more than
6291 21. The RFC2047 encoding function was originally intended for short strings
6292 such as real names; it was not keeping to the 75-character limit for
6293 encoded words that the RFC imposes. It now respects the limit, and
6294 generates multiple encoded words if necessary. To be on the safe side, I
6295 have increased the buffer size for the ${rfc2047: expansion operator from
6298 22. Failure to deliver a bounce message always caused it to be frozen, even if
6299 there was an errors_to setting on the router. The errors_to setting is now
6302 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6303 canonical form (fully expanded) before being placed in
6304 $sender_host_address.
6306 24. Updated eximstats to version 1.33
6308 25. Include certificate and key file names in error message when GnuTLS fails
6309 to set them up, because the GnuTLS error message doesn't include the name
6310 of the failing file when there is a problem reading it.
6312 26. Expand error message when OpenSSL has problems setting up cert/key files.
6315 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6316 (this can affect the format of dates).
6318 28. exim_tidydb, when checking for the continued existence of a message for
6319 which it has found a message-specific retry record, was not finding
6320 messages that were in split spool directories. Consequently, it was
6321 deleting retry records that should have stayed in existence.
6323 29. eximstats updated to version 1.35
6324 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6325 1.35 - bugfix such that pie charts by volume are generated correctly
6327 30. The SPA authentication driver was not abandoning authentication and moving
6328 on to the next authenticator when an expansion was forced to fail,
6329 contradicting the general specification for all authenticators. Instead it
6330 was generating a temporary error. It now behaves as specified.
6332 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6333 (the order specifies the preference for clients). The order is now AES256,
6334 AES128, 3DES, ARCFOUR128.
6336 31. Small patch to Sieve code - explicitly set From: when generating an
6339 32. Exim crashed if a remote delivery caused a very long error message to be
6340 recorded - for instance if somebody sent an entire SpamAssassin report back
6341 as a large number of 550 error lines. This bug was coincidentally fixed by
6342 increasing the size of one of Exim's internal buffers (big_buffer) that
6343 happened as part of the Exiscan merge. However, to be on the safe side, I
6344 have made the code more robust (and fixed the comments that describe what
6347 33. Some experimental protocols are using DNS PTR records for new purposes. The
6348 keys for these records are domain names, not reversed IP addresses. The
6349 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6350 leaves it alone. Component reversal etc. now happens only for IP addresses.
6353 34. The host_aton() function is supposed to be passed a string that is known
6354 to be a valid IP address. However, in the case of IPv6 addresses, it was
6355 not checking this. This is a hostage to fortune. Exim now panics and dies
6356 if the condition is not met. A case was found where this could be provoked
6357 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6358 components; fortuitously, this particular loophole had already been fixed
6359 by change 4.50/55 or 4.44/33 above.
6361 If there are any other similar loopholes, the new check in host_aton()
6362 itself should stop them being exploited. The report I received stated that
6363 data on the command line could provoke the exploit when Exim was running as
6364 exim, but did not say which command line option was involved. All I could
6365 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6366 running as the user.
6369 35. There was a buffer overflow vulnerability in the SPA authentication code
6370 (which came originally from the Samba project). I have added a test to the
6371 spa_base64_to_bits() function which I hope fixes it.
6374 36. The daemon start-up calls getloadavg() while still root for those OS that
6375 need the first call to be done as root, but it missed one case: when
6376 deliver_queue_load_max is set with deliver_drop_privilege. This is
6377 necessary for the benefit of the queue runner, because there is no re-exec
6378 when deliver_drop_privilege is set.
6380 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6381 was in use, was not putting the data itself into the right store pool;
6382 consequently, it could be overwritten for a subsequent message in the same
6383 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6386 38. Sometimes the final signoff response after QUIT could fail to get
6387 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6388 < 0 before doing a fflush(). This bug looks as though it goes back to the
6389 introduction of TLS in release 3.20, but "sometimes" must have been rare
6390 because the tests only now provoked it.
6396 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6397 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6398 Now it returns an integer. A number of calls were still expecting a T/F
6399 return. Fortuitously, in all cases, the tests worked in OK situations,
6400 which is the norm. However, things would have gone wrong on any write
6401 failures on the smtp file descriptor. This function is used when sending
6402 messages over SMTP and also when doing verify callouts.
6404 2. When Exim is called to do synchronous delivery of a locally submitted
6405 message (the -odf or -odi options), it no longer closes stderr before doing
6408 3. Implemented the mua_wrapper option.
6410 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6412 5. Implemented the functions header_remove(), header_testname(),
6413 header_add_at_position(), and receive_remove_recipient(), and exported them
6416 6. If an ACL "warn" statement specified the addition of headers, Exim already
6417 inserted X-ACL-Warn: at the start if there was no header name. However, it
6418 was not making this test for the second and subsequent header lines if
6419 there were newlines in the string. This meant that an invalid header could
6420 be inserted if Exim was badly configured.
6422 7. Allow an ACL "warn" statement to add header lines at the start or after all
6423 the Received: headers, as well as at the end.
6425 8. Added the rcpt_4xx retry error code.
6427 9. Added postmaster_mailfrom=xxx to callout verification option.
6429 10. Added mailfrom=xxxx to the callout verification option, for verify=
6432 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6433 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6435 12. Inserted some casts to stop certain compilers warning when using pointer
6436 differences as field lengths or precisions in printf-type calls (mostly
6437 affecting debugging statements).
6439 13. Added optional readline() support for -be (dynamically loaded).
6441 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6442 same clock tick as a message's arrival, so that its received time was the
6443 same as the "first fail" time on the retry record, and that message
6444 remained on the queue past the ultimate address timeout, every queue runner
6445 would try a delivery (because it was past the ultimate address timeout) but
6446 after another failure, the ultimate address timeout, which should have then
6447 bounced the address, did not kick in. This was a "< instead of <=" error;
6448 in most cases the first failure would have been in the next clock tick
6449 after the received time, and all would be well.
6451 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6452 being recognized when the domain list was tested by the match_domain
6453 condition in an expansion string.
6455 16. Added the ${str2b64: operator.
6457 17. Exim was always calling setrlimit() to set a large limit for the number of
6458 processes, without checking whether the existing limit was already
6459 adequate. (It did check for the limit on file descriptors.) Furthermore,
6460 errors from getrlimit() and setrlimit() were being ignored. Now they are
6461 logged to the main and panic logs, but Exim does carry on, to try to do its
6462 job under whatever limits there are.
6464 18. Imported PCRE 5.0.
6466 19. Trivial typo in log message " temporarily refused connection" (the leading
6469 20. If the log selector return_path_on_delivery was set and an address was
6470 redirected to /dev/null, the delivery process crashed because it assumed
6471 that a return path would always be set for a "successful" delivery. In this
6472 case, the whole delivery is bypassed as an optimization, and therefore no
6475 21. Internal re-arrangement: the function for sending a challenge and reading
6476 a response while authentication was assuming a zero-terminated challenge
6477 string. It's now changed to take a pointer and a length, to allow for
6478 binary data in such strings.
6480 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6482 23. Exim was not respecting finduser_retries when seeking the login of the
6483 uid under which it was called; it was always trying 10 times. (The default
6484 setting of finduser_retries is zero.) Also, it was sleeping after the final
6485 failure, which is pointless.
6487 24. Implemented tls_on_connect_ports.
6489 25. Implemented acl_smtp_predata.
6491 26. If the domain in control=submission is set empty, Exim assumes that the
6492 authenticated id is a complete email address when it generates From: or
6493 Sender: header lines.
6495 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6496 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6497 chown and chgrp in /bin and hostname in /usr/bin.
6499 28. Exim was keeping the "process log" file open after each use, just as it
6500 does for the main log. This opens the possibility of it remaining open for
6501 long periods when the USR1 signal hits a daemon. Occasional processlog
6502 errors were reported, that could have been caused by this. Anyway, it seems
6503 much more sensible not to leave this file open at all, so that is what now
6506 29. The long-running daemon process does not normally write to the log once it
6507 has entered its main loop, and it closes the log before doing so. This is
6508 so that log files can straightforwardly be renamed and moved. However,
6509 there are a couple of unusual error situations where the daemon does write
6510 log entries, and I had neglected to close the log afterwards.
6512 30. The text of an SMTP error response that was received during a remote
6513 delivery was being truncated at 512 bytes. This is too short for some of
6514 the long messages that one sometimes sees. I've increased the limit to
6517 31. It is now possible to make retry rules that apply only when a message has a
6518 specific sender, in particular, an empty sender.
6520 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6521 it possible to be selective about when SMTP synchronization is enforced.
6523 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6525 32. Implemented hosts_connection_nolog.
6527 33. Added an ACL for QUIT.
6529 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6532 35. Added mailbox_size and mailbox_filecount to appendfile.
6534 36. Added control = no_multiline_responses to ACLs.
6536 37. There was a bug in the logic of the code that waits for the clock to tick
6537 in the case where the clock went backwards by a substantial amount such
6538 that the microsecond fraction of "now" was more than the microsecond
6539 fraction of "then" (but the whole seconds number was less).
6541 38. Added support for the libradius Radius client library this is found on
6542 FreeBSD (previously only the radiusclient library was supported).
6548 1. When certain lookups returned multiple values in the form name=value, the
6549 quoting of the values was not always being done properly. Specifically:
6550 (a) If the value started with a double quote, but contained no whitespace,
6552 (b) If the value contained whitespace other than a space character (i.e.
6553 tabs or newlines or carriage returns) it was not quoted.
6554 This fix has been applied to the mysql and pgsql lookups by writing a
6555 separate quoting function and calling it from the lookup code. The fix
6556 should probably also be applied to nisplus, ibase and oracle lookups, but
6557 since I cannot test any of those, I have not disturbed their existing code.
6559 2. A hit in the callout cache for a specific address caused a log line with no
6560 reason for rejecting RCPT. Now it says "Previous (cached) callout
6561 verification failure".
6563 3. There was an off-by-one bug in the queryprogram router. An over-long
6564 return line was truncated at 256 instead of 255 characters, thereby
6565 overflowing its buffer with the terminating zero. As well as fixing this, I
6566 have increased the buffer size to 1024 (and made a note to document this).
6568 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6569 when Exim is waiting for an SMTP response from a remote server, Exim
6570 restarts its select() call on the socket, thereby resetting its timeout.
6571 This is not a problem when such interrupts are rare. Somebody set up a cron
6572 job to run exiwhat every 2 minutes, which is less than the normal select()
6573 timeout (5 or 10 minutes). This meant that the select() timeout never
6574 kicked in because it was always reset. I have fixed this by comparing the
6575 time when an interrupt arrives with the time at the start of the first call
6576 to select(). If more time than the timeout has elapsed, the interrupt is
6577 treated as a timeout.
6579 5. Some internal re-factoring in preparation for the addition of Sieve
6580 extensions (by MH). In particular, the "personal" test is moved to a
6581 separate function, and given an option for scanning Cc: and Bcc: (which is
6582 not set for Exim filters).
6584 6. When Exim created an email address using the login of the caller as the
6585 local part (e.g. when creating a From: or Sender: header line), it was not
6586 quoting the local part when it contained special characters such as @.
6588 7. Installed new OpenBSD configuration files.
6590 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6591 try to make them clearer.
6593 9. Callout options, other than the timeout value, were being ignored when
6594 verifying sender addresses in header lines. For example, when using
6596 verify = header_sender/callout=no_cache
6598 the cache was (incorrectly) being used.
6600 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6601 only the Cygwin environment.
6603 11. When return_path_on_delivery was set as a log selector, if different remote
6604 addresses in the same message used different return paths and parallel
6605 remote delivery occurred, the wrong values would sometimes be logged.
6606 (Whenever a remote delivery process finished, the return path value from
6607 the most recently started remote delivery process was logged.)
6609 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6610 lines when AUTH and/or TLS are in use. This is the "received protocol"
6611 field. Exim used to use "asmtp" for authenticated SMTP, without any
6612 indication (in the protocol name) for TLS use. Now it follows the RFC and
6613 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6614 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6615 names appear in log lines as well as in Received: header lines.
6617 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6618 extensions, and comparison tests, and to fix some bugs.
6620 14. Changes to the "personal" filter test:
6622 (1) The test was buggy in that it was just doing the equivalent of
6623 "contains" tests on header lines. For example, if a user's address was
6624 anne@some.where, the "personal" test would incorrectly be true for
6626 To: susanne@some.where
6628 This test is now done by extracting each address from the header in turn,
6629 and checking the entire address. Other tests that are part of "personal"
6630 are now done using regular expressions (for example, to check local parts
6631 of addresses in From: header lines).
6633 (2) The list of non-personal local parts in From: addresses has been
6634 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6635 taken from the Sieve specification recommendations.
6637 (3) If the message contains any header line starting with "List-" it is
6638 treated as non-personal.
6640 (4) The test for "circular" in the Subject: header line has been removed
6641 because it now seems ill-conceived.
6643 15. Minor typos in src/EDITME comments corrected.
6645 16. Installed latest exipick from John Jetmore.
6647 17. If headers_add on a router specified a text string that was too long for
6648 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6649 of string_sprintf() is now avoided.
6651 18. $message_body_size was not set (it was always zero) when running the DATA
6652 ACL and the local_scan() function.
6654 19. For the "mail" command in an Exim filter, no default was being set for
6655 the once_repeat time, causing a random time value to be used if "once" was
6656 specified. (If the value happened to be <= 0, no repeat happened.) The
6657 default is now 0s, meaning "never repeat". The "vacation" command was OK
6658 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6659 (I found it when inspecting the code).
6661 20. There is now an overall timeout for performing a callout verification. It
6662 defaults to 4 times the callout timeout, which applies to individual SMTP
6663 commands during the callout. The overall timeout applies when there is more
6664 than one host that can be tried. The timeout is checked before trying the
6665 next host. This prevents very long delays if there are a large number of
6666 hosts and all are timing out (e.g. when the network connections are timing
6667 out). The value of the overall timeout can be changed by specifying an
6668 additional sub-option for "callout", called "maxwait". For example:
6670 verify = sender/callout=5s,maxwait=20s
6672 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6673 to the end before writing, but this should make it even safer).
6675 22. Exim was forgetting that it had advertised PIPELINING for the second and
6676 subsequent messages on an SMTP connection. It was also not resetting its
6677 memory on STARTTLS and an internal HELO.
6679 23. When Exim logs an SMTP synchronization error within a session, it now
6680 records whether PIPELINING has been advertised or not.
6682 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6683 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6684 rather than long int.
6686 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6688 26. Added the never_mail option to autoreply.
6694 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6695 crash if the getsockname() call failed; this can happen if a connection is
6696 closed very soon after it is established. The problem was simply in the
6697 order in which certain operations were done, causing Exim to try to write
6698 to the SMTP stream before it had set up the file descriptor. The bug has
6699 been fixed by making things happen in the correct order.
6705 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6706 before the connection was closed, thus losing the rejection response.
6708 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6709 some early Solaris releases, but causes trouble in current releases where
6710 socklen_t is defined.
6712 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6715 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6718 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6719 that releasing the top of it at the end releases what was used for sub-
6720 expansions (unless the block got too big). However, discard this block if
6721 the first thing is a variable or header, so that we can use its block when
6722 it is dynamic (useful for very large $message_headers, for example).
6724 6. Lookups now cache *every* query, not just the most recent. A new, separate
6725 store pool is used for this. It can be recovered when all lookup caches are
6726 flushed. Lookups now release memory at the end of their result strings.
6727 This has involved some general refactoring of the lookup sources.
6729 7. Some code has been added to the store_xxx() functions to reduce the amount
6730 of flapping under certain conditions.
6732 8. log_incoming_interface used to affect only the <= reception log lines. Now
6733 it causes the local interface and port to be added to several more SMTP log
6734 lines, for example "SMTP connection from", and rejection lines.
6736 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6738 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6740 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6741 could be overwritten at the end of the current message (or the start of a
6742 new message if it was set in a HELO ACL). The value is now preserved for
6743 the duration of the SMTP connection.
6745 12. If a transport had a headers_rewrite setting, and a matching header line
6746 contained an unqualified address, that address was qualified, even if it
6747 did not match any rewriting rules. The underlying bug was that the values
6748 of the flags that permit the existence of unqualified sender and recipient
6749 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6750 non-local messages, and by -bnq for local messages) were not being
6751 preserved with the message after it was received.
6753 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6754 "next input=" as part of the text comprising the host identity instead of
6755 the correct text. The code was using the same buffer for two different
6756 strings. However, depending on which order the printing function evaluated
6757 its arguments, the bug did not always show up. Under Linux, for example, my
6758 test suite worked just fine.
6760 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6761 doesn't work with some older versions of Perl. It has been changed to "my",
6762 which in any case is probably the better facility to use.
6764 15. A really picky compiler found some instances of statements for creating
6765 error messages that either had too many or two few arguments for the format
6768 16. The size of the buffer for calls to the DNS resolver has been increased
6769 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6770 for addresses that have a lot of PTR records. This alleviates a problem; it
6771 does not fully solve it.
6773 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6774 buffer now truncates the list and logs the incident, which is the same
6775 action as happens when Exim is looking up a host name and its aliases.
6776 Previously in this situation something unpredictable would happen;
6777 sometimes it was "internal error: store_reset failed".
6779 18. If a server dropped the connection unexpectedly when an Exim client was
6780 using GnuTLS and trying to read a response, the client delivery process
6781 crashed while trying to generate an error log message.
6783 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6784 string, for example:
6786 warn message = H1: something\nH2: something
6788 the text was added as a single header line from Exim's point of view
6789 though it ended up OK in the delivered message. However, searching for the
6790 second and subsequent header lines using $h_h2: did not work. This has been
6791 fixed. Similarly, if a system filter added multiple headers in this way,
6792 the routers could not see them.
6794 20. Expanded the error message when iplsearch is called with an invalid key to
6795 suggest using net-iplsearch in a host list.
6797 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6798 ACLs are no longer actually imposed (and a message to that effect is
6801 22. If a "gecos" field in a passwd entry contained escaped characters, in
6802 particular, if it contained a \" sequence, Exim got it wrong when building
6803 a From: or a Sender: header from that name. A second bug also caused
6804 incorrect handling when an unquoted " was present following a character
6805 that needed quoting.
6807 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6808 was not being matched caselessly.
6810 24. Arranged for all hyphens in the exim.8 source to be escaped with
6813 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6814 properly. Recipient callout cache records were still being keyed to include
6815 the sender, even when use_sender was set false. This led to far more
6816 callouts that were necessary. The sender is no longer included in the key
6817 when use_sender is false.
6819 26. Added "control = submission" modifier to ACLs.
6821 27. Added the ${base62d: operator to decode base 62 numbers.
6823 28. dnsdb lookups can now access SRV records.
6825 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6826 the configuration file.
6828 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6829 (-v) mode. This makes the output for a verbose queue run more intelligible.
6831 31. Added a use_postmaster feature to recipient callouts.
6833 32. Added the $body_zerocount variable, containing the number of binary zero
6834 bytes in the message body.
6836 33. The time of last modification of the "new" subdirectory is now used as the
6837 "mailbox time last read" when there is a quota error for a maildir
6840 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6842 35. Added +ignore_unknown as a special item in host lists.
6844 36. Code for decoding IPv6 addresses in host lists is now included, even if
6845 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6846 address was recognized as an IP address, but was then not correctly decoded
6847 into binary, causing unexpected and incorrect effects when compared with
6854 1. Very minor rewording of debugging text in manualroute to say "list of
6855 hosts" instead of "hostlist".
6857 2. If verify=header_syntax was set, and a header line with an unqualified
6858 address (no domain) and a large number of spaces between the end of the
6859 name and the colon was received, the reception process suffered a buffer
6860 overflow, and (when I tested it) crashed. This was caused by some obsolete
6861 code that should have been removed. The fix is to remove it!
6863 3. When running in the test harness, delay a bit after writing a bounce
6864 message to get a bit more predictability in the log output.
6866 4. Added a call to search_tidyup() just before forking a reception process. In
6867 theory, someone could use a lookup in the expansion of smtp_accept_max_
6868 per_host which, without the tidyup, could leave open a database connection.
6870 5. Added the variables $recipient_data and $sender_data which get set from a
6871 lookup success in an ACL "recipients" or "senders" condition, or a router
6872 "senders" option, similar to $domain_data and $local_part_data.
6874 6. Moved the writing of debug_print from before to after the "senders" test
6877 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6878 problems for message scanning, either using a data ACL, or using
6879 local_scan() because the Received: header was not generated till after they
6880 were called (in order to set the time as the time of reception completion).
6881 I have revised the way this works. The header is now generated after the
6882 body is received, but before the ACL or local_scan() are called. After they
6883 are run, the timestamp in the header is updated.
6889 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6890 before starting a queue runner without re-exec. This happened only when
6891 deliver_drop_privilege was set or when the Exim user was set to root. The
6892 effect of the bug was that timeouts during subsequent deliveries caused
6893 crashes instead of being properly handled. The handler is now left at its
6894 default (and expected) setting.
6896 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6897 message, again when deliver_drop_privilege is set or Exim is run as root.
6898 The bug described in (1) was not present in this case, but the tidying up
6899 of the other signals was missing. I have made the two cases consistent.
6901 3. The ignore_target_hosts setting on a manualroute router was being ignored
6902 for hosts that were looked up using the /MX notation.
6904 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6907 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6908 operated on the sender address. After changing the $sender_address to <>
6909 for the sender address verify, Exim was re-instated it as the original
6910 (before rewriting) address, but remembering that it had rewritten it, so it
6911 wasn't rewriting it again. This bug also had the effect of breaking the
6912 sender address verification caching when the sender address was rewritten.
6914 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6915 This has been changed so that if the ip literal address matches
6916 ignore_target_hosts, the router declines.
6918 7. Added expansion conditions match_domain, match_address, and match_local_
6919 part (NOT match_host).
6921 8. The placeholder for the Received: header didn't have a length field set.
6923 9. Added code to Exim itself and to exim_lock to test for a specific race
6924 condition that could lead to file corruption when using MBX delivery. The
6925 issue is with the lockfile that is created in /tmp. If this file is removed
6926 after a process has opened it but before that process has acquired a lock,
6927 there is the potential for a second process to recreate the file and also
6928 acquire a lock. This could lead to two Exim processes writing to the file
6929 at the same time. The added code performs the same test as UW imapd; it
6930 checks after acquiring the lock that its file descriptor still refers to
6931 the same named file.
6933 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6934 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6937 11. Added the smtp_active_hostname option. If used, this will typically be made
6938 to depend on the incoming interface address. Because $interface_address is
6939 not set up until the daemon has forked a reception process, error responses
6940 that can happen earlier (such as "too many connections") no longer contain
6943 12. If an expansion in a condition on a "warn" statement fails because a lookup
6944 defers, the "warn" statement is abandoned, and the next ACL statement is
6945 processed. Previously this caused the whole ACL to be aborted.
6947 13. Added the iplsearch lookup type.
6949 14. Added ident_timeout as a log selector.
6951 15. Added tls_certificate_verified as a log selector.
6953 16. Added a global option tls_require_ciphers (compare the smtp transport
6954 option of the same name). This controls incoming TLS connections.
6956 17. I finally figured out how to make tls_require_ciphers do a similar thing
6957 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6958 before starting the TLS session.
6960 18. Tabs are now shown as \t in -bP output.
6962 19. If the log selector return_path_on_delivery was set, Exim crashed when
6963 bouncing a message because it had too many Received: header lines.
6965 20. If two routers both had headers_remove settings, and the first one included
6966 a superfluous trailing colon, the final name in the first list and the
6967 first name in the second list were incorrectly joined into one item (with a
6968 colon in the middle).
6974 1. Added -C and -D options to the exinext utility, mainly to make it easier
6975 to include in the automated testing, but these could be helpful when
6976 multiple configurations are in use.
6978 2. The exinext utility was not formatting the output nicely when there was
6979 an alternate port involved in the retry record key, nor when there was a
6980 message id as well (for retries that were specific to a specific message
6981 and a specific host). It was also confused by IPv6 addresses, because of
6982 the additional colons they contain. I have fixed the IPv4 problem, and
6983 patched it up to do a reasonable job for IPv6.
6985 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
6986 delivery, the log line now contains "pipelined" if PIPELINING was used.
6988 4. An SMTP transport process used to panic and die if the bind() call to set
6989 an explicit outgoing interface failed. This has been changed; it is now
6990 treated in the same way as a connect() failure.
6992 5. A reference to $sender_host_name in the part of a conditional expansion
6993 that was being skipped was still causing a DNS lookup. This no longer
6996 6. The def: expansion condition was not recognizing references to header lines
6997 that used bh_ and bheader_.
6999 7. Added the _cache feature to named lists.
7001 8. The code for checking quota_filecount in the appendfile transport was
7002 allowing one more file than it should have been.
7004 9. For compatibility with Sendmail, the command line option
7012 and sets the incoming protocol and host name (for trusted callers). The
7013 host name and its colon can be omitted when only the protocol is to be set.
7014 Note the Exim already has two private options, -pd and -ps, that refer to
7015 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7016 "s", but I don't think that's a major issue.
7018 10. A number of refactoring changes to the code, none of which should affect
7021 (a) The number of logging options was getting close to filling up the
7022 32-bit word that was used as a bit map. I have split them into two classes:
7023 those that are passed in the argument to log_write(), and those that are
7024 only ever tested independently outside of that function. These are now in
7025 separate 32-bit words, so there is plenty of room for expansion again.
7026 There is no change in the user interface or the logging behaviour.
7028 (b) When building, for example, log lines, the code previously used a
7029 macro that called string_cat() twice, in order to add two strings. This is
7030 not really sufficiently general. Furthermore, there was one instance where
7031 it was actually wrong because one of the argument was used twice, and in
7032 one call a function was used. (As it happened, calling the function twice
7033 did not affect the overall behaviour.) The macro has been replaced by a
7034 function that can join an arbitrary number of extra strings onto a growing
7037 (c) The code for expansion conditions now uses a table and a binary chop
7038 instead of a serial search (which was left over from when there were very
7039 few conditions). Also, it now recognizes conditions like "pam" even when
7040 the relevant support is not compiled in: a suitably worded error message is
7041 given if an attempt is made to use such a condition.
7043 11. Added ${time_interval:xxxxx}.
7045 12. A bug was causing one of the ddress fields not to be passed back correctly
7046 from remote delivery subprocesses. The field in question was not being
7047 subsequently used, so this caused to problems in practice.
7049 13. Added new log selectors queue_time and deliver_time.
7051 14. Might have fixed a bug in maildirsizefile handling that threw up
7052 "unexpected character" debug warnings, and recalculated the data
7053 unnecessarily. In any case, I expanded the warning message to give more
7056 15. Added the message "Restricted characters in address" to the statements in
7057 the default ACL that block characters like @ and % in local parts.
7059 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7060 Three changes have been made:
7062 (a) There was a serious bug; a negative response to MAIL caused the whole
7063 recipient domain to be cached as invalid, thereby blocking all messages
7064 to all local parts at the same domain, from all senders. This bug has
7065 been fixed. The domain is no longer cached after a negative response to
7066 MAIL if the sender used is not empty.
7068 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7071 (c) A new callout option, "use_sender" has been added for people who want
7072 the modified behaviour.
7078 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7081 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7082 indeed breaks things for older releases.
7084 3. Added additional logging to the case where there is a problem reading data
7085 from a filter that is running in a subprocess using a pipe, in order to
7086 try to track down a specific problem.
7088 4. Testing facility fudge: when running in the test harness and attempting
7089 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7090 getting "No route to host". Convert this to a timeout.
7092 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7095 6. Some OS don't have socklen_t but use size_t instead. This affects the
7096 fifth argument of getsockopt() amongst other things. This is now
7097 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7098 can be set for individual OS. I have set it for SunOS5, OSF1, and
7099 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7100 some earlier ones do not.
7102 7. Change 4.30/15 was not doing the test caselessly.
7104 8. The standard form for an IPv6 address literal was being rejected by address
7105 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7106 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7107 this, as well as the form without the "IPv6" on the front (but only when
7108 address literals are enabled, of course).
7110 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7112 10. Exim crashed if a message with an empty sender address specified by -f
7113 encountered a router with an errors_to setting. This could be provoked only
7114 by a command such as
7118 where an empty string was supplied; "<>" did not hit this bug.
7120 11. Installed PCRE release 4.5.
7122 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7123 remained set. It is now erased.
7125 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7126 times from message ids (which are base 36 rather than the normal 62).
7128 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7129 were being counted as actual protocol errors, and logged if the log
7130 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7131 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7132 and DATA following a set of rejected RCPTs do not count as protocol errors.
7133 In other words, Exim assumes they were pipelined, though this may not
7134 actually be the case. Of course, in all cases the client gets an
7135 appropriate error code.
7137 15. If a lookup fails in an ACL condition, a message about the failure may
7138 be available; it is used if testing the ACL cannot continue, because most
7139 such messages specify what the cause of the deferral is. However, some
7140 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7141 that caused an old message to be retained and used if a later statement
7142 caused a defer, replacing the real cause of the deferral.
7144 16. If an IP address had so many PTR records that the DNS lookup buffer
7145 was not large enough to hold them, Exim could crash while trying to process
7146 the truncated data. It now detects and logs this case.
7148 17. Further to 4.21/58, another change has been made: if (and only if) the
7149 first line of a message (the first header line) ends with CRLF, a bare LF
7150 in a subsequent header line has a space inserted after it, so as not to
7151 terminate the header.
7153 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7154 unnecessarily, used atoi() instead of strtol(), and didn't check the
7155 termination when getting file sizes from file names by regex.
7157 19. Completely re-implemented the support for maildirsize files, in the light
7158 of a number of problems with the previous contributed implementation
7159 (4.30/29). In particular:
7161 . If the quota is zero, the maildirsize file is maintained, but no quota is
7164 . If the maildir directory does not exist, it is created before any attempt
7165 to write a maildirsize file.
7167 . The quota value in the file is just a cache; if the quota is changed in
7168 the transport, the new value overrides.
7170 . A regular expression is available for excluding directories from the
7173 20. The autoreply transport checks the characters in options that define the
7174 message's headers; it allows continued headers, but it was checking with
7175 isspace() after an embedded newline instead of explicitly looking for a
7178 21. If all the "regular" hosts to which an address was routed had passed their
7179 expiry times, and had not reached their retry times, the address was
7180 bounced, even if fallback hosts were defined. Now Exim should go on to try
7183 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7184 equivalent code in the SMTP transport. Some hosts send humungous responses
7185 to HELO/EHLO, more than 1024 it seems.
7187 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7188 gives compiler warnings in some environments. I've now done it "properly",
7191 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7192 (because of problems with the built-in one) was declared to return uschar *
7193 instead of char *, causing compiler failure.
7195 25. Fixed a file descriptor leak when processing alias/forward files.
7197 26. Fixed a minor format string issue in dbfn.c.
7199 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7201 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7202 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7205 29. When a sender address is verified, it is cached, to save repeating the test
7206 when there is more than one recipient in a message. However, when the
7207 verification involves a callout, it is possible for different callout
7208 options to be set for different recipients. It is too complicated to keep
7209 track of this in the cache, so now Exim always runs a verification when a
7210 callout is required, relying on the callout cache for the optimization.
7211 The overhead is duplication of the address routing, but this should not be
7214 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7215 to be verified with callout=postmaster, and the main callout worked but the
7216 postmaster check failed, the verification correctly failed. However, if a
7217 subsequent RCPT command asked for sender verification *without* the
7218 postmaster check, incorrect caching caused this verification also to fail,
7221 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7222 it was not caching the DNS options (qualify_single, search_parents) that
7223 were used when the lookup failed. A subsequent lookup with different
7224 options therefore always gave the same answer, though there were cases
7225 where it should not have. (Example: a "domains = !$mx_any" option on a
7226 dnslookup router: the "domains" option is always processed without any
7227 widening, but the router might have qualify_single set.) Now Exim uses the
7228 cached value only when the same options are set.
7230 32. Added John Jetmore's "exipick" utility to the distribution.
7232 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7233 than a timeout (e.g. a certificate is required, and is not provided), an
7234 Exim server now closes the connection immediately. Previously it waited for
7235 the client to close - but if the client is SSL, it seems that they each
7236 wait for each other, leading to a delay before one of them times out.
7238 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7239 maintained 0.8.x compatibility because I don't think many are using it, and
7240 it is clearly obsolete.
7242 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7245 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7246 client certificate was expired. A simple patch fixes this, though I don't
7247 understand the full logic of why the verify callback is called multiple
7250 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7251 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7252 which causes problems with some clients (such as the Certicom SSL Plus
7253 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7254 disables the coutermeasure allowing Eudora to connect."
7256 38. Exim was not checking that a write() to a log file succeeded. This could
7257 lead to Bad Things if a log got too big, in particular if it hit a file
7258 size limit. Exim now panics and dies if it cannot write to a log file, just
7259 as it does if it cannot open a log file.
7261 39. Modified OS/Makefile-Linux so that it now contains
7263 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7265 The two -D definitions ensure that Exim is compiled with large file
7266 support, which makes it possible to handle log files that are bigger than
7269 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7270 instance) a domain was checked against a named list that involved a lookup,
7271 causing $domain_data to be set, then another domain was checked against the
7272 same list, then the first domain was re-checked, the value of $domain_data
7273 after the final check could be wrong. In particular, if the second check
7274 failed, it could be set empty. This bug probably also applied to
7277 41. The strip_trailing_dot option was not being applied to the address given
7278 with the -f command-line option.
7280 42. The code for reading a message's header from the spool was incrementing
7281 $received_count, but never initializing it. This meant that the value was
7282 incorrect (doubled) while delivering a message in the same process in which
7283 it was received. In the most common configuration of Exim, this never
7284 happens - a fresh exec is done - but it can happen when
7285 deliver_drop_privilege is set.
7287 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7288 it now includes up to 150 characters of the unexpected data in the log
7291 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7292 and building data strings. The size of both of these buffers was 10 000
7293 bytes - far larger than anybody would *ever* want, thought I. Needless to
7294 say, somebody hit the limit. I have increased the maximum line length to
7295 20 000 and the maximum data length of concatenated lines to 100 000. I have
7296 also fixed two bugs, because there was no checking on these buffers. Tsk,
7297 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7298 buffer is too small.
7300 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7301 lsearch lookups. Now it does.
7303 46. When parsing a route_list item in a manualroute router, a fixed-length
7304 buffer was used for the list of hosts. I made this 1024 bytes long,
7305 thinking that nobody would ever have a list of hosts that long. Wrong.
7306 Somebody had a whole pile of complicated expansion conditions, and the
7307 string was silently truncated, leading to an expansion error. It turns out
7308 that it is easier to change to an unlimited length (owing to other changes
7309 that have happened since this code was originally written) than to build
7310 structure for giving a limitation error. The length of the item that
7311 expands into the list of hosts is now unlimited.
7313 47. The lsearch lookup could not handle data where the length of text line was
7314 more than 4095 characters. Such lines were truncated, leading to shortened
7315 data being returned. It should now handle lines of any length.
7317 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7318 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7321 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7322 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7323 no longer happens in comment lines. A list of replacements is now placed
7324 at the head of all of the source files, except those whose only change is
7325 to replace PERL_COMMAND in the very first #! line.
7327 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7328 messages on the queue, with a bottom-up merge sort, using code contributed
7329 by Michael Haardt. This should make operations like -bp somewhat faster on
7330 large queues. It won't affect queue runners, except when queue_run_in_order
7333 51. Installed eximstats 1.31 in the distribution.
7335 52. Added support for SRV lookups to the dnslookup router.
7337 53. If an ACL referred to $message_body or $message_body_end, the value was not
7338 reset for any messages that followed in the same SMTP session.
7340 54. The store-handling optimization for building very long strings was not
7341 differentiating between the different store pools. I don't think this
7342 actually made any difference in practice, but I've tidied it.
7344 55. While running the routers to verify a sender address, $sender_address
7345 was still set to the sender address. This is wrong, because when routing to
7346 send a bounce to the sender, it would be empty. Therefore, I have changed
7347 it so that, while verifying a sender address, $sender_address is set to <>.
7348 (There is no change to what happens when verifying a recipient address.)
7350 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7351 target A or AAAA records (if not already returned) without resetting the
7352 qualify_single or search_parents options of the DNS resolver. These are
7353 inappropriate in this case because the targets of MX and SRV records must
7354 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7355 target that, when qualified, matched something in the local domain. These
7356 two options are now turned off when doing these lookups.
7358 57. It seems that at least some releases of Reiserfs (which does not have the
7359 concept of a fixed number of inodes) returns zero and not -1 for the
7360 number of available inodes. This interacted badly with check_spool_inodes,
7361 which assumed that -1 was the "no such thing" setting. What I have done is
7362 to check that the total number of inodes is greater than zero before doing
7363 the test of how many are available.
7365 58. When a "warn" ACL statement has a log_message modifier, the message is
7366 remembered, and not repeated. This is to avoid a lot of repetition when a
7367 message has many recipients that cause the same warning to be written.
7368 However, Exim was preserving the list of already written lines for an
7369 entire SMTP session, which doesn't seem right. The memory is now reset if a
7370 new message is started.
7372 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7373 debugging output unless log_rewrite was also set.
7375 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7376 of (int)(handle) when we know that handle contains (void *)(-1).
7378 61. The Exim daemon panic-logs an error return when it closes the incoming
7379 connection. However "connection reset by peer" seems to be common, and
7380 isn't really an error worthy of noting specially, so that particular error
7383 62. When Exim is trying to find all the local interfaces, it used to panic and
7384 die if the ioctl to get the interface flags failed. However, it seems that
7385 on at least one OS (Solaris 9) it is possible to have an interface that is
7386 included in the list of interfaces, but for which you get a failure error
7387 for this call. This happens when the interface is not "plumbed" into a
7388 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7389 failure of the "get flags" call assumes that the interface is down.
7391 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7392 makes life easier for people who are doing arithmetic on fields extracted
7393 from dates, where you often get leading zeros that should not be
7394 interpreted as octal.
7396 64. Added qualify_domain to the redirect router, to override the global
7399 65. If a pathologically long header line contained very many addresses (the
7400 report of this problem mentioned 10 000) and each of them was rewritten,
7401 Exim could use up a very large amount of memory. (It kept on making new
7402 copies of the header line as it rewrote, and never released the old ones.)
7403 At the expense of a bit more processing, the header rewriting function has
7404 been changed so that it no longer eats memory in this way.
7406 66. The generation of the Received: header has been moved from the time that a
7407 message starts to be received, to the time that it finishes. The timestamp
7408 in the Received: header should now be very close to that of the <= log
7409 line. There are two side-effects of this change:
7411 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7412 logged header lines no longer include the local Received: line, because
7413 it has not yet been created. The same applies to a copy of the message
7414 that is returned to a non-SMTP sender when a message is rejected.
7416 (b) When a filter file is tested using -bf, no additional Received: header
7417 is added to the test message. After some thought, I decided that this
7420 This change does not affect the value of $received_for. It is still set
7421 after address rewriting, but before local_scan() is called.
7423 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7425 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7426 gave an unhelpful panic error message, and a defer error. I have managed to
7427 change this behaviour so that it now rejects any supplied certificate,
7428 which seems right, as the list of acceptable certificates is empty.
7430 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7431 gave an unhelpful defer error. I have not managed to make this reject any
7432 supplied certificates, but the error message it gives is "no certificate
7433 supplied", which is not helpful.
7435 70. exigrep's output now also includes lines that are not associated with any
7436 message, but which match the given pattern. Implemented by a patch from
7437 Martin Sluka, which also tidied up the Perl a bit.
7439 71. Recipient callout verification, like sender verification, was using <> in
7440 the MAIL FROM command. This isn't really the right thing, since the actual
7441 sender may affect whether the remote host accepts the recipient or not. I
7442 have changed it to use the actual sender in the callout; this means that
7443 the cache record is now keyed on a recipient/sender pair, not just the
7444 recipient address. There doesn't seem to be a real danger of callout loops,
7445 since a callout by the remote host to check the sender would use <>.
7446 [SEE ABOVE: changed after hitting problems.]
7448 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7449 temporary errors. However, in the case of such a code being given after
7450 the end of a data transmission (i.e. after ".") Exim was failing to write
7451 a retry record for the message. (Yes, there was some broken host that was
7452 actually sending 8xx at this point.)
7454 73. An unknown lookup type in a host list could cause Exim to panic-die when
7455 the list was checked. (An example that provoked this was putting <; in the
7456 middle of a list instead of at the start.) If this happened during a DATA
7457 ACL check, a -D file could be left lying around. This kind of configuration
7458 error no longer causes Exim to die; instead it causes a defer error. The
7459 incident is still logged to the main and panic logs.
7461 74. Buglet left over from Exim 3 conversion. The message "too many messages
7462 in one connection" was written to the rejectlog but not the mainlog, except
7463 when address rewriting (yes!) was being logged.
7465 75. Added write_rejectlog option.
7467 76. When a system filter was run not as root (that is, when system_filter_user
7468 was set), the values of the $n variables were not being returned to the
7469 main process; thus, they were not subsequently available in the $sn
7472 77. Added +return_path_on_delivery log selector.
7474 78. A connection timeout was being treated differently from recipients deferred
7475 when testing hosts_max_try with a message that was older than the host's
7476 retry timeout. (The host should not be counted, thus allowing all hosts to
7477 be tried at least once before bouncing.) This may have been the cause of an
7478 occasionally reported bug whereby a message would remain on the queue
7479 longer than the retry timeout, but would be bounced if a delivery was
7480 forced. I say "may" because I never totally pinned down the problem;
7481 setting up timeout/retry tests is difficult. See also the next item.
7483 79. The ultimate address timeout was not being applied to errors that involved
7484 a combination of host plus message (for example, a timeout on a MAIL
7485 command). When an address resolved to a number of possible hosts, and they
7486 were not all tried for each delivery (e.g. because of hosts_max_try), a
7487 message could remain on the queue longer than the retry timeout.
7489 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7492 81. Fixed an obscure SMTP outgoing bug which required at least the following
7493 conditions: (a) there was another message waiting for the same server;
7494 (b) the server returned 5xx to all RCPT commands in the first message so
7495 that the message was not completed; (c) the server dropped the connection
7496 or gave a negative response to the RSET that Exim sends to abort the
7497 transaction. The observed case was a dropped connection after DATA that had
7498 been sent in pipelining mode. That is, the server had advertised PIPELINING
7499 but was not implementing it correctly. The effect of the bug was incorrect
7500 behaviour, such as trying another host, and this could lead to a crash.
7506 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7507 and daemon.c were passed as pointers to ints; they should have been
7508 pointers to socklen_t variables (which are typically unsigned ints).
7510 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7513 3. Fixed a really odd bug that affected only the testing scheme; patching a
7514 certain fixed string in the binary changed the value of another string that
7515 happened to be identical to the end of the original first string.
7517 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7518 name", it returns that address as the IP address. On some operating
7519 systems (e.g. Solaris), it also passes back the IP address string as the
7520 "host name". However, on others (e.g. Linux), it passes back an empty
7521 string. Exim wasn't checking for this, and was changing the host name to an
7522 empty string, assuming it had been canonicalized.
7524 5. Although rare, it is permitted to have more than one PTR record for a given
7525 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7526 all the names associated with an address, because they do in Solaris.
7527 However, it seems that they do not in Linux for data that comes from the
7528 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7529 I found this out when I moved to a new Linux workstation and tried to run
7530 the Exim test suite.
7532 To get round this problem I have changed the code so that it now does its
7533 own call to the DNS to look up PTR records when searching for a host name.
7534 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7535 addresses that are only in /etc/hosts are still found.
7537 This behaviour is, however, controlled by an option called host_lookup_
7538 order, which defaults to "bydns:byaddr". If people want to use the other
7539 order, or indeed, just use one or the other means of lookup, they can
7540 specify it in this variable.
7542 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7543 some operating systems, this comes back from gethostbyaddr() as an empty
7544 string, and this is what Exim used to test for. However, it seems that in
7545 other systems, "." is yielded. Exim now tests for this case too.
7547 7. The values of check_spool_space and check_log_space are now held internally
7548 as a number of kilobytes instead of an absolute number of bytes. If a
7549 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7550 kilobyte. This means that much larger values can be stored.
7552 8. Exim monitor: an attempt to get the action menu when not actually pointing
7553 at a message produces an empty menu entitled "No message selected". This
7554 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7555 no entries in it ("Shell widget menu has zero width and/or height"). So I
7556 have added a single, blank menu entry in this case.
7558 9. Added ${quote_local_part.
7560 10. MIME decoding is now applied to the contents of Subject: header lines when
7563 11. Now that a reference to $sender_host_address automatically causes a reverse
7564 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7565 host lookup before query-style lookups in lists that might use this
7566 variable. This has therefore been abolished, and the "net-" prefix is no
7567 longer necessary for query-style lookups.
7569 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7570 have been a typo for LFLAGS, so it has been changed.
7572 13. The install script calls Exim with "-C /dev/null" in order to find the
7573 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7574 to be output. However, since Exim outputs its version number before the
7575 error, it didn't break the script. It just looked ugly. I fixed this by
7576 always allowing "-C /dev/null" if the caller is root.
7578 14. Ignore overlarge ACL variable number when reading spool file - insurance
7579 against a later release with more variables having written the file.
7581 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7582 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7583 this, as well as the form without the "IPv6" on the front.
7585 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7586 OS/Makefile-Darwin file.
7588 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7589 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7591 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7592 variable. However, some OS use ints for inodes. I've added cast to long int
7593 to get rid of the compiler warning.
7595 19. I had forgotten to lock out "/../" in configuration file names when
7596 ALT_CONFIG_PREFIX was set.
7598 20. Routers used for verification do not need to specify transports. However,
7599 if such a router generated a host list, and callout was configured, Exim
7600 crashed, because it could not find a port number from the (non-existent)
7601 transport. It now assumes port 25 in this circumstance.
7603 21. Added the -t option to exigrep.
7605 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7606 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7607 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7608 them is set, LOOKUP_LSEARCH is forced.
7610 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7612 24. Added sender and host information to the "rejected by local_scan()" log
7613 line; previously there was no indication of these.
7615 25. Added .include_if_exists.
7617 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7618 Linux. It turns out that not all file systems support this. Apparently some
7619 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7620 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7621 not supported on the file descriptor, is now ignored when Exim is trying to
7622 sync a directory. This applies only to Linux.
7624 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7626 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7627 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7630 29. Added support for maildirsize files from supplied patch (modified a bit).
7632 30. The use of :fail: followed by an empty string could lead Exim to respond to
7633 sender verification failures with (e.g.):
7635 550 Verification failed for <xxx>
7636 550 Sender verify failed
7638 where the first response line was missing the '-' that indicates it is not
7639 the final line of the response.
7641 31. The loop for finding the name of the user that called Exim had a hardwired
7642 limit of 10; it now uses the value of finduser_retries, which is used for
7643 all other user lookups.
7645 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7648 33. Exim was neglecting to zero errno before one call of strtol() when
7649 expanding a string and expecting an integer value. On some systems this
7650 resulted in spurious "integer overflow" errors. Also, it was casting the
7651 result into an int without checking.
7653 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7654 did not work. The code looks as if it has *never* worked, though it appears
7655 to have been documented since at least release 1.62. I have made it work.
7657 35. The "timeout_DNS" error in retry rules, also documented since at least
7658 1.62, also never worked. As it isn't clear exactly what this means, and
7659 clearly it isn't a major issue, I have abolished the feature by treating it
7660 as "timeout", and writing a warning to the main and panic logs.
7662 36. The display of retry rules for -brt wasn't always showing the error code
7665 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7666 timeout_connect_A, timeout_connect_MX.
7668 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7669 to the empty sender.
7671 39. The daemon was not analysing the content of -oX till after it had closed
7672 stderr and disconnected from the controlling terminal. This meant that any
7673 syntax errors were only noted on the panic log, and the return code from
7674 the command was 0. By re-arranging the code a little, I've made the
7675 decoding happen first, so such errors now appear on stderr, and the return
7676 code is 1. However, the actual setting up of the sockets still happens in
7677 the disconnected process, so errors there are still only recorded on the
7680 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7681 connections (as happens on some IP stacks) was logged at start up time as
7682 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7683 it from "IPv6 and IPv4", which means that two separate sockets are being
7686 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7687 says whether AF_INET or AF_INET6 was passed as an argument.
7689 42. Exiwhat output was messed up when time zones were included in log
7692 43. Exiwhat now gives more information about the daemon's listening ports,
7693 and whether -tls-on-connect was used.
7695 44. The "port" option of the smtp transport is now expanded.
7697 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7698 silently ignored. Now an error message is written to the main and panic
7701 46. There's a new ACL modifier called "logwrite" which writes to a log file
7702 as soon as it is encountered.
7704 47. Added $local_user_uid and $local_user_gid at routing time.
7706 48. Exim crashed when trying to verify a sender address that was being
7709 49. Exim was recognizing only a space character after ".include". It now also
7710 recognizes a tab character.
7712 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7713 extracting the relevant information from the specification. The man page no
7714 longer contains scrambled data for the -d option, and I've added a section
7715 at the front about calling Exim under different names.
7717 51. Added "extra_headers" argument to the "mail" command in filter files.
7719 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7722 53. Installed eximstats 1.29.
7724 54. Added transport_filter_timeout as a generic transport option.
7726 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7727 Cc: header lines. This was required by RFC 822, but it not required by RFC
7730 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7731 incoming messages that did not have them. Now it does so only if the
7732 message originates locally, that is, if there is no associated remote host
7733 address. When Resent- header lines are present, this applies to the Resent-
7734 lines rather than the non-Resent- lines.
7736 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7737 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7739 58. Messages for configuration errors now include the name of the main
7740 configuration file - useful now that there may be more than one file in a
7741 list (.included file names were always shown).
7743 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7744 for those rare installations that do not start the daemon as root or run it
7745 setuid root. I've cut out the call to initgroups() if the daemon is not
7748 60. The Exim user and group can now be bound into the binary as text strings
7749 that are looked up at the start of Exim's processing.
7751 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7753 62. Added $mailstore_basename variable.
7755 63. Installed patch to sieve.c from Michael Haardt.
7757 64. When Exim failed to open the panic log after failing to open the main log,
7758 the original message it was trying to log was written to stderr and debug
7759 output, but if they were not available (the usual case in production), it
7760 was lost. Now it is written to syslog before the two lines that record the
7761 failures to open the logs.
7763 65. Users' Exim filters run in subprocesses under the user's uid. It is
7764 possible for a "deliver" command or an alias in a "personal" command to
7765 provoke an address rewrite. If logging of address rewriting is configured,
7766 this fails because the process is not running as root or exim. There may be
7767 a better way of dealing with this, but for the moment (because 4.30 needs
7768 to be released), I have disabled address rewrite logging when running a
7769 filter in a non-root, non-exim process.
7775 1. The buildconfig auxiliary program wasn't quoting the value set for
7776 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7777 not defined. This bug was masked in 4.22 by the effect that was fixed in
7780 2. Some messages that were rejected after a message id was allocated were
7781 shown as "incomplete" by exigrep. It no longer does this for messages that
7782 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7784 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7785 have allow_domain_literals set, the ID did not get logged in the <= line.
7786 Domain literals are now always recognized in Message-ID: header lines.
7788 4. The first argument for a ${extract expansion item is the key name or field
7789 number. Leading and trailing spaces in this item were not being ignored,
7790 causing some misleading effects.
7792 5. When deliver_drop_privilege was set, single queue runner processes started
7793 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7794 same command in the process it spins off) were not dropping privilege.
7796 6. When the daemon running as "exim" started a queue runner, it always
7797 re-executed Exim in the spun-off process. This is a waste of effort when
7798 deliver_drop_privilege is set. The new process now just calls the
7799 queue-runner function directly.
7805 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7808 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7809 ignored. Though the use of -oP was forcing the writing of a pid file, it
7810 was always written to the default place.
7812 3. If the message "no IP address found for host xxxx" is generated during
7813 incoming verification, it is now followed by identification of the incoming
7814 connection (so you can more easily find what provoked it).
7816 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7818 5. Added some features to "harden" Exim a bit more against certain attacks:
7820 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7821 be put in Local/Makefile. This is like the never_users runtime option,
7822 but it cannot be overridden. The default setting is "root".
7824 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7825 prefix string with which any file named in a -C command line option
7828 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7829 is retained for -C and -D only if the caller of Exim is root. Without
7830 it, the exim user may also use -C and -D and retain privilege.
7832 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7833 command line option is disabled.
7835 6. Macro names set by the -D option must start with an upper case letter, just
7836 like macro names defined in the configuration file.
7838 7. Added "dereference=" facility to LDAP.
7840 8. Two instances of the typo "uknown" in the source files are fixed.
7842 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7843 the Configure-Makefile script screwed up while processing it.
7845 10. Incorporated PCRE 4.4.
7847 11. The SMTP synchronization check was not operating right at the start of an
7848 SMTP session. For example, it could not catch a HELO sent before the client
7849 waited for the greeting. There is now a check for outstanding input at the
7850 point when the greeting is written. Because of the duplex, asynchronous
7851 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7852 way, but not yet received, when the check is performed.
7854 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7855 on TCP/IP sockets, because this apparently causes some broken clients to
7858 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7859 unchanged) from the Cygwin maintainer.
7861 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7862 was supported instead of testing for mbx. Effectively a typo.
7864 15. The spa authenticator server code was not checking that the input it
7865 received was valid base64.
7867 16. The debug output line for the "set" modifier in ACLs was not showing the
7868 name of the variable that was being set.
7870 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7872 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7873 lookup on its own. Something else was needed to trigger the lookup. For
7874 example, a match in host_lookup or the need for a host name in a host list.
7875 Now, if $sender_host_name is referenced and the host name has not yet been
7876 looked up, a lookup is performed. If the lookup fails, the variable remains
7877 empty, and $host_lookup_failed is set to "1".
7879 19. Added "eqi" as a case-independent comparison operator.
7881 20. The saslauthd authentication condition could segfault if neither service
7882 nor realm was specified.
7884 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7885 error message that was logged was misleading, and incoming SMTP
7886 connections were dropped. The message is now more accurate, and temporary
7887 errors are given to SMTP connections.
7889 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7890 (except RSET) with a 421 error, until QUIT is received. However, it was
7891 failing to send a response to QUIT.
7893 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7894 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7895 ACL is now run after the helo_try_verify_hosts code.
7897 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7898 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7899 case-independent, so other case variants are also recognized). Apparently
7900 some systems use these upper case variants.
7902 25. If more than two messages were waiting for the same host, and a transport
7903 filter was specified for the transport, Exim sent two messages over the
7904 same TCP/IP connection, and then failed with "socket operation on non-
7905 socket" when it tried to send the third.
7907 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7909 27. The extern definition of crypt16() in expand.c was not being excluded when
7910 the OS had its own crypt16() function.
7912 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7913 as a preferred synonym for return_size_limit, both as an option and as an
7916 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7918 30. Changed the default configuration ACL to relax the local part checking rule
7919 for addresses that are not in any local domains. For these addresses,
7920 slashes and pipe symbols are allowed within local parts, but the sequence
7921 /../ is explicitly forbidden.
7923 31. SPA server authentication was not clearing the challenge buffer before
7926 32. log_message in a "warn" ACL statement was writing to the reject log as
7927 well as to the main log, which contradicts the documentation and doesn't
7928 seem right (because no rejection is happening). So I have stopped it.
7930 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7931 However, I am unable to do any testing of this.
7933 34. Fixed an infelicity in the appendfile transport. When checking directories
7934 for a mailbox, to see if any needed to be created, it was accidentally
7935 using path names with one or more superfluous leading slashes; tracing
7936 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7938 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7939 contents are added to the log line that is written for every discarded
7940 recipient. (Previously a log_message setting was ignored.)
7942 36. The ${quote: operator now quotes the string if it is empty.
7944 37. The install script runs exim in order to find its version number. If for
7945 some reason other than non-existence or emptiness, which it checks, it
7946 could not run './exim', it was installing it with an empty version number,
7947 i.e. as "exim-". This error state is now caught, and the installation is
7950 38. An argument was missing from the function that creates an error message
7951 when Exim fails to connect to the socket for saslauthd authentication.
7952 This could cause Exim to crash, or give a corrupted message.
7954 39. Added isip, isip4, and isip6 to ${if conditions.
7956 40. The ACL variables $acl_xx are now saved with the message, and can be
7957 accessed later in routers, transports, and filters.
7959 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7960 strings in the file are not string-expanded.
7962 42. If a MAIL command specified a SIZE value that was too large to fit into an
7963 int variable, the check against message_size_limit failed. Such values are
7964 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7965 day this will have to be increased, but I don't think I want to be around
7966 when emails are that large.
7973 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7974 iconv() is not standard in FreeBSD.
7976 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7977 IPv6 enabled. The observed symptom was a segmentation fault on return from
7978 the function os_common_find_running_interfaces() in src/os.c.
7980 3. In the check_special_case() function in daemon.c I had used "errno" as an
7981 argument name, which causes warnings on some systems. This was basically a
7982 typo, since it was named "eno" in the comments!
7984 4. The code that waits for the clock to tick (at a resolution of some fraction
7985 of a second) so as to ensure message-id uniqueness was always waiting for
7986 at least one whole tick, when it could have waited for less. [This is
7987 almost certainly not relevant at current processor speeds, where it is
7988 unlikely to ever wait at all. But we try to future-proof.]
7990 5. The function that sleeps for a time interval that includes fractions of a
7991 second contained a race. It did not block SIGALRM between setting the
7992 timer, and suspending (a couple of lines later). If the interval was short
7993 and the sigsuspend() was delayed until after it had expired, the suspension
7994 never ended. On busy systems this could lead to processes getting stuck for
7997 6. Some uncommon configurations may cause a lookup to happen in a queue runner
7998 process, before it forks any delivery processes. The open lookup caching
7999 mechanism meant that the open file or database connection was passed into
8000 the delivery process. The problem was that delivery processes always tidy
8001 up cached lookup data. This could cause a problem for the next delivery
8002 process started by the queue runner, because the external queue runner
8003 process does not know about the closure. So the next delivery process
8004 still has data in the lookup cache. In the case of a file lookup, there was
8005 no problem because closing a file descriptor in a subprocess doesn't affect
8006 the parent. However, if the lookup was caching a connection to a database,
8007 the connection was closed, and the second delivery process was likely to
8008 see errors such as "PGSQL: query failed: server closed the connection
8009 unexpectedly". The problem has been fixed by closing all cached lookups
8010 in a queue runner before running a delivery process.
8012 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8013 seem to have the "const" qualifier which it has on other OS. I've
8016 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8017 *of the same type* (client or server) with the same public name that an
8018 error should be diagnosed.
8020 9. When Exim looked up a host name for an IP address, but failed to find the
8021 original IP address when looking up the host name (a safety check), it
8022 output the message "<ip address> does not match any IP for NULL", which was
8023 confusing, to say the least. The bug was that the host name should have
8024 appeared instead of "NULL".
8026 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8027 user that is built into the binary, and the -C or -D options is used, root
8028 privilege is dropped before the configuration file is read. In addition,
8029 logging is switched to stderr instead of the normal log files. If the
8030 configuration then re-defines the Exim user, the unprivileged environment
8031 is probably not what is expected, so Exim logs a panic warning message (but
8034 However, if deliver_drop_privilege is set, the unprivileged state may well
8035 be exactly what is intended, so the warning has been cut out in that case,
8036 and Exim is allowed to try to write to its normal log files.
8042 1. smtp_return_error_details was not giving details for temporary sender
8043 or receiver verification errors.
8045 2. Diagnose a configuration error if two authenticators have the same public
8048 3. Exim used not to create the message log file for a message until the first
8049 delivery attempt. This could be confusing when incoming messages were held
8050 for policy or load reasons. The message log file is now created at the time
8051 the message is received, and an initial "Received" line is written to it.
8053 4. The automatically generated man page for command line options had a minor
8054 bug that caused no ill effects; however, a more serious problem was that
8055 the procedure for building the man page automatically didn't always
8056 operate. Consequently, release 4.20 contains an out-of-date version. This
8057 shouldn't happen again.
8059 5. When building Exim with embedded Perl support, the script that builds the
8060 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8061 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8063 6. The freeze_tell option was not being used for messages that were frozen on
8064 arrival, either by an ACL or by local_scan().
8066 7. Added the smtp_incomplete_transaction log selector.
8068 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8069 was accepting AUTH without a new EHLO.
8071 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8072 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8075 10. Logging of TCP/IP connections (when configured) now happens in the main
8076 daemon process instead of the child process, so that the TCP/IP connection
8077 count is more accurate (but it can never be perfect).
8079 11. The use of "drop" in a nested ACL was not being handled correctly in the
8080 outer ACL. Now, if condition failure induced by the nested "drop" causes
8081 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8082 or "require"), the connection is dropped.
8084 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8085 that yield "discard" can now be used with an "accept" or a "discard" verb,
8086 but an error is generated for any others (because I can't see a useful way
8087 to define what should happen).
8089 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8090 are now processed in the same way as lines in the Exim configuration file.
8091 In particular, continuation lines are supported.
8093 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8095 15. Added -ti meaning -t -i.
8097 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8098 domains, and warn by logging if others are found.
8100 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8101 returned from GIFCONF when Exim is trying to find the list of interfaces on
8102 a host. The code in os.c has been modified to copy each ifreq to an aligned
8103 structure in all cases.
8105 Also, in some cases, the returned ifreq's were being copied to a 'struct
8106 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8107 means the last couple of bytes of an IPv6 address could be chopped if the
8108 ifreq contained only a normal sockaddr (14 bytes storage).
8110 18. Named domain lists were not supported in the hosts_treat_as_local option.
8111 An entry such as +xxxx was not recognized, and was treated as a literal
8114 19. Ensure that header lines added by a DATA ACL are included in the reject log
8115 if the ACL subsequently rejects the message.
8117 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8118 MD5 (which is deprecated).
8120 21. When testing a filter file using -bf, Exim was writing a message when it
8121 took the sender from a "From " line in the message, but it was not doing so
8122 when it took $return_path from a Return-Path: header line. It now does.
8124 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8125 with a valid header line field name (a series of printing characters
8126 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8128 23. Changed "disc" in the source to "disk" to conform to the documentation and
8129 the book and for uniformity.
8131 24. Ignore Sendmail's -Ooption=value command line item.
8133 25. When execve() failed while trying to run a command in a pipe transport,
8134 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8135 could be confused with a return value of 69 from the command itself. This
8136 has been changed to 127, the value the shell returns if it is asked to run
8137 a non-existent command. The wording for the related log line suggests a
8138 non-existent command as the problem.
8140 26. If received_header_text expands to an empty string, do not add a Received:
8141 header line to the message. (Well, it adds a token one on the spool, but
8142 marks it "old" so that it doesn't get used or transmitted.)
8144 27. Installed eximstats 1.28 (addition of -nt option).
8146 28. There was no check for failure on the call to getsockname() in the daemon
8147 code. This can fail if there is a shortage of resources on the system, with
8148 ENOMEM, for example. A temporary error is now given on failure.
8150 29. Contrary to the C standard, it seems that in some environments, the
8151 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8152 program. Exim now does this explicitly; it affects the formatting of
8153 timestamps using strftime().
8155 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8156 complaints. I've now initialized all the variables, to avoid this.
8158 32. Header lines added by a system filter were not being "seen" during
8159 transport-time rewrites.
8161 33. The info_callback() function passed to OpenSSL is set up with type void
8162 (*)(SSL *, int, int), as described somewhere. However, when calling the
8163 function (actually a macro) that sets it up, the type void(*)() is
8164 expected. I've put in a cast to prevent warnings from picky compilers.
8166 34. If a DNS black list lookup found a CNAME record, but there were no A
8167 records associated with the domain it pointed at, Exim crashed.
8169 35. If a DNS black list lookup returned more than one A record, Exim ignored
8170 all but the first. It now scans all returned addresses if a particular IP
8171 value is being sought. In this situation, the contents of the
8172 $dnslist_value variable are a list of all the addresses, separated by a
8175 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8176 to accept a host name and all its aliases if the forward lookup for any of
8177 them yielded the IP address of the incoming connection. Now it accepts only
8178 those names whose forward lookup yields the correct IP address. Any other
8179 names are discarded. This closes a loophole whereby a rogue DNS
8180 administrator could create reverse DNS records to break through a
8181 wildcarded host restriction in an ACL.
8183 37. If a user filter or a system filter that ran in a subprocess used any of
8184 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8185 the wrong values were passed to the pipe command ($thisaddress had the
8186 value of $0, $0 had the value of $1, etc). This bug was introduced by
8187 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8189 38. Improved the line breaking for long SMTP error messages from ACLs.
8190 Previously, if there was no break point between 40 and 75 characters, Exim
8191 left the rest of the message alone. Two changes have been made: (a) I've
8192 reduced the minimum length to 35 characters; (b) if it can't find a break
8193 point between 35 and 75 characters, it looks ahead and uses the first one
8194 that it finds. This may give the occasional overlong line, but at least the
8195 remaining text gets split now.
8197 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8198 file descriptors might be low, and that setting 1000 would always raise it.
8199 It turns out that in some environments, the limit is already over 1000 and
8200 that lowering it causes trouble. So now Exim takes care not to decrease it.
8202 40. When delivering a message, the value of $return_path is set to $sender_
8203 address at the start of routing (routers may change the value). By an
8204 oversight, this default was not being set up when an address was tested by
8205 -bt or -bv, which affected the outcome if any router or filter referred to
8208 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8209 log only when "message" or "log_message" are set. However, if one of the
8210 conditions was an address verification, or a call to a nested ACL, the
8211 messages generated by the underlying test were being passed through. This
8212 no longer happens. The underlying message is available in $acl_verify_
8213 message for both "message" and "log_message" expansions, so it can be
8214 passed through if needed.
8216 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8217 new expansion $bh_ to give the encoded byte string without charset
8218 translation. Translation happens only if iconv() is available; HAVE_ICONV
8219 indicates this at build time. HEADERS_CHARSET gives the charset to
8220 translate to; headers_charset can change it in the configuration, and
8221 "headers charset" can change it in an individual filter file.
8223 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8224 that creates RFC 2047 encoded "words" labels them as that charset instead
8225 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8226 expansion operator; (ii) when Exim creates a From: line for a local
8227 message; (iii) when a header line is rewritten to include a "phrase" part.
8229 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8230 buggy, causing it to skip the first lines of messages whose message ID
8231 ended in 'D'. This would not have bitten before Exim release 4.14, because
8232 message IDs were unlikely to end in 'D' before then. The effect was to have
8233 incorrect size information for certain domains.
8235 45. #include "config.h" was missing at the start of the crypt16.c module. This
8236 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8239 46. If there was a timeout during a "random" callout check, Exim treated it as
8240 a failure of the random address, and carried on sending RSET and the real
8241 address. If the delay was just some slowness somewhere, the response to the
8242 original RCPT would be taken as a response to RSET and so on, causing
8243 mayhem of various kinds.
8245 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8246 when I implemented it. It didn't allow for the fact that some option values
8247 may legitimately be negative (e.g. size_addition), and it didn't even do
8248 the right test for positive values.
8250 48. Domain names in DNS records are case-independent. Exim always looks them up
8251 in lower case. Some resolvers return domain names in exactly the case they
8252 appear in the zone file, that is, they may contain uppercase letters. Not
8253 all resolvers do this - some return always lower case. Exim was treating a
8254 change of case by a resolver as a change of domain, similar to a widening
8255 of a domain abbreviation. This triggered its re-routing code and so it was
8256 trying to route what was effectively the same domain again. This normally
8257 caused routing to fail (because the router wouldn't handle the domain
8258 twice). Now Exim checks for this case specially, and just changes the
8259 casing of the domain that it ultimately uses when it transmits the message
8262 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8265 50. If a filter generated a file delivery with a non-absolute name (possible if
8266 no home directory exists for the router), the forbid_file option was not
8269 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8270 the existing equality matching.
8272 52. Exim was using ints instead of ino_t variables in some places where it was
8273 dealing with inode numbers.
8275 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8276 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8277 called TMPDIR, and if it finds it is different, it changes its value.
8279 54. The smtp_printf() function is now made available to local_scan() so
8280 additional output lines can be written before returning. There is also an
8281 smtp_fflush() function to enable the detection of a dropped connection.
8282 The variables smtp_input and smtp_batched_input are exported to
8285 55. Changed the default runtime configuration: the message "Unknown user"
8286 has been removed from the ACL, and instead placed on the localuser router,
8287 using the cannot_route_message feature. This means that any verification
8288 failures that generate their own messages won't get overridden. Similarly,
8289 the "Unrouteable address" message that was in the ACL for unverifiable
8290 relay addresses has also been removed.
8292 56. Added hosts_avoid_esmtp to the smtp transport.
8294 57. The exicyclog script was not checking for the esoteric option
8295 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8296 will work only if exicyclog is run under the appropriate euid.
8298 58. Following a discussion on the list, the rules by which Exim recognises line
8299 endings on incoming messages have been changed. The -dropcr and drop_cr
8300 options are now no-ops, retained only for backwards compatibility. The
8301 following line terminators are recognized: LF CRLF CR. However, special
8302 processing applies to CR:
8304 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8305 nor a local message in the state where . is a terminator.
8307 (ii) If a bare CR is encountered in a header line, an extra space is added
8308 after the line terminator so as not to end the header. The reasoning
8309 behind this is that bare CRs in header lines are most likely either
8310 to be mistakes, or people trying to play silly games.
8312 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8313 was being incorrectly given as 18 bytes larger than it should have been.
8314 This is a VOB (very old bug).
8316 60. This may never have affected anything current, but just in case it has:
8317 When the local host is found other than at the start of a list of hosts,
8318 the local host, those with the same MX, and any that follow, are discarded.
8319 When the list in question was part of a longer list of hosts, the following
8320 hosts (not currently being processed) were also being discarded. This no
8321 longer happens. I'm not sure if this situation could ever has previously
8324 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8327 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8328 header. This is something that is recommended in a new Internet Draft, and
8329 is something that is documented as being done by Sendmail. There are two
8330 possible values. For messages generated by the autoreply transport, Exim
8333 Auto-Submitted: auto-replied
8335 whereas for all other generated messages (e.g. bounces) it adds
8337 Auto-Submitted: auto-generated
8339 63. The "personal" condition in filters now includes a test for the
8340 Auto-Submitted: header. If it contains the string "auto-" the message it
8341 not considered personal.
8343 64. Added rcpt_include_affixes as a generic transport option.
8345 65. Added queue_only_override (default true).
8347 66. Added the syslog_duplication option.
8349 67. If what should have been the first header line of a message consisted of
8350 a space followed by a colon, Exim was mis-interpreting it as a header line.
8351 It isn't of course - it is syntactically invalid and should therefore be
8352 treated as the start of the message body. The misbehaviour could have
8353 caused a number of strange effects, including loss of data in subsequent
8354 header lines, and spool format errors.
8356 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8357 client host had authenticated. This control can now be exercised by an ACL
8358 for more flexibility.
8360 69. By default, callouts do not happen when testing with -bh. There is now a
8361 variant, -bhc, which does actually run the callout code, including
8362 consulting and updating the callout cache.
8364 70. Added support for saslauthd authentication, courtesy of Alexander
8367 71. If statvfs() failed on the spool or log directories while checking their
8368 size for availability, Exim confusingly gave the error "space shortage".
8369 Furthermore, in debugging mode it crashed with a floating point exception.
8370 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8371 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8372 serious problem, Exim now writes to the main and panic logs when this
8373 happens, with details of the failure. It then refuses to accept the
8374 incoming message, giving the message "spool directory problem" or "log
8375 directory problem" with a 421 code for SMTP messages.
8377 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8378 0, 1, and 2 exist, because some OS complain for execs without them (see
8379 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8380 descriptors. However, the code omitted to check that the open succeeded,
8381 causing mysterious errors if for some reason the permissions on /dev/null
8382 got screwed. Now Exim writes a message to the main and panic logs, and
8383 bombs out if it can't open /dev/null.
8385 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8386 interact so that it is all more flexible. It is supposed to remain
8387 backwards compatible. Also added extra_local_interfaces.
8389 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8390 to bomb out with an assertion failure - to the client this appears as a
8391 connection drop. This problem occurs in the part of the code that was taken
8392 from the Samba project. Fortunately, the assertion is in a very simple
8393 function, so I have fixed this by reproducing the function inline in the
8394 one place where it is called, and arranging for authentication to fail
8395 instead of killing the process with assert().
8397 75. The SPA client code was not working when the server requested OEM rather
8398 than Unicode encoding.
8400 76. Added code to make require_files with a specific uid setting more usable in
8401 the case where statting the file as root fails - usually a non-root-mounted
8402 NFS file system. When this happens and the failure is EACCES, Exim now
8403 forks a subprocess and does the per-uid checking as the relevant uid.
8405 77. Added process_log_path.
8407 78. If log_file_path was not explicitly set, a setting of check_log_space or
8408 check_log_inodes was ignored.
8410 79. If a space check for the spool or log partitions fails, the incident is now
8411 logged. Of course, in the latter case the data may get lost...
8413 80. Added the %p formatting code to string_format() so that it can be used to
8414 print addresses in debug_print(). Adjusted all the address printing in the
8415 debugging in store.c to use %p rather than %d.
8417 81. There was a concern that a line of code in smtp_in.c could overflow a
8418 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8419 initially expressed, the concern was not well-founded, because trailing
8420 spaces are removed early. However, if the trailing spaces were followed by
8421 a NULL, they did not get removed, so the overflow was possible. Two fixes
8424 (a) I re-wrote the offending code in a cleaner fashion.
8425 (b) If an incoming SMTP command contains a NULL character, it is rejected
8428 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8429 runs initgroups(), so that if the Exim user is in any additional groups,
8430 they will be used during message reception.
8436 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob