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
166 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
167 Found and fixed by Jasen Betts. No testcase for this as my usual text
168 editor insists on emitting only valid UTF-8.
170 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
171 a null-indirection SIGSEGV for the receive process.
173 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
174 -bh use. Internal checking would spot this and take a panic.
176 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
177 In 4.96 this would expand to empty.
179 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
180 certificate. Find and fix by Andreas Metzler.
182 JH/36 Add ARC info to DMARC hostory recordsl
188 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
189 after reception to before a subsequent reception. This should
190 mean slightly faster delivery, and also confirmation of reception
193 JH/02 Move from using the pcre library to pcre2. The former is no longer
194 being developed or supported (by the original developer).
196 JH/03 Constification work in the filters module required a major version
197 bump for the local-scan API. Specifically, the "headers_charset"
198 global which is visible via the API is now const and may therefore
199 not be modified by local-scan code.
201 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
202 sendfile() didi not account for the way the ClamAV driver code called it.
204 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
205 time check was being done for every character; replace that with one
208 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
209 sent was prefixed with a length byte.
211 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
212 RFC 5321. Previously Dovecot (at least) would log errors during
215 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
216 in 4.61 without a reason logged. Binaries will be bigger, which might
217 matter on diskspace-constrained systems, but debug is easier.
219 JH/09 Fix macro-definition during "-be" expansion testing. The move to
220 write-protected store for macros had not accounted for these runtime
221 additions; fix by removing this protection for "-be" mode.
223 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
224 handing out large-numbered file descriptors, violating the usual Unix
225 assumption (and required by Posix) that the lowest possible number will be
226 allocated by the kernel when a new one is needed. In the daemon, and any
227 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
228 are not useable for FD_SET() [and hence select()] and overwrite the stack.
229 Assorted crashes happen.
231 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
232 main-section options or in a connect ACL, the value from the first ever
233 connection was never replaced for subsequent connections. Found by
236 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
237 though only once PCRE2 was introduced: the memory accounting used under
238 debug offset allocations by an int, giving a hard trap in early startup.
239 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
241 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
242 with underbars is given. The write-protection of configuration introduced
243 in 4.95 trapped when normalisation was applied to an option not needing
246 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
248 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
249 failing to close the certificates directory, every hour or any time it
252 JH/16 Debugging initiated by an ACL control now continues through into routing
253 and transport processes. Previously debugging stopped any time Exim
254 re-execs, or for processing a queued message.
256 JH/17 The "expand" debug selector now gives more detail, specifically on the
257 result of expansion operators and items.
259 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
260 bad comparison between the option value and the name of the file to
261 be included was done, and a mismatch was wrongly identified.
262 4.88 to 4.95 are affected.
264 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
266 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
267 supplied as the db file-pair basename. Previously, if a directory
268 path was given, for example via the autoreply "once" option, the DB
269 file.pag and file.dir files would be created in that directory's
272 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
273 "taint" log_selector. These were previously deprecated.
275 JH/22 Fix static address-list lookups to properly return the matched item.
276 Previously only the domain part was returned.
278 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
279 the call into OpenSSL to send a TLS Close was being repeated; this
280 resulted in the library waiting for the peer's Close. If that was never
281 sent we waited forever. Fix by tracking send calls.
283 JH/24 The ${run} expansion item now expands its command string elements after
284 splitting. Previously it was before; the new ordering makes handling
285 zero-length arguments simpler. The old ordering can be obtained by
286 appending a new option "preexpand", after a comma, to the "run".
288 JH/25 Taint-check exec arguments for transport-initiated external processes.
289 Previously, tainted values could be used. This affects "pipe", "lmtp" and
290 "queryprogram" transport, transport-filter, and ETRN commands.
291 The ${run} expansion is also affected: in "preexpand" mode no part of
292 the command line may be tainted, in default mode the executable name
295 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
296 the the facility was not passed across execs, and only the first message
297 passed over a connection could use BDAT; any further ones using DATA.
299 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
300 uses $sending_ip_address and an interface is specified.
301 Previously any use of the local address in the EHLO name disabled
302 PIPECONNECT, the common case being to use the rDNS of it.
304 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
305 resumption. Previously verify failed because no certificate status is
306 passed on the wire for the restarted session. Fix by using the recorded
307 ocsp status of the stored session for the new connection.
309 JH/29 TLS resumption: the key for session lookup in the client now includes
310 more info that a server could potentially use in configuring a TLS
311 session, avoiding oferring mismatching sessions to such a server.
312 Previously only the server IP was used.
314 JH/30 Fix string_copyn() for limit greater than actual string length.
315 Previously the copied amount was the limit, which could result in a
316 overlapping memcpy for newly allocated destination soon after a
317 source string shorter than the limit. Found/investigated by KM.
319 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
320 close; it may be needed for a subsequent connection. This caused a
321 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
323 JH/32 Fix CHUNKING for a second message on a connection when the first was
324 rejected. Previously we did not reset the chunking-offered state, and
325 erroneously rejected the BDAT command. Investigation help from
328 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
329 an empty address. Previously the expansion returned an error.
331 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
332 proxy. Previously these were misparsed, leading to paniclog entries.
338 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
339 related applications. Previously an "H" was used where available info
340 says that "M" should be, so change to match.
342 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
343 as arguments, so an implementation trying to copy these into a local
344 buffer was taking a taint-enforcement trap. Fix by using dynamically
345 created buffers. Similar fix for radius expansion condition.
347 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
348 reasonable, eg. to count headers. Fix by using dynamically created
349 buffers rather than a local. Do similar fixes for ACL actions "dcc",
350 "log_reject_target", "malware" and "spam"; the arguments are expanded
351 so could be handling tainted values.
353 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
354 broken the (no-op) support for this sendmail command. Restore it
355 to doing nothing, silently, and returning good status.
357 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
358 record path was given (or the default used) without a leading directory
359 path, an error occurred on trying to open it. Use the transport's working
362 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
363 transport. Previously it was the name on the DNS A-record; use instead
364 the head of the CNAME chain leading there (if there is one). This seems
365 to align better with RFC 6125.
367 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
368 smtp_accept_max_per_host allocated resources which were not released
369 when the limit was exceeded. This eventually crashed the daemon. Fix
370 by adding a release action in that path.
372 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
373 expanded; previously using tainted values was rejected. Fix by using
374 dynamically-created buffers.
376 JH/09 Relax restrictions on ACL verify condition needing access to message
377 headers. Previously they were only permitted in data and non-smtp ACLs;
378 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
379 not_blind, header_sender and header_names_ascii verification.
381 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
382 Previously a macro used one argument twice; when called with the
383 argument as an expression having side-effects, incorrect operation
384 resulted. Use an inlineable function.
386 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
387 held open for a verify callout. Previously this wan not accounted for
388 and a corrupt onward SMTP conversation resulted.
390 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
391 Previously they were embedded naively in the constructed address; when
392 needed, strip the quoting and quote the entire local_part.
393 Also make the inbound_srs expansion condition handle quoting.
395 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
396 excluded, not matching the documentation.
398 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
399 was given for the sqlite_dbfile a trap resulted.
401 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
402 "name" argument resulted in a trap. There is no reason to disallow such;
403 this was a coding error.
405 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
406 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
407 spent suspended, ignoring the POSIX definition. Previously we assumed
408 it did and a constant offset from real time could be used as a correction.
409 Change to using the same clock source for the start-of-message and the
410 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
411 exists, just to get a clock slightly more aligned to reality.
413 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
414 RFC says it is optional some validators care. The missing char was not
415 intended but triggered by a line-wrap alignment. Discovery and fix by
416 Guillaume Outters, hacked on by JH.
418 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
419 name being quoted was tainted a trap would be taken. Fix by using
420 dynamically created buffers. The routine could have been called by a
421 rewrite with the "h" flag, by using the "-F" command-line option, or
422 by using a "name=" option on a control=submission ACL modifier.
424 JH/19 SPF: change the Authentication-Results expansion component to give
425 smtp.helo when the sender domain is empty. Previously it gave
428 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
429 not in 127.0.0.0/8 to help in spotting list domains taken over by a
430 domain-parking registrar.
432 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
433 Previously when a whitespace character was specified it was not inserted
434 after removing the newline.
436 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
437 the domain part of the recipient address. This overrides any tls_sni
438 option set, which was previously used.
440 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
443 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
444 is_tainted() had an off-by-one error in the overenthusiastic direction.
445 Find and fix by Gavan. Although NetBSD is not a supported platform for
446 4.94 this bug could affect other platforms.
448 PP/01 Fix default prime selection to be consistent.
449 One path used ike23 still, instead of exim.dev.20160529.3; now both
450 execution flows will use the same DH primes (currently
451 exim.dev.20160529.3).
453 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
454 option in smtp transport, to match the documentation. Previously
455 verification was not being done in some cases where it should have been.
457 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
458 than one server was defined and depending on the platform memory layout
459 details, an internal consistency trap could be hit while walking the list
462 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
463 files. The value is used by the authresults expansion item. Previously
464 if this was used in a router or transport, a crash could result.
466 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
467 have a sufficient incidence of EINTR returns from select that an
468 interaction with other operations done by the main daemon loop exposed
469 a bug in the error-handling. This was benign apart from the log
472 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
473 for consistency with delivered "=>" and failed "**" lines. While we're
474 there, handle PRX and TFO.
476 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
477 applied. This resulted, if any header-line rewrite rules were configured,
478 in a panic-log triggerable by sending a message with a long address in
479 a header. Fix by increasing the arbitrary limit to larger than a single
480 (dewrapped) 5322 header line maximum size.
482 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
483 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
484 RFC 6648 which deprecates X- options in protocols as a general practice.
485 Changeover between the implementations is handled by the mechanisms
488 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
489 a remote which uses callout/hold. Previously the local was queued.
491 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
494 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
495 log line. The asterisk is now consistently appended to the remote IP
496 (and port, if given), and will also be provided on defer and fail log
497 lines. Previously it could be placed on the local IP if that was being
498 logged, and was only provided on delivery lines.
500 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
502 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
503 authenticator client_send option. Previously the next char, after a pair
504 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
505 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
506 documentation. There is still no way to get a leading ^ immediately
507 after a NUL (ie. for the password of a PLAIN method authenticator.
509 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
510 files. For bad sizes read, delete the record and whine to paniclog.
512 JH/38 When logging an AUTH failure, as server, do not include sensitive
513 information. Previously, the credentials would be included if given
514 as part of the AUTH command line and an ACL denied authentication.
516 JH/39 Bug 2691: fix $local_part_data. When the matching list element
517 referred to a file, bad data was returned. This likely also affected
520 JH/40 The gsasl authenticator now supports caching of the salted password
521 generated by the client-side implementation. This required the addition
522 of a new variable: $auth4.
524 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
525 left undeleted; the attempt to re-create it then failed - resulting in
526 the usual "SIGHUP tp have daemon reload configuration" to not work.
527 This affected any platform not supporting "abstract" Unix-domain
528 sockets (i.e. not Linux).
530 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
531 recipients" response to RCPT in a later response, with a 250. The
532 previous coding assumed this would not happen, and under PIPELINING
533 would result in both lost and duplicate recipients for a message.
535 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
536 Previously the weighting was incorrectly applied. Similar fix for socks
537 proxies. Found and fixed by Heiko Schlichting.
539 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
540 not handle sub-lists included using the +namedlist syntax. While
541 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
542 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
545 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
546 Previously the permanent pool was used, so the sore could not be freed.
547 This meant a connection with many messages would use continually-growing
550 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
551 per-pool so as not to waste too much space. Previously a constant size
552 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
553 DOS attacks harder. The cost is wasted memory use in the larger blocks.
555 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
556 same space repeatedly, and a smaller process footprint.
558 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
559 for GnuTLS. Previously the name started "%s" which, while not a bug,
560 looked as if if might be one.
562 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
563 "continued connection") make the $proxy_* variables available. Previously
564 the information was not passed across the exec() call for subsequent
565 transport executions. This also mean that the log lines for the
566 messages can show the proxy information.
568 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
569 receive time. With modern systems the difference is significant.
570 The historical behaviour can be restored by disabling (a new) log_selector
571 "queue_time_exclusive".
573 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
574 content) were specifically tested for. Now, also cover expansions
575 resulting in ACL names and inline ACL content.
577 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
578 making it unusable in complex expressions.
580 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
581 fail with a taint-check on the spoolfile name, and leave the message
584 HS/01 Enforce absolute PID file path name.
586 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
588 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
590 PP/02 Bug 2643: Correct TLS DH constants.
591 A missing NUL termination in our code-generation tool had led to some
592 incorrect Diffie-Hellman constants in the Exim source.
593 Reported by kylon94, code-gen tool fix by Simon Arlott.
595 PP/03 Impose security length checks on various command-line options.
596 Fixes CVE-2020-SPRSS reported by Qualys.
598 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
599 better. Reported by Qualys.
601 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
602 providing a particularly obnoxious sender full name.
605 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
607 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
610 PP/08 Change default for recipients_max from unlimited to 50,000.
612 PP/09 Fix security issue with too many recipients on a message (to remove a
613 known security problem if someone does set recipients_max to unlimited,
614 or if local additions add to the recipient list).
615 Fixes CVE-2020-RCPTL reported by Qualys.
617 PP/10 Fix security issue in SMTP verb option parsing
618 Fixes CVE-2020-EXOPT reported by Qualys.
620 PP/11 Fix security issue in BDAT state confusion.
621 Ensure we reset known-good where we know we need to not be reading BDAT
622 data, as a general case fix, and move the places where we switch to BDAT
623 mode until after various protocol state checks.
624 Fixes CVE-2020-BDATA reported by Qualys.
626 HS/03 Die on "/../" in msglog file names
628 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
629 the Exim runtime user are allowed to create files.
631 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
634 QS/03 When reading the output from interpreted forward files we do not
635 pass the pipe between the parent and the interpreting process to
636 executed child processes (if any).
638 QS/04 Always die if requested from internal logging, even is logging is
641 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
642 the API; compilation noo longer completes with DMARC support included.
643 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
644 been reported on other platforms.
646 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
648 JH/56 Make the majority of info read from config files readonly, for defence-in-
649 depth against exploits. Suggestion by Qualys.
650 Not supported on Solaris 10.
652 JH/57 Fix control=fakreject for a custom message containing tainted data.
653 Previously this resulted in a log complaint, due to a re-expansion present
654 since fakereject was originally introduced.
656 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
657 modifier was given, a loop resulted.
659 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
660 pipelined SMTP command followed the BDAT LAST then it would be
661 incorrectly treated as part of the message body, causing a verification
664 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
665 require looser validation rules than those for 821-level addresses,
666 which only permit IP addresses.
672 JH/01 Avoid costly startup code when not strictly needed. This reduces time
673 for some exim process initialisations. It does mean that the logging
674 of TLS configuration problems is only done for the daemon startup.
676 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
678 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
679 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
682 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
683 DKIM signing is being done. Previously a transport_filter would always
684 disable CHUNKING, falling back to traditional DATA.
686 JH/05 Regard command-line recipients as tainted.
688 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
690 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
691 PAM library frees one of the arguments given to it, despite the
692 documentation. Therefore a plain malloc must be used.
694 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
695 on-stack buffers were used, resulting in a taint trap when DSN information
696 copied from a received message was written into the buffer.
698 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
699 the ordering of its ARC headers. This caused a crash.
701 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
702 a new record was being constructed with information from the peer, a trap
705 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
706 installation would get error messages from DMARC verify, when it hit the
707 nonexistent file indicated by the default. Distros wanting DMARC enabled
708 should both provide the file and set the option.
709 Also enforce no DMARC verification for command-line sourced messages.
711 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
712 could, depending on the platform, hang at the STARTTLS response.
714 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
715 message on a connection. Previously if one message had ARC headers and
716 the following one did not, a crash could result when adding an
717 Authentication-Results: header.
719 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
720 Exim main code and Exim-related utities. The introduction of taint
721 tracking also did many adjustments to string handling. Since then, eximon
722 frequently terminated with an assert failure.
724 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
725 check for 452 responses. This slightly helps the inefficieny of doing
726 a large alias-expansion into a recipient-limited target. The max_rcpt
727 transport option still applies (and at the current default, will override
728 the new feature). The check is done for either cause of synch, and forces
729 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
730 connection. The new facility is not tunable at this time.
732 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
733 library live data was being used, so the results became garbage. Make
734 copies while it is still usable.
736 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
737 on delivery deferred (==) and failed (**) lines (if a delivery was
738 attemtped). Previously it was only on completion (=>) lines.
740 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
741 for the expansion of the server_scram_iter and server_scram_salt options.
743 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
744 are now specifically given a NO_DATA response without hitting the system
745 resolver. The library goes on to do the now-standard TXT lookup.
746 Use of dnsdb lookups is not affected.
748 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
749 only retrieve the errormessage once. Previously two calls to dlerror()
750 were used, and the second one (for mainlog/paniclog) retrieved null
753 JH/20 Taint checking: disallow use of tainted data for
754 - the appendfile transport file and directory options
755 - the pipe transport command
756 - the autoreply transport file, log and once options
757 - file names used by the redirect router (including filter files)
759 - paths used by single-key lookups
760 Previously this was permitted.
762 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
763 adjusted the size of a major service buffer; this failed because the
764 buffer was in use at the time. Change to a compile-time increase in the
765 buffer size, when this authenticator is compiled into exim.
767 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
768 previous fast-mode was untenable in the face of glibs using mmap to
769 support larger malloc requests.
771 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
772 New values supported, if defined on system where compiled:
773 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
774 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
776 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
777 running a limited number of proceses in parallel, a benefit is gained. The
778 amount varies with the platform hardware and load. The use of the option
779 queue_run_in_order means we cannot do this, as ordering becomes
782 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
783 had introduced a string-copy (for ensuring NUL-termination) which was not
784 appropriate for that case, which can include embedded NUL bytes in the
785 block of data. Investigation showed the copy to actually be needless, the
786 data being length-specified.
788 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
789 done during a receiving connection, and both used TLS, global info was
790 used rather than per-connection info for tracking the state of data
791 queued for transmission. This could result in a connection hang.
793 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
794 Previously, when delivering serveral messages down a single connection
795 only the first would provide a SIZE. This was due to the size information
796 not being properly tracked.
798 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
799 TAI (at 37 seconds currently), pretend to be in UTC for time-related
800 expansion and logging. Previously, spurious values such as a future
801 minute could be seen.
803 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
804 it could crash from a null-deref. This could also affect the
805 ${addresses: } operator and ${readsock } item.
807 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
808 message following a mime one, the variable was not reset.
810 JH/30 When an pipelined-connect fails at the first response, assume incorrect
811 cached capability (perhaps the peer reneged?) and immediately retry in
814 JH/31 Fix spurious detection of timeout while writing to transport filter.
816 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
817 an attempt to copy the string was made before checking it.
819 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
820 taint of the lookup key was maintained; we now regard the presence in the
821 filesystem as sufficient validation.
823 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
824 argument is supplied.
826 JH/35 The dsearch lookup now requires that the directory is an absolute path.
827 Previously this was not checked, and nonempty relative paths made an
828 access under Exim's current working directory.
830 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
831 Previously no event was raised.
833 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
834 parameter supplied by the sender MAIL FROM command. Previously it was
835 ignored, and only the check_spool_space option value for the required
838 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
839 the size of the signing public-key. Previously it was instead giving
840 the size of the signature hash.
842 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
843 the default. See the (new) dkim_verify_min_keysizes option.
845 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
846 and an ACL use a lookup for checking either the local_part or domain,
847 stale data could be accessed. Ensure that variable references are
848 dropped between messages.
850 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
851 by the client was not checked as pointing within response data before
852 being used. A malicious client could thus cause an out-of-bounds read and
853 possibly gain authentication. Fix by adding the check.
855 JH/42 Internationalisation: change the default for downconversion in the smtp
856 transport to be "if needed". Previously it was "as previously set" for
857 the message, which usually meant "if needed" for message-submission but
858 "no" for everything else. However, MTAs have been seen using SMTPUTF8
859 even when the envelope addresses did not need it, resulting in forwarding
860 failures to non-supporting MTAs. A downconvert in such cases will be
861 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
862 transport. The change does mean that addresses needing conversion will
863 be converted when previously a delivery failure would occur.
865 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
866 response was received it would be used unchecked in a fail-DSN, violating
867 standards on line-length limits. Truncate if needed.
869 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
870 doesn't work. (Additionally add a new main config option to configure the
877 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
878 side, to decode a TLS 1.3 packet capture.
880 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
881 Previously the default library behaviour applied, sending two, each in
884 JH/03 Debug output for ACL now gives the config file name and line number for
887 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
889 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
891 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
892 buffer overrun for (non-chunking) other transports.
894 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
895 TLS1.3, means that a server rejecting a client certificate is not visible
896 to the client until the first read of encrypted data (typically the
897 response to EHLO). Add detection for that case and treat it as a failed
898 TLS connection attempt, so that the normal retry-in-clear can work (if
899 suitably configured).
901 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
902 and/or domain. Found and fixed by Jason Betts.
904 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
905 configuration). If a CNAME target was not a wellformed name pattern, a
908 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
909 the OS reports them interleaved with other addresses.
911 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
912 used both for input and for a verify callout, both encrypted, SMTP
913 responses being sent by the server could be lost. This resulted in
914 dropped connections and sometimes bounces generated by a peer sending
917 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
918 string. Previously it was possible to cause undefined behaviour in a
919 library routine (usually a crash). Found by "zerons".
921 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
924 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
925 API was removed, so update to use the newer ones.
927 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
928 any timeout set, is taking a long time. Previously we would hang on to a
929 rotated logfile "forever" if the input was arriving with long gaps
930 (a previous attempt to fix addressed lack, for a long time, of initial
933 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
934 shared (NFS) environment. The length of the tempfile name is now
935 4 + 16 ("hdr.$message_exim_id") which might break on file
936 systems which restrict the file name length to lower values.
939 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
940 shared (NFS) environment.
942 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
943 did for all versions <4.90). Notably -M, -m, --invert, -I may be
946 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
947 on some platforms for bit 31.
949 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
950 to changes apparently associated with TLS1.3 handling some of the APIs
951 previously used were either nonfunctional or inappropriate. Strings
952 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
953 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
954 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
955 This affects log line X= elements, the $tls_{in,out}_cipher variables,
956 and the use of specific cipher names in the encrypted= ACL condition.
958 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
960 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
961 verification result was not updated unless hosts_require_ocsp applied.
963 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
964 queue_list_requires_admin set to false, non-admin users were denied the
967 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
968 directory-of-certs mode. Previously they were advertised despite the
971 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
972 A single TCP connection by a client will now hold a TLS connection open
973 for multiple message deliveries, by default. Previously the default was to
976 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
977 default. If built with the facility, DANE will be used. The facility
978 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
980 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
981 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
982 must be defined and you must still, unless you define DISABLE_TLS, manage
983 the the include-dir and library-file requirements that go with that
984 choice. Non-TLS builds are still supported.
986 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
989 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
990 default. If the platform supports and has the facility enabled, it will
991 be requested on all coneections.
993 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
994 controlled by the build-time option SUPPORT_PIPE_CONNECT.
996 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
998 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
999 success-DSN messages. Previously the From: header was always the default
1000 one for these; the option was ignored.
1002 JH/28 Fix the timeout on smtp response to apply to the whole response.
1003 Previously it was reset for every read, so a teergrubing peer sending
1004 single bytes within the time limit could extend the connection for a
1005 long time. Credit to Qualsys Security Advisory Team for the discovery.
1007 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1008 delivery address, which leaked information of the results of local
1009 forwarding. Change to the original envelope recipient address, per
1012 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1013 requested. Previously not bounce was generated and a log entry of
1014 error ignored was made.
1016 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1018 JH/32 Introduce a general tainting mechanism for values read from the input
1019 channel, and values derived from them. Refuse to expand any tainted
1020 values, to catch one form of exploit.
1022 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1023 was unused and the unexpanded text used for the test. Found and
1024 fixed by Ruben Jenster.
1026 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1027 an attempt to use a TLS library read routine dereffed a nul pointer,
1030 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1031 them in our smtp response.
1033 JH/36 Have the general router option retry_use_local_part default to true when
1034 any of the restrictive preconditions are set (to anything). Previously it
1035 was only for check_local user. The change removes one item of manual
1036 configuration which is required for proper retries when a remote router
1037 handles a subset of addresses for a domain.
1039 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1040 link count into consideration.
1042 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1043 caused the extension of big_buffer, the following lines were ignored.
1045 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1046 accordance with RFC 2308. Previously there was no expiry, so a longlived
1047 receive process (eg. due to ACL delays) versus a short SOA value could
1050 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1052 JH/39 Promote DMARC support to mainline.
1054 JH/40 Bug 2452: Add a References: header to DSNs.
1056 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1057 parameters. The relevant library call is documented as "Deprecated: This
1058 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1059 3.6.0, DH parameters are negotiated following RFC7919."
1061 HS/06 Change the default of dnssec_request_domains to "*"
1063 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1064 carried on and emitted a BDAT command, even when PIPELINING was not
1067 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1068 buffer was used for the filename, resulting in a trap when tainted
1069 arguments (eg. $domain) were used.
1071 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1072 recommended to avoid a possible server-load attack. The feature can be
1073 re-enabled via the openssl_options main cofiguration option.
1075 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1076 for version 4.90 - adding a "more data" boolean to the arguments.
1077 Bumped the ABI version number also, this having been missed previously;
1078 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1079 respect of usage of smtp_printf() by either local_scan code or libraries
1080 accessed via the ${dlfunc } expansion item. Both will need coding
1081 adjustment for any calls to smtp_printf() to match the new function
1082 signature; a FALSE value for the new argument is always safe.
1084 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1085 the file-offset (which the Linux syscall does, and exim expects); this
1086 resulted in an indefinite loop.
1088 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1089 to do ARC verification. The Authentication-Results: header line added
1090 by the configuration then had no ARC item.
1092 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1093 an incorrect mode was used for reading the result, resulting in it being
1100 JH/01 Remove code calling the customisable local_scan function, unless a new
1101 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1103 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1104 non-signal-safe functions being used.
1106 JH/03 Bug 2269: When presented with a received message having a stupidly large
1107 number of DKIM-Signature headers, disable DKIM verification to avoid
1108 a resource-consumption attack. The limit is set at twenty.
1110 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1111 report of oldest_pass in ${authres } in consequence, and separate out
1112 some descriptions of reasons for verification fail.
1114 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1115 files in the spool were present and unlocked. A queue-runner could spot
1116 them, resulting in a duplicate delivery. Fix that by doing the unlock
1117 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1118 add more error-checking on spoolfile handling while that code is being
1121 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1122 No known attacks, no CVE, this is defensive hardening.
1124 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1125 a queue-runner could start a delivery while other operations were ongoing.
1126 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1127 Found and investigated by Tim Stewart. Fix by using the open message data
1128 file handle rather than opening another, and not locally closing it (which
1129 releases a lock) for that case, while creating the temporary .eml format
1130 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1132 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1133 $sender_verify_failure/$recipient_verify_failure to "random".
1135 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1138 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1139 Previously this would segfault.
1141 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1144 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1145 like zero, since the resolver should be doing this for us, But we need one
1146 as a CNAME but no MX presence gets the CNAME returned; we need to check
1147 that doesn't point to an MX to declare it "no MX returned" rather than
1148 "error, loop". A new main option is added so the older capability of
1149 following some limited number of chain links is maintained.
1151 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1153 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1154 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1155 now-deprecated earlier definitions used only specified the range up to TLS
1156 1.2 (in the older-version library docs).
1158 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1160 JH/15 Rework TLS client-side context management. Stop using a global, and
1161 explicitly pass a context around. This enables future use of TLS for
1162 connections to service-daemons (eg. malware scanning) while a client smtp
1163 connection is using TLS; with cutthrough connections this is quite likely.
1165 JH/16 Fix ARC verification to do AS checks in reverse order.
1167 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1169 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1170 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1171 Previously the "utf8" would be re-prepended for every additional message.
1173 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1174 Previously thery were accepted, resulting in issues when attempting to
1175 forward messages to a non-supporting MTA.
1177 PP/02 Let -n work with printing macros too, not just options.
1179 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1180 one parent address was copied, and bogus data was used at delivery-logging
1181 time. Either a crash (after delivery) or bogus log data could result.
1182 Discovery and analysis by Tim Stewart.
1184 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1185 Previously if the string ended mid-character, we did not insert the
1186 promised '?' replacement.
1188 PP/04 Documentation: current string operators work on bytes, not codepoints.
1190 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1191 should pack well giving a smaller memory footprint so better caching and
1192 therefore performance. Group the declarations where this can't be done so
1193 that the byte-sized flag variables are not interspersed among pointer
1194 variables, giving a better chance of good packing by the compiler.
1196 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1197 non-null, to avoid issues with sites running BATV. Previously reports were
1198 sent with an empty envelope sender so looked like bounces.
1200 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1201 The ignore_error flag wasn't being returned from the filter subprocess so
1202 was not set for later routers. Investigation and fix by Matthias Kurz.
1204 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1205 and a msg:complete for the whole, when a message is manually removed using
1206 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1208 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1209 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1210 hope that is portable enough.
1212 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1213 requiring a known-CA anchor certificate; make it now rely entirely on the
1214 TLSA as an anchor. Checking the name on the leaf cert against the name
1215 on the A-record for the host is still done for TA (but not for EE mode).
1217 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1218 included in delivery lines for non-proxied connections, when compiled with
1219 SUPPORT_SOCKS and running with proxy logging enabled.
1221 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1222 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1223 move the existing event to fire before the normal logging of message
1224 failure so that custom logging is bracketed by normal logging.
1226 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1227 msg:fail:internal event. Developement by Matthias Kurz.
1229 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1230 far too small for todays use of crypto signatures stored there. Go all
1231 the way to the max DNS message size of 64kB, even though this might be
1232 overmuch for IOT constrained device use.
1234 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1235 copy a string over itself. The library routine is documented as not
1236 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1238 JH/32 For main options check_spool_space and check_inode_space, where the
1239 platform supports 64b integers, support more than the previous 2^31 kB
1240 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1241 the previous G, M, k.
1243 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1244 $authenticated_fail_id variable on authentication failure. Previously
1247 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1248 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1249 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1250 GNUTLS_SEC_PARAM_MEDIUM.
1252 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1253 side. Previously we would continue as if no SNI had been received.
1255 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1256 "<" character, which should be a list-separator specification, we walked
1257 off past the nul-terimation.
1259 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1260 causes) even when the retry time is not yet met. Previously they were
1261 not, meaning that when (say) an account was over-quota and temp-rejecting,
1262 and multiple senders' messages were queued, only one sender would get
1263 notified on each configured delay_warning cycle.
1265 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1267 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1268 cipher-suites, an error can be left on the stack even for a succeeding
1269 accept; this results in impossible error messages when a later operation
1272 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1273 return error codes indicating retry. Under TLS1.3 this becomes required.
1275 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1276 it only wrote the new authenticators, resulting in a lack of tracking of
1277 peer changes of ESMTP extensions until the next cache flush.
1279 JH/41 Fix the loop reading a message header line to check for integer overflow,
1280 and more-often against header_maxsize. Previously a crafted message could
1281 induce a crash of the recive process; now the message is cleanly rejected.
1283 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1284 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1290 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1291 When redis_servers is set to a list of > 1 element, and the Redis servers
1292 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1293 case of MOVED into a DEFER case instead, thus moving the query onto the
1294 next server in the list. For a cluster of N elements, all N servers must
1295 be defined in redis_servers.
1297 GF/02 Catch and remove uninitialized value warning in exiqsumm
1298 Check for existence of @ARGV before looking at $ARGV[0]
1300 JH/01 Replace the store_release() internal interface with store_newblock(),
1301 which internalises the check required to safely use the old one, plus
1302 the allocate and data copy operations duplicated in both (!) of the
1303 extant use locations.
1305 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1306 modifier. This matches the restriction on the commandline.
1308 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1309 Previously only the last row was returned.
1311 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1312 we assumed that tags in the header were well-formed, and parsed the
1313 element content after inspecting only the first char of the tag.
1314 Assumptions at that stage could crash the receive process on malformed
1317 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1318 While running the DKIM ACL we operate on the Permanent memory pool so that
1319 variables created with "set" persist to the DATA ACL. Also (at any time)
1320 DNS lookups that fail create cache records using the Permanent pool. But
1321 expansions release any allocations made on the current pool - so a dnsdb
1322 lookup expansion done in the DKIM ACL releases the memory used for the
1323 DNS negative-cache, and bad things result. Solution is to switch to the
1324 Main pool for expansions.
1325 While we're in that code, add checks on the DNS cache during store_reset,
1326 active in the testsuite.
1327 Problem spotted, and debugging aided, by Wolfgang Breyha.
1329 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1330 When none of the hosts presented to a transport match an already-open
1331 connection, close it and proceed with the list. Previously we would
1332 queue the message. Spotted by Lena with Yahoo, probably involving
1335 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1336 Previously a spurious "250 OK id=" response was appended to the proper
1339 JH/08 The "support for" informational output now, which built with Content
1340 Scanning support, has a line for the malware scanner interfaces compiled
1341 in. Interface can be individually included or not at build time.
1343 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1344 by the template makefile "src/EDITME". The "STREAM" support for an older
1345 ClamAV interface method is removed.
1347 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1348 rows affected is given instead).
1350 JH/11 The runtime Berkeley DB library version is now additionally output by
1351 "exim -d -bV". Previously only the compile-time version was shown.
1353 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1354 SMTP connection. Previously, when one had more recipients than the
1355 first, an abortive onward connection was made. Move to full support for
1356 multiple onward connections in sequence, handling cutthrough connection
1357 for all multi-message initiating connections.
1359 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1360 routers. Previously, a multi-recipient message would fail to match the
1361 onward-connection opened for the first recipient, and cause its closure.
1363 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1364 a timeout on read on a GnuTLS initiating connection, resulting in the
1365 initiating connection being dropped. This mattered most when the callout
1366 was marked defer_ok. Fix to keep the two timeout-detection methods
1369 JH/15 Relax results from ACL control request to enable cutthrough, in
1370 unsupported situations, from error to silently (except under debug)
1371 ignoring. This covers use with PRDR, frozen messages, queue-only and
1374 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1376 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1377 metadata, resulting in a crash in free().
1379 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1380 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1381 Broken also in d185889f4, with init system revamp.
1383 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1384 Previously we abruptly closed the connection after reading a malware-
1385 found indication; now we go on to read the "scan ok" response line,
1388 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1389 ACL. Previously, a crash would result.
1391 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1392 macro text after a replacement (previously it was only once per line) and
1393 by skipping builtin macros when searching for an uppercase lead character.
1395 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1396 for the build is renamed.
1398 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1399 was allocated for every new TLS startup, meaning one per message. Fix
1400 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1402 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1403 reported the original. Fix to report (as far as possible) the ACL
1404 result replacing the original.
1406 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1407 OpenSSL. Certificate information is loaded for every new TLS startup,
1408 and the resources needed to be freed.
1410 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1412 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1415 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1416 DATA response info to the (existing) per-recipient response info for
1417 the "C=" log element. It can have useful tracking info from the
1418 destination system. Patch from Simon Arlott.
1420 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1421 length value. Previously this would segfault.
1423 HS/02 Support Avast multiline protoocol, this allows passing flags to
1424 newer versions of the scanner.
1426 JH/28 Ensure that variables possibly set during message acceptance are marked
1427 dead before release of memory in the daemon loop. This stops complaints
1428 about them when the debug_store option is enabled. Discovered specifically
1429 for sender_rate_period, but applies to a whole set of variables.
1430 Do the same for the queue-runner and queue-list loops, for variables set
1431 from spool message files. Do the same for the SMTP per-message loop, for
1432 certain variables indirectly set in ACL operations.
1434 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1435 as a multi-recipient message from a mailinglist manager). The coding had
1436 an arbitrary cutoff number of characters while checking for more input;
1437 enforced by writing a NUL into the buffer. This corrupted long / fast
1438 input. The problem was exposed more widely when more pipelineing of SMTP
1439 responses was introduced, and one Exim system was feeding another.
1440 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1441 receiving system, and refused recipients seen by the sending system
1442 (propating to people being dropped from mailing lists).
1443 Discovered and pinpointed by David Carter.
1445 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1446 replaced by the ${authresults } expansion.
1448 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1450 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1451 allows proper process termination in container environments.
1453 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1454 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1456 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1457 and "err_perm", deprecated since 4.83 when the RFC-defined words
1458 "temperror" and "permerror" were introduced.
1460 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1461 transport-filters or DKIM-signing. The restriction was lost in the
1462 consolidation of verify-callout and delivery SMTP handling.
1463 Extend the restriction to also cover ARC-signing.
1465 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1466 in defer=pass mode supply a 450 to the initiator. Previously the message
1469 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1470 tls_require_ciphers is used as before.
1472 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1473 "pass_unscanned". Only tmpfails from the scanner are written to
1474 the paniclog, as they may require admin intervention (permission
1475 denied, license issues). Other scanner errors (like decompression
1476 bombs) do not cause a paniclog entry.
1478 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1479 Previously it was possible to log spurious information in receive log
1482 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1483 triggered odd behaviour from Outlook Express clients.
1485 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1488 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1489 since the IETF WG has not yet settled on that versus the original
1490 "bare" representation.
1492 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1493 Previously the millisecond value corrupted the output.
1494 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1495 corrupted the output.
1501 JH/01 Rework error string handling in TLS interface so that the caller in
1502 more cases is responsible for logging. This permits library-sourced
1503 string to be attached to addresses during delivery, and collapses
1504 pairs of long lines into single ones.
1506 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1507 during configuration. Wildcards are allowed and expanded.
1509 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1510 This permits better logging.
1512 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1513 do not close it down and have the child transport start it up again on
1514 the passed-on TCP connection. Instead, proxy the child (and any
1515 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1516 affected: the continued delivery log lines do not have any DNSSEC, TLS
1517 Certificate or OCSP information. TLS cipher information is still logged.
1519 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1520 identical IP addresses on different listening ports. Will also affect
1523 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1524 add noisy ifdef guards to special-case this sillyness.
1525 Patch from Bernd Kuhls.
1527 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1528 than 255 are no longer allowed.
1530 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1531 Disable the session-cache too, which might reduce our load. Since we
1532 currrectly use a new context for every connection, both as server and
1533 client, there is no benefit for these.
1534 GnuTLS appears to not support tickets server-side by default (we don't
1535 call gnutls_session_ticket_enable_server()) but client side is enabled
1536 by default on recent versions (3.1.3 +) unless the PFS priority string
1539 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1540 <https://reproducible-builds.org/specs/source-date-epoch/>.
1542 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1543 the check for any unsuccessful recipients did not notice the limit, and
1544 erroneously found still-pending ones.
1546 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1547 MSG_MORE. Only in-clear (not on TLS connections).
1549 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1550 transport-filter is involved we can buffer the headers in memory for
1551 creating the signature, and read the spool data file once for the
1552 signature and again for transmission.
1554 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1555 4.77 as the kernel support then wasn't solid, having issues in 64bit
1556 mode. Now, it's been long enough. Add support for FreeBSD also.
1558 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1559 case where the routing stage had gathered several addresses to send to
1560 a host before calling the transport for the first, we previously failed
1561 to close down TLS in the old transport process before passing the TCP
1562 connection to the new process. The new one sent a STARTTLS command
1563 which naturally failed, giving a failed delivery and bloating the retry
1564 database. Investigation and fix prototype from Wolfgang Breyha.
1566 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1567 false-negatives in the check that the sender had not preempted a response
1568 or prompt from Exim (running as a server), due to that code's lack of
1569 awareness of the SMTP input buffering.
1571 PP/04 Add commandline_checks_require_admin option.
1572 Exim drops privileges sanely, various checks such as -be aren't a
1573 security problem, as long as you trust local users with access to their
1574 own account. When invoked by services which pass untrusted data to
1575 Exim, this might be an issue. Set this option in main configuration
1576 AND make fixes to the calling application, such as using `--` to stop
1579 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1580 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1581 responses to those, into a single TLS record each way (this usually means
1582 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1585 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1586 affects you only if you're dancing at the edge of the param size limits.
1587 If you are, and this message makes sense to you, then: raise the
1588 configured limit or use OpenSSL 1.1. Nothing we can do for older
1591 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1592 cmdline element to get the documented default one. Previously it was
1595 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1596 are unsuable for non-message-specific reasons - eg. connection timeout,
1597 banner-time rejection.
1599 JH/16 Fix logging of delivery remote port, when specified by router, under
1602 PP/06 Repair manualroute's ability to take options in any order, even if one
1603 is the name of a transport.
1606 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1608 JH/17 Change the list-building routines interface to use the expanding-string
1609 triplet model, for better allocation and copying behaviour.
1611 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1612 Previously it was constructed the first time a possibly-matching string
1613 was met in the configuration file input during startup; now it is done
1616 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1617 compatible one, to avoid the (poorly documented) possibility of a config
1618 file in the working directory redirecting the DB files, possibly correpting
1619 some existing file. CVE-2017-10140 assigned for BDB.
1621 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1622 cache-hot. Previously, although the result was properly cached, the
1623 initial verify call returned a defer.
1625 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1626 the main verify for receipient in uncached-mode.
1628 JH/22 Retire historical build files to an "unsupported" subdir. These are
1629 defined as "ones for which we have no current evidence of testing".
1631 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1632 if present. Previously it was ignored.
1634 JH/24 Start using specified-initialisers in C structure init coding. This is
1635 a C99 feature (it's 2017, so now considered safe).
1637 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1638 if was a fixed-sized field and bitmask ops via macros; it is now more
1641 PP/07 GitHub PR 56: Apply MariaDB build fix.
1642 Patch provided by Jaroslav Škarvada.
1644 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1645 during Coverity cleanups [4.87 JH/47]
1646 Diagnosis and fix provided by Michael Fischer v. Mollard.
1648 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1649 the right size to place the terminating semicolon on its own folded
1650 line, the header hash was calculated to an incorrect value thanks to
1651 the (relaxed) space the fold became.
1653 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1654 and confused the parent.
1656 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1657 which could crash as a result. This could lead to undeliverable messages.
1659 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1662 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1663 matters on fast-turnover and PID-randomising systems, which were getting
1664 out-of-order delivery.
1666 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1667 a possibly-overlapping copy. The symptom was that "Remote host closed
1668 connection in response to HELO" was logged instead of the actual 4xx
1671 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1672 Previously only that bufferd was discarded, resulting in SYMTP command
1675 JH/32 DKIM: when a message has multiple signatures matching an identity given
1676 in dkim_verify_signers, run the dkim acl once for each. Previously only
1677 one run was done. Bug 2189.
1679 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1680 "panic the current process" to "deliberately defer". The panic log is
1681 still written with the problem list name; the mail and reject logs now
1682 get a temp-reject line for the message that was being handled, saying
1683 something like "domains check lookup or other defer". The SMTP 451
1684 message is still "Temporary local problem".
1686 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1687 A crafted sequence of BDAT commands could result in in-use memory beeing
1688 freed. CVE-2017-16943.
1690 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1691 from SMTP input. Previously it was always done; now only done for DATA
1692 and not BDAT commands. CVE-2017-16944.
1694 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1695 to the message (such as an overlong header line). Previously this was
1696 not done and we did not exit BDAT mode. Followon from the previous item
1697 though a different problem.
1703 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1704 than -2003 did; needs libidn2 in addition to libidn.
1706 JH/02 The path option on a pipe transport is now expanded before use.
1708 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1709 Patch provided by "Björn", documentation fix added too.
1711 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1712 missing a wire-to-host endian conversion.
1714 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1715 close after a BDAT command line could be taken as a following command,
1716 giving a synch failure. Fix by only checking for synch immediately
1717 before acknowledging the chunk.
1719 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1720 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1721 macro. Patches provided by Josh Soref.
1723 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1724 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1725 by IANA as a service extension. However, John Klensin suggests that we
1728 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1729 the dkim code may be unix-mode line endings rather than smtp wire-format
1730 CRLF, so prepend a CR to any bare LF.
1732 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1733 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1735 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1736 the first were themselves being wrongly included in the feed into dkim
1737 processing; with most chunk sizes in use this resulted in an incorrect
1738 body hash calculated value.
1740 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1741 DKIM signature block, for verification. Although advised against by
1742 standards it is specifically not ruled illegal.
1744 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1746 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1747 missing a body hash (the bh= tag).
1749 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1750 It seems that HAProxy sends the Proxy Protocol information in clear and
1751 only then does a TLS startup, so do the same.
1753 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1754 TCP connections (such as for Spamd) unless the daemon successfully set
1755 Fast Open mode on its listening sockets. This fixes breakage seen on
1756 too-old kernels or those not configured for Fast Open, at the cost of
1757 requiring both directions being enabled for TFO, and TFO never being used
1758 by non-daemon-related Exim processes.
1760 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1761 endings, at least on the first header line. Try to canonify any that get
1762 past that check, despite the cost.
1764 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1765 now limited to an arbitrary five deep, while parsing addresses with the
1766 strip_excess_angle_brackets option enabled.
1768 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1769 instead leave the unprompted TLS handshake in socket buffer for the
1770 TLS library to consume.
1772 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1774 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1776 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1777 region in one operation, for speed, and this leaves assigned pointers
1778 dangling. Add checks run only under the testsuite which checks all
1779 variables at a store-reset and panics on a dangling pointer; add code
1780 explicitly nulling out all the variables discovered. Fixes one known
1781 bug: a transport crash, where a dangling pointer for $sending_ip_address
1782 originally assigned in a verify callout, is re-used.
1784 PP/06 Drop '.' from @INC in various Perl scripts.
1786 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1788 PP/08 Reduce a number of compilation warnings under clang; building with
1789 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1790 should be warning-free.
1792 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1794 HS/01 Fix portability problems introduced by PP/08 for platforms where
1795 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1797 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1798 chunk. This allows us to accept broken chunked messages. We need a more
1799 general solution here.
1801 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1802 already-broken messages in the queue.
1804 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1806 JH/19 Fix reference counting bug in routing-generated-address tracking.
1812 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1813 supports it and a size is available (ie. the sending peer gave us one).
1815 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1816 years of being deprecated). The replacements are the ACLs
1817 acl_smtp_mime and acl_not_smtp_mime.
1819 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1820 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1821 or even an in-clear connection were permitted. Now, if the host lookup
1822 was dnssec and dane was requested then the host is only used if the
1823 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1824 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1825 if one fails this test.
1826 This means that a poorly-configured remote DNS will make it incommunicado;
1827 but it protects against a DNS-interception attack on it.
1829 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1830 non-noisy when a race steals the message being considered.
1832 JH/05 If main configuration option tls_certificate is unset, generate a
1833 self-signed certificate for inbound TLS connections.
1835 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1836 in rewrites and routers.
1838 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1839 and logged a warning sing 4.83; now they are a configuration file error.
1841 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1842 (lacking @domain). Apply the same qualification processing as RCPT.
1844 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1846 JH/10 Support ${sha256:} applied to a string (as well as the previous
1849 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1850 a cutthrough deliver is pending, as we always want to make a connection.
1851 This also avoids re-routing the message when later placing the cutthrough
1852 connection after a verify cache hit.
1853 Do not update it with the verify result either.
1855 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1856 when routing results in more than one destination address.
1858 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1859 signing (which inhibits the cutthrough capability). Previously only
1860 the presence of an option was tested; now an expansion evaluating as
1861 empty is permissible (obviously it should depend only on data available
1862 when the cutthrough connection is made).
1864 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1865 the relevant preceding SMTP command did not note the pipelining mode.
1867 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1868 Previously they were not counted.
1870 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1871 as one having no matching records. Previously we deferred the message
1872 that needed the lookup.
1874 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1875 distinguished as "(=".
1877 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1878 for missing MX records. Previously it only worked for missing A records.
1880 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1882 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1883 after the data-go-ahead and data-ack. Patch from Jason Betts.
1885 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1886 even for a "none" policy. Patch from Tony Meyer.
1888 JH/22 Fix continued use of a connection for further deliveries. If a port was
1889 specified by a router, it must also match for the delivery to be
1892 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1893 When one of the recipients of a message was unsuitable for the connection
1894 (has no matching addresses), we lost track of needing to mark it
1895 deferred. As a result mail would be lost.
1897 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1899 JH/25 Decoding ACL controls is now done using a binary search; the source code
1900 takes up less space and should be simpler to maintain. Merge the ACL
1901 condition decode tables also, with similar effect.
1903 JH/26 Fix problem with one_time used on a redirect router which returned the
1904 parent address unchanged. A retry would see the parent address marked as
1905 delivered, so not attempt the (identical) child. As a result mail would
1908 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1909 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1910 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1913 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1914 Main config options check_{log,spool}_{inodes,space} are now
1915 100 inodes, 10MB unless set otherwise in the configuration.
1917 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1918 Previously it only applied to the main-section connection policy
1921 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1923 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1924 by me. Added RFC7919 DH primes as an alternative.
1926 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1927 are not in the system include path.
1929 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1930 GnuTLS, when a session startup failed (eg because the client disconnected)
1931 Exim did stdio operations after fclose. This was exposed by a recent
1932 change which nulled out the file handle after the fclose.
1934 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1935 signed directly by the cert-signing cert, rather than an intermediate
1936 OCSP-signing cert. This is the model used by LetsEncrypt.
1938 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1940 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1941 an incoming connection.
1943 HS/02 Bug 1802: Do not half-close the connection after sending a request
1946 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1947 fallback to "prime256v1".
1949 JH/34 SECURITY: Use proper copy of DATA command in error message.
1950 Could leak key material. Remotely exploitable. CVE-2016-9963.
1956 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1957 and 3.4.4 - once the server is enabled to respond to an OCSP request
1958 it does even when not requested, resulting in a stapling non-aware
1959 client dropping the TLS connection.
1961 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1962 support variable-length bit vectors. No functional change.
1964 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1965 The I= interface field on outgoing lines is now after the H= remote
1966 host field, same as incoming lines. There is a separate
1967 outgoing_interface log selector which allows you to disable the
1970 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1971 If not running log_selector +smtp_connection the mainlog would be held
1972 open indefinitely after a "too many connections" event, including to a
1973 deleted file after a log rotate. Leave the per net connection logging
1974 leaving it open for efficiency as that will be quickly detected by the
1975 check on the next write.
1977 HS/01 Bug 1671: Fix post transport crash.
1978 Processing the wait-<transport> messages could crash the delivery
1979 process if the message IDs didn't exist for some reason. When
1980 using 'split_spool_directory=yes' the construction of the spool
1981 file name failed already, exposing the same netto behaviour.
1983 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1984 mime_regex ACL conditions.
1986 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1987 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1988 to HELO, local diagnostic string.
1990 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
1991 log line to debug. Even when configured with a tls authenticator many
1992 client connections are expected to not authenticate in this way, so
1993 an authenticate fail is not an error.
1995 HS/02 Add the Exim version string to the process info. This way exiwhat
1996 gives some more detail about the running daemon.
1998 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
1999 matter for fast-change records such as DNSBLs.
2001 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2002 retry record, even if constant. There may be multiple transports with
2003 different interface settings and the retry behaviour needs to be kept
2006 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2008 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2010 JH/10 Bug 840: fix log_defer_output option of pipe transport
2012 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2013 hosts_randomize. This matters a lot when combined with mua_wrapper.
2015 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2016 ${quote_pgsql:<string>} operator.
2018 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2020 JH/14 Change header file name for freeradius-client. Relevant if compiling
2021 with Radius support; from the Gentoo tree and checked under Fedora.
2023 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2025 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2026 option transport_filter, meaning no filtering.
2028 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2030 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2031 defaults to "*" (all hosts). The variable is now available when not built
2032 with TLS, default unset, mainly to enable keeping the testsuite sane.
2033 If a server certificate is not supplied (via tls_certificate) an error is
2034 logged, and clients will find TLS connections fail on startup. Presumably
2035 they will retry in-clear.
2036 Packagers of Exim are strongly encouraged to create a server certificate
2037 at installation time.
2039 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2040 with the $config_file variable.
2042 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2043 in transport context, after the attempt, and per-recipient. The latter type
2044 is per host attempted. The event data is the error message, and the errno
2045 information encodes the lookup type (A vs. MX) used for the (first) host,
2046 and the trailing two digits of the smtp 4xx response.
2048 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2049 to write to mainlog (or rejectlog, paniclog) in the window between file
2050 creation and permissions/ownership being changed. Particularly affects
2051 installations where exicyclog is run as root, rather than exim user;
2052 result is that the running daemon panics and dies.
2054 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2056 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2057 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2058 "pri" and "weight". Note that the previous implicit priority given by the
2059 list order is no longer honoured.
2061 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2062 for DKIM processing.
2064 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2065 by defining SUPPORT_SOCKS.
2067 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2068 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2069 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2070 variables are renamed to proxy_{local,external}_{address,port}.
2072 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2073 for a build by defining SUPPORT_I18N
2075 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2076 of the query string, and make ${quote_redis:} do that quoting.
2078 JH/29 Move Events support from Experimental to mainline, enabled by default
2079 and removable for a build by defining DISABLE_EVENT.
2081 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2083 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2084 cached by the daemon.
2086 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2087 by defining LOOKUP_REDIS. The libhiredis library is required.
2089 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2090 keys are given for lookup.
2092 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2093 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2094 only supported when built with TLS support. The PolarSSL SHA routines
2095 are still used when the TLS library is too old for convenient support.
2097 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2098 openssl_options), for security. OpenSSL forces this from version 1.1.0
2099 server-side so match that on older versions.
2101 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2102 allocation for $value could be released as the expansion processing
2103 concluded, but leaving the global pointer active for it.
2105 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2106 and to use the domains and local_parts ACL conditions.
2108 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2109 incorrectly not doubled on cutthrough transmission, hence seen as a
2110 body-termination at the receiving system - resulting in truncated mails.
2111 Commonly the sender saw a TCP-level error, and retransmitted the message
2112 via the normal store-and-forward channel. This could result in duplicates
2113 received - but deduplicating mailstores were liable to retain only the
2114 initial truncated version.
2116 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2118 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2120 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2121 we're in there, support oversigning also; bug 1309.
2123 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2125 HS/04 Add support for keep_environment and add_environment options.
2127 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2128 either intentional arithmetic overflow during PRNG, or testing config-
2131 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2132 delivery resulted in actual delivery. Cancel cutthrough before DATA
2135 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2136 rejects a recipient: pass the reject to the originator.
2138 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2139 Many were false-positives and ignorable, but it's worth fixing the
2142 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2143 for the new environment-manipulation done at startup. Move the routines
2144 from being local to tls.c to being global via the os.c file.
2146 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2147 an extract embedded as result-arg for a map, the first arg for extract
2148 is unavailable so we cannot tell if this is a numbered or keyed
2149 extraction. Accept either.
2155 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2158 JH/02 The smtp transport option "multi_domain" is now expanded.
2160 JH/03 The smtp transport now requests PRDR by default, if the server offers
2163 JH/04 Certificate name checking on server certificates, when exim is a client,
2164 is now done by default. The transport option tls_verify_cert_hostnames
2165 can be used to disable this per-host. The build option
2166 EXPERIMENTAL_CERTNAMES is withdrawn.
2168 JH/05 The value of the tls_verify_certificates smtp transport and main options
2169 default to the word "system" to access the system default CA bundle.
2170 For GnuTLS, only version 3.0.20 or later.
2172 JH/06 Verification of the server certificate for a TLS connection is now tried
2173 (but not required) by default. The verification status is now logged by
2174 default, for both outbound TLS and client-certificate supplying inbound
2177 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2180 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2181 Status Notification (bounce) messages are now MIME format per RFC 3464.
2182 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2183 under the control of the dsn_advertise_hosts option, and routers may
2184 have a dsn_lasthop option.
2186 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2187 default, modifiable by a malware= option. The list separator for
2188 the options can now be changed in the usual way. Bug 68.
2190 JH/10 The smtp_receive_timeout main option is now expanded before use.
2192 JH/11 The incoming_interface log option now also enables logging of the
2193 local interface on delivery outgoing connections.
2195 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2196 if the interface and destination host and port all match.
2198 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2201 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2202 Patch from Andrew Lewis.
2204 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2205 now supports optional time-restrictions, weighting, and priority
2206 modifiers per server. Patch originally by <rommer@active.by>.
2208 JH/16 The spamd_address main option now supports a mixed list of local
2209 and remote servers. Remote servers can be IPv6 addresses, and
2210 specify a port-range.
2212 JH/17 Bug 68: The spamd_address main option now supports an optional
2213 timeout value per server.
2215 JH/18 Bug 1581: Router and transport options headers_add/remove can
2216 now have the list separator specified.
2218 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2221 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2224 JH/21 Support for the A6 type of dns record is withdrawn.
2226 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2227 rather than the verbs used.
2229 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2230 from 255 to 1024 chars.
2232 JH/24 Verification callouts now attempt to use TLS by default.
2234 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2235 are generic router options now. The defaults didn't change.
2237 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2238 Original patch from Alexander Shikoff, worked over by JH.
2240 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2241 files if ZCAT_COMMAND is not executable.
2243 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2245 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2247 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2248 Normally benign, it bites when the pair was led to by a CNAME;
2249 modern usage is to not canonicalize the domain to a CNAME target
2250 (and we were inconsistent anyway for A-only vs AAAA+A).
2252 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2254 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2255 when evaluating $sender_host_dnssec.
2257 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2258 $sender_helo_dnssec variable.
2260 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2262 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2264 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2266 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2267 documented as working, but never had. Support all but $spam_report.
2269 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2270 added for tls authenticator.
2272 HS/03 Add perl_taintmode main config option
2278 TL/01 When running the test suite, the README says that variables such as
2279 no_msglog_check are global and can be placed anywhere in a specific
2280 test's script, however it was observed that placement needed to be near
2281 the beginning for it to behave that way. Changed the runtest perl
2282 script to read through the entire script once to detect and set these
2283 variables, reset to the beginning of the script, and then run through
2284 the script parsing/test process like normal.
2286 TL/02 The BSD's have an arc4random API. One of the functions to induce
2287 adding randomness was arc4random_stir(), but it has been removed in
2288 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2289 function when detected.
2291 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2292 cause callback expansion.
2294 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2295 syntax errors in an expansion can be treated as a string instead of
2296 logging or causing an error, due to the internal use of bool_lax
2297 instead of bool when processing it.
2299 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2300 server certificates when making smtp deliveries.
2302 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2304 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2306 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2308 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2309 Merged patch from Sebastian Wiedenroth.
2311 JH/05 Fix results-pipe from transport process. Several recipients, combined
2312 with certificate use, exposed issues where response data items split
2313 over buffer boundaries were not parsed properly. This eventually
2314 resulted in duplicates being sent. This issue only became common enough
2315 to notice due to the introduction of connection certificate information,
2316 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2318 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2319 size buffer was used, resulting in syntax errors when an expansion
2322 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2323 version 3.3.6 or later.
2325 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2326 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2327 both become "event_action", the variables become $event_name, $event_data
2328 and $event_defer_errno. There is a new variable $verify_mode, usable in
2329 routers, transports and related events. The tls:cert event is now also
2330 raised for inbound connections, if the main configuration event_action
2333 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2334 early OCSP support, but no stapling (appears to be less than 1.0.0).
2336 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2337 server certificate names available under the smtp transport option
2338 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2341 JH/10 Time-related extraction expansions from certificates now use the main
2342 option "timezone" setting for output formatting, and are consistent
2343 between OpenSSL and GnuTLS compilations. Bug 1541.
2345 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2346 encoded parameter in the incoming message. Bug 1558.
2348 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2349 include certificate info, eximon was claiming there were spoolfile
2352 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2354 JH/14 Log delivery-related information more consistently, using the sequence
2355 "H=<name> [<ip>]" wherever possible.
2357 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2358 are problematic for Debian distribution, omit them from the release
2361 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2363 JH/16 Fix string representation of time values on 64bit time_t architectures.
2366 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2367 output list separator was used.
2372 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2373 checkers that were complaining about end of non-void function with no
2376 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2377 This was a regression introduced in 4.83 by another bugfix.
2379 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2381 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2382 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2388 TF/01 Correctly close the server side of TLS when forking for delivery.
2390 When a message was received over SMTP with TLS, Exim failed to clear up
2391 the incoming connection properly after forking off the child process to
2392 deliver the message. In some situations the subsequent outgoing
2393 delivery connection happened to have the same fd number as the incoming
2394 connection previously had. Exim would try to use TLS and fail, logging
2395 a "Bad file descriptor" error.
2397 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2398 utilities have not been installed.
2400 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2401 temporary space as the ACL may create new global variables.
2403 TL/01 LDAP support uses per connection or global context settings, depending
2404 upon the detected version of the libraries at build time.
2406 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2407 to extract and use the src ip:port in logging and expansions as if it
2408 were a direct connection from the outside internet. PPv2 support was
2409 updated based on HAProxy spec change in May 2014.
2411 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2413 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2414 Properly escape header and check for NULL return.
2416 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2419 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2421 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2422 characters in header names, implemented as a verify condition.
2423 Contributed by Michael Fischer v. Mollard.
2425 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2426 results permerror and temperror. Previous values are deprecated but
2427 still accepted. In a future release, err_perm and err_temp will be
2428 completely removed, which will be a backward incompatibility if the
2429 ACL tests for either of these two old results. Patch contributed by
2430 user bes-internal on the mailing list.
2432 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2434 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2435 selectors, in both main and reject logs.
2437 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2440 JH/07 Add malware type "sock" for talking to simple daemon.
2442 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2444 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2445 routers/transports under cutthrough routing.
2447 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2448 numbers. Touch up "bool" conditional to keep the same definition.
2450 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2452 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2454 JH/12 Expand items in router/transport headers_add or headers_remove lists
2455 individually rather than the list as a whole. Bug 1452.
2457 Required for reasonable handling of multiple headers_ options when
2458 they may be empty; requires that headers_remove items with embedded
2459 colons must have them doubled (or the list-separator changed).
2461 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2462 view the policy declared in the DMARC record. Currently, $dmarc_status
2463 is a combined value of both the record presence and the result of the
2466 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2468 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2469 dnslookup router and the smtp transport (applying to the forward
2472 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2473 of ldap servers used for a specific lookup. Patch provided by Heiko
2476 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2477 New variable $lookup_dnssec_authenticated for observability.
2479 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2480 Patch submitted by Lars Timman.
2482 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2484 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2485 Requires trusted mode and valid format message id, aborts otherwise.
2486 Patch contributed by Heiko Schlichting.
2488 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2489 certextract with support for various fields. Bug 1358.
2491 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2492 is requested by default, modifiable by smtp transport option
2495 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2496 operate on certificate variables to give certificate fingerprints
2497 Also new ${sha256:cert_variable}.
2499 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2501 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2502 Christian Aistleitner.
2504 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2506 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2507 file. Patch from Wolfgang Breyha.
2509 JH/25 Expand the coverage of the delivery $host and $host_address to
2510 client authenticators run in verify callout. Bug 1476.
2512 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2513 align with daemon_smtp_ports. Bug 72.
2515 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2516 support and error reporting did not work properly.
2518 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2519 and is readable. Patch from Andrew Colin Kissa.
2521 TL/14 Enhance documentation of ${run expansion and how it parses the
2522 commandline after expansion, particularly in the case when an
2523 unquoted variable expansion results in an empty value.
2525 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2527 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2528 by renaming away from it. Addresses GH issue 15, reported by
2531 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2533 TL/15 SECURITY: prevent double expansion in math comparison functions
2534 (can expand unsanitized data). Not remotely exploitable.
2541 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2543 PP/02 Make -n do something, by making it not do something.
2544 When combined with -bP, the name of an option is not output.
2546 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2549 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2550 $sender_host_name and config options to manage this, and basic check
2553 PP/05 DSCP support for outbound connections and control modifier for inbound.
2555 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2556 (Only plugin which currently uses this is kerberos4, which nobody should
2557 be using, but we should make it available and other future plugins might
2558 conceivably use it, even though it would break NAT; stuff *should* be
2559 using channel bindings instead).
2561 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2562 name; added for Sendmail compatibility; requires admin caller.
2563 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2564 just ignore it); requires trusted caller.
2565 Also parse but ignore: -Ac -Am -X<logfile>
2568 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2570 TL/02 Add +smtp_confirmation as a default logging option.
2572 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2573 Patch by Magnus Holmgren from 2007-02-20.
2575 TL/04 Bugzilla 1281 - Spec typo.
2576 Bugzilla 1283 - Spec typo.
2577 Bugzilla 1290 - Spec grammar fixes.
2579 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2581 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2583 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2584 mailing list by by Dmitry Isaikin.
2586 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2588 JH/02 Support "G" suffix to numbers in ${if comparisons.
2590 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2592 NM/01 Bugzilla 1197 - Spec typo
2593 Bugzilla 1196 - Spec examples corrections
2595 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2597 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2598 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2601 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2602 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2604 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2605 "acl {{name}{arg}...}", and optional args on acl condition
2608 JH/05 Permit multiple router/transport headers_add/remove lines.
2610 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2612 JH/07 Avoid using a waiting database for a single-message-only transport.
2613 Performance patch from Paul Fisher. Bugzilla 1262.
2615 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2618 JH/09 Add $headers_added variable, with content from use of ACL modifier
2619 add_header (but not yet added to the message). Bugzilla 199.
2621 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2622 Pulled from Bugzilla 817 by Wolfgang Breyha.
2624 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2626 (nb: this is the same fix as in Exim 4.80.1)
2628 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2631 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2634 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2636 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2637 advertise SMTP AUTH mechanism to us, instead of a generic
2638 protocol violation error. Also, make Exim more robust to bad
2639 data from the Dovecot auth socket.
2641 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2643 When a queue runner is handling a message, Exim first routes the
2644 recipient addresses, during which it prunes them based on the retry
2645 hints database. After that it attempts to deliver the message to
2646 any remaining recipients. It then updates the hints database using
2649 So if a recipient address works intermittently, it can get repeatedly
2650 deferred at routing time. The retry hints record remains fresh so the
2651 address never reaches the final cutoff time.
2653 This is a fairly common occurrence when a user is bumping up against
2654 their storage quota. Exim had some logic in its local delivery code
2655 to deal with this. However it did not apply to per-recipient defers
2656 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2658 This change adds a proper retry rule check during routing so that the
2659 final cutoff time is checked against the message's age. We only do
2660 this check if there is an address retry record and there is not a
2661 domain retry record; this implies that previous attempts to handle
2662 the address had the retry_use_local_parts option turned on. We use
2663 this as an approximation for the destination being like a local
2664 delivery, as in LMTP.
2666 I suspect this new check makes the old local delivery cutoff check
2667 redundant, but I have not verified this so I left the code in place.
2669 TF/02 Correct gecos expansion when From: is a prefix of the username.
2671 Test 0254 submits a message to Exim with the header
2675 When I ran the test suite under the user fanf2, Exim expanded
2676 the header to contain my full name, whereas it should have added
2677 a Resent-Sender: header. It erroneously treats any prefix of the
2678 username as equal to the username.
2680 This change corrects that bug.
2682 GF/01 DCC debug and logging tidyup
2683 Error conditions log to paniclog rather than rejectlog.
2684 Debug lines prefixed by "DCC: " to remove any ambiguity.
2686 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2688 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2689 Bug spotted by Jeremy Harris; was flawed since initial commit.
2690 Would have resulted in OCSP responses post-SNI triggering an Exim
2691 NULL dereference and crash.
2693 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2695 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2696 Bug detection, analysis and fix by Samuel Thibault.
2697 Bugzilla 1331, Debian bug #698092.
2699 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2701 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2702 Server implementation by Todd Lyons, client by JH.
2703 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2704 config variable "prdr_enable" controls whether the server
2705 advertises the facility. If the client requests PRDR a new
2706 acl_data_smtp_prdr ACL is called once for each recipient, after
2707 the body content is received and before the acl_smtp_data ACL.
2708 The client is controlled by both of: a hosts_try_prdr option
2709 on the smtp transport, and the server advertisement.
2710 Default client logging of deliveries and rejections involving
2711 PRDR are flagged with the string "PRDR".
2713 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2714 fclose(). Diagnosis by Todd Lyons.
2716 PP/17 Update configure.default to handle IPv6 localhost better.
2717 Patch by Alain Williams (plus minor tweaks).
2720 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2721 This is now consistent with GnuTLS, and is now documented: the
2722 previous undocumented portable approach to treating the option as
2723 unset was to force an expansion failure. That still works, and
2724 an empty string is now equivalent.
2726 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2727 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2728 not performing validation itself.
2730 PP/20 Added force_command boolean option to pipe transport.
2731 Patch from Nick Koston, of cPanel Inc.
2733 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2736 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2738 PP/21 Fix eximon continuous updating with timestamped log-files.
2739 Broken in a format-string cleanup in 4.80, missed when I repaired the
2740 other false fix of the same issue.
2741 Report and fix from Heiko Schlichting.
2744 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2745 Report from Prashanth Katuri.
2747 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2748 It's SecureTransport, so affects any MacOS clients which use the
2749 system-integrated TLS libraries, including email clients.
2751 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2752 using a MIME ACL for non-SMTP local injection.
2753 Report and assistance in diagnosis by Warren Baker.
2755 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2757 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2759 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2760 last id that failed so it may be referenced in subsequent ACL's.
2762 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2765 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2766 ldap library initialization, allowing self-signed CA's to be
2767 used. Also properly sets require_cert option later in code by
2768 using NULL (global ldap config) instead of ldap handle (per
2769 session). Bug diagnosis and testing by alxgomz.
2771 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2772 the src/util/ subdirectory.
2774 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2775 renamed to Transport Post Delivery Action by Jeremy Harris, as
2778 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2779 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2780 redis_servers = needs to be configured which will be used by the redis
2781 lookup. Patch from Warren Baker, of The Packet Hub.
2783 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2785 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2786 hostname or reverse DNS when processing a host list. Used suggestions
2787 from multiple comments on this bug.
2789 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2791 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2792 Missed a few lines, added it to make the runtest require no keyboard
2795 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2796 contains upper case chars. Make router use caseful_local_part.
2798 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2799 support when GnuTLS has been built with p11-kit.
2805 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2807 This, or similar/improved, will also be change PP/11 of 4.82.
2813 PP/01 Handle short writes when writing local log-files.
2814 In practice, only affects FreeBSD (8 onwards).
2815 Bugzilla 1053, with thanks to Dmitry Isaikin.
2817 NM/01 Bugzilla 949 - Documentation tweak
2819 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2822 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2824 PP/02 Implemented gsasl authenticator.
2826 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2828 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2829 `pkg-config foo` for cflags/libs.
2831 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2832 with rest of GSASL and with heimdal_gssapi.
2834 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2835 `pkg-config foo` for cflags/libs for the TLS implementation.
2837 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2838 properties get this fed in as external SSF. A number of robustness
2839 and debugging improvements to the cyrus_sasl authenticator.
2841 PP/08 cyrus_sasl server now expands the server_realm option.
2843 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2844 Patch by Jeremy Harris.
2846 PP/10 Added dbmjz lookup type.
2848 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2850 PP/12 MAIL args handles TAB as well as SP, for better interop with
2851 non-compliant senders.
2852 Analysis and variant patch by Todd Lyons.
2854 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2855 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2856 Patch from Dirk Mueller <dmueller@suse.com>
2858 PP/13 tls_peerdn now print-escaped for spool files.
2859 Observed some $tls_peerdn in wild which contained \n, which resulted
2860 in spool file corruption.
2862 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2863 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2864 or write after TLS renegotiation, which otherwise led to messages
2867 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2868 as a tracking header (ie: a signed header comes before the signature).
2869 Patch from Wolfgang Breyha.
2871 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2872 comma-sep list; embedded commas doubled.
2874 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2876 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2878 Report and patch from Dmitry Banschikov.
2880 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2881 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2882 failures. We appear to now support TLS1.1+ with Exim.
2884 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2885 lets Exim select keys and certificates based upon TLS SNI from client.
2886 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2887 before an outbound SMTP session. New log_selector, +tls_sni.
2889 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2890 NULL dereference. Report and patch from Alun Jones.
2892 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2893 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2894 Not seeing resolver debug output on NetBSD, but suspect this is a
2895 resolver implementation change.
2897 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2898 Left warnings. Added "eximon gdb" invocation mode.
2900 PP/21 Defaulting "accept_8bitmime" to true, not false.
2902 PP/22 Added -bw for inetd wait mode support.
2904 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2905 locate the relevant includes and libraries. Made this the default.
2907 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2908 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2910 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2911 This may cause build issues on older platforms.
2913 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2914 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2915 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2916 Added SNI support via GnuTLS too.
2917 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2919 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2921 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2922 Applied second patch from Janne, implementing suggestion to default
2923 multiple-strings-in-record handling to match SPF spec.
2925 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2927 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2928 read-only, out of scope).
2929 Patch from Wolfgang Breyha, report from Stuart Northfield.
2931 PP/29 Fix three issues highlighted by clang analyser static analysis.
2932 Only crash-plausible issue would require the Cambridge-specific
2933 iplookup router and a misconfiguration.
2934 Report from Marcin Mirosław.
2936 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2938 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2939 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2940 As part of this, removing so much warning spew let me fix some minor
2941 real issues in debug logging.
2943 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2944 assignment on my part. Fixed.
2946 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2947 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2948 Janne Snabb (who went above and beyond: thank you).
2950 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2951 string otherwise requires a connection and a bunch more work and it's
2952 relatively easy to get wrong. Should also expose TLS library linkage
2955 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2956 64-bit ${eval} (JH/03).
2958 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2959 GNU libc to support some of the 64-bit stuff, should not lead to
2960 conflicts. Defined before os.h is pulled in, so if a given platform
2961 needs to override this, it can.
2963 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2964 protection layer was required, which is not implemented.
2965 Bugzilla 1254, patch from Wolfgang Breyha.
2967 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2968 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2969 tls_dhparam take prime identifiers. Also unbreak combination of
2970 OpenSSL+DH_params+TLSSNI.
2972 PP/39 Disable SSLv2 by default in OpenSSL support.
2978 PP/01 Solaris build fix for Oracle's LDAP libraries.
2979 Bugzilla 1109, patch from Stephen Usher.
2981 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2983 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2986 TF/02 Fix a couple more cases where we did not log the error message
2987 when unlink() failed. See also change 4.74-TF/03.
2989 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2990 lock up or crash if it happened to be inside a call to libc when it
2991 got a SIGUSR1 from exiwhat.
2993 The SIGUSR1 handler appends the current process status to the process
2994 log which is later printed by exiwhat. It used to use the general
2995 purpose logging code to do this, but several functions it calls are
2996 not safe for signals.
2998 The new output code in the SIGUSR1 handler is specific to the process
2999 log, and simple enough that it's easy to inspect for signal safety.
3000 Removing some special cases also simplifies the general logging code.
3001 Removing the spurious timestamps from the process log simplifies
3004 TF/04 Improved ratelimit ACL condition.
3006 The /noupdate option has been deprecated in favour of /readonly which
3007 has clearer semantics. The /leaky, /strict, and /readonly update modes
3008 are mutually exclusive. The update mode is no longer included in the
3009 database key; it just determines when the database is updated. (This
3010 means that when you upgrade Exim will forget old rate measurements.)
3012 Exim now checks that the per_* options are used with an update mode that
3013 makes sense for the current ACL. For example, when Exim is processing a
3014 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3015 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3016 must specify per_mail/readonly. If you omit the update mode it defaults to
3017 /leaky where that makes sense (as before) or /readonly where required.
3019 The /noupdate option is now undocumented but still supported for
3020 backwards compatibility. It is equivalent to /readonly except that in
3021 ACLs where /readonly is required you may specify /leaky/noupdate or
3022 /strict/noupdate which are treated the same as /readonly.
3024 A useful new feature is the /count= option. This is a generalization
3025 of the per_byte option, so that you can measure the throughput of other
3026 aggregate values. For example, the per_byte option is now equivalent
3027 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3029 The per_rcpt option has been generalized using the /count= mechanism
3030 (though it's more complicated than the per_byte equivalence). When it is
3031 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3032 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3033 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3034 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3035 Note that using per_rcpt with a non-readonly update mode in more than
3036 one ACL will cause the recipients to be double-counted. (The per_mail
3037 and per_byte options don't have this problem.)
3039 The handling of very low rates has changed slightly. If the computed rate
3040 is less than the event's count (usually one) then this event is the first
3041 after a long gap. In this case the rate is set to the same as this event's
3042 count, so that the first message of a spam run is counted properly.
3044 The major new feature is a mechanism for counting the rate of unique
3045 events. The new per_addr option counts the number of different
3046 recipients that someone has sent messages to in the last time period. It
3047 behaves like per_rcpt if all the recipient addresses are different, but
3048 duplicate recipient addresses do not increase the measured rate. Like
3049 the /count= option this is a general mechanism, so the per_addr option
3050 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3051 example, measure the rate that a client uses different sender addresses
3052 with the options per_mail/unique=$sender_address. There are further
3053 details in the main documentation.
3055 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3057 TF/06 Removed a few PCRE remnants.
3059 TF/07 Automatically extract Exim's version number from tags in the git
3060 repository when doing development or release builds.
3062 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3063 Bugzilla 879. Patch from Paul Fisher.
3065 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3066 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3069 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3071 PP/05 Variable $av_failed, true if the AV scanner deferred.
3072 Bugzilla 1078. Patch from John Horne.
3074 PP/06 Stop make process more reliably on build failure.
3075 Bugzilla 1087. Patch from Heiko Schlittermann.
3077 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3078 Bugzilla 1089. Patch from Heiko Schlittermann.
3080 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3081 Bugzilla 1131. Patch from Holger Weiß.
3083 PP/09 Handle IPv6 addresses with SPF.
3084 Bugzilla 860. Patch from Wolfgang Breyha.
3086 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3088 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3091 PP/11 match_* no longer expand right-hand-side by default.
3092 New compile-time build option, EXPAND_LISTMATCH_RHS.
3093 New expansion conditions, "inlist", "inlisti".
3095 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3097 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3099 PP/14 fix log_write() format string regression from TF/03.
3100 Bugzilla 1152. Patch from Dmitry Isaikin.
3106 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3108 PP/02 Harmonised TLS library version reporting; only show if debugging.
3109 Layout now matches that introduced for other libraries in 4.74 PP/03.
3111 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3113 PP/04 New "dns_use_edns0" global option.
3115 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3118 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3119 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3121 TK/01 Updated PolarSSL code to 0.14.2.
3122 Bugzilla 1097. Patch from Andreas Metzler.
3124 PP/07 Catch divide-by-zero in ${eval:...}.
3125 Fixes bugzilla 1102.
3127 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3130 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3131 format-string attack -- SECURITY: remote arbitrary code execution.
3133 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3134 time unintentionally subject to list matching rules, letting the header
3135 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3136 arbitrary string expansion). This allowed for information disclosure.
3138 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3139 INT_MIN/-1 -- value coerced to INT_MAX.
3145 NM/01 Workaround for PCRE version dependency in version reporting
3148 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3149 This fixes portability to compilers other than gcc, notably
3150 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3152 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3153 makefiles for portability to HP-UX and POSIX correctness.
3155 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3156 Also via indented variable definition in the Makefile.
3157 (Debugging by Oliver Heesakkers).
3159 PP/02 Restore caching of spamd results with expanded spamd_address.
3160 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3162 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3163 Improves build reliability. Fix from: Frank Elsner
3165 NM/02 Fix wide character breakage in the rfc2047 coding
3166 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3168 NM/03 Allow underscore in dnslist lookups
3169 Fixes bug 1026. Patch from Graeme Fowler
3171 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3172 Code patches from Adam Ciarcinski of NetBSD.
3174 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3177 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3178 is logged, to avoid truncation. Patch from John Horne.
3180 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3181 Patch from Jakob Hirsch.
3183 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3184 SQL string expansion failure details.
3185 Patch from Andrey Oktyabrski.
3187 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3188 Patch from Simon Arlott.
3190 PP/09 New lookups functionality failed to compile on old gcc which rejects
3191 extern declarations in function scope.
3192 Patch from Oliver Fleischmann
3194 PP/10 Use sig_atomic_t for flags set from signal handlers.
3195 Check getgroups() return and improve debugging.
3196 Fixed developed for diagnosis in bug 927 (which turned out to be
3199 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3200 Patch from Mark Zealey.
3202 PP/12 Bugzilla 1056: Improved spamd server selection.
3203 Patch from Mark Zealey.
3205 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3206 Based on patch from Heiko Schlittermann.
3208 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3209 Patch from Uwe Doering, sign-off by Michael Haardt.
3211 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3212 variable declaration deep within a block. Bug and patch from
3215 PP/15 lookups-Makefile IRIX compatibility coercion.
3217 PP/16 Make DISABLE_DKIM build knob functional.
3219 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3220 Patch by Simon Arlott
3222 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3223 variable argument macros. Our copy now differs from upstream.
3229 TF/01 Failure to get a lock on a hints database can have serious
3230 consequences so log it to the panic log.
3232 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3233 controlled using the smtp_confirmation log selector.
3235 TF/03 Include the error message when we fail to unlink a spool file.
3237 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3238 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3239 for maintaining out-of-tree patches for some time.
3241 PP/01 Bugzilla 139: Documentation and portability issues.
3242 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3243 Handle per-OS dynamic-module compilation flags.
3245 PP/02 Let /dev/null have normal permissions.
3246 The 4.73 fixes were a little too stringent and complained about the
3247 permissions on /dev/null. Exempt it from some checks.
3248 Reported by Andreas M. Kirchwitz.
3250 PP/03 Report version information for many libraries, including
3251 Exim version information for dynamically loaded libraries. Created
3252 version.h, now support a version extension string for distributors
3253 who patch heavily. Dynamic module ABI change.
3255 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3256 privilege escalation vulnerability whereby the Exim run-time user
3257 can cause root to append content of the attacker's choosing to
3260 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3263 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3264 If dropping privileges for untrusted macros, we disabled normal logging
3265 on the basis that it would fail; for the Exim run-time user, this is not
3266 the case, and it resulted in successful deliveries going unlogged.
3267 Fixed. Reported by Andreas Metzler.
3273 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3274 only prepend for the Resent-* case. Fixes regression introduced in
3275 Exim 4.70 by NM/22 for Bugzilla 607.
3277 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3278 increasing numbers of administrators be bitten by this.
3280 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3282 PP/03 Bugzilla 994: added openssl_options main configuration option.
3284 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3286 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3288 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3290 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3291 without a peer certificate, leading to a segfault because of an
3292 assumption that peers always have certificates. Be a little more
3293 paranoid. Problem reported by Martin Tscholak.
3295 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3296 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3297 NB: ClamAV planning to remove STREAM in "middle of 2010".
3298 CL also introduces -bmalware, various -d+acl logging additions and
3299 more caution in buffer sizes.
3301 PP/09 Implemented reverse_ip expansion operator.
3303 PP/10 Bugzilla 937: provide a "debug" ACL control.
3305 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3307 PP/12 Bugzilla 973: Implement --version.
3309 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3311 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3313 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3315 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3316 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3317 ignore trailing whitespace.
3319 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3321 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3324 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3325 Notification from Dr Andrew Aitchison.
3327 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3328 ExtendedDetectionInfo response format.
3329 Notification from John Horne.
3331 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3334 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3335 XSL and documented dependency on system catalogs, with examples of how
3338 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3341 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3342 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3343 configuration file which is writeable by the Exim user or group.
3345 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3346 of configuration files to cover files specified with the -C option if
3347 they are going to be used with root privileges, not just the default
3350 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3351 option (effectively making it always true).
3353 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3354 files to be used while preserving root privileges.
3356 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3357 that rogue child processes cannot use them.
3359 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3360 run-time user, instead of root.
3362 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3363 Exim run-time user without dropping privileges.
3365 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3366 result string, instead of calling string_vformat() twice with the same
3369 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3370 for other users. Others should always drop root privileges if they use
3371 -C on the command line, even for a whitelisted configure file.
3373 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3375 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3381 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3382 $data_path, and $header_path variables; fixed documentation bugs and
3385 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3386 exipick to access non-standard spools, including the "frozen" queue
3389 NM/01 Bugzilla 965: Support mysql stored procedures.
3390 Patch from Alain Williams
3392 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3394 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3395 Patch from Andreas Metzler
3397 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3398 Patch from Kirill Miazine
3400 NM/05 Bugzilla 671: Added umask to procmail example.
3402 JJ/03 installed exipick 20100323.0, fixing doc bug
3404 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3405 directory. Notification and patch from Dan Rosenberg.
3407 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3409 TK/02 Improve log output when DKIM signing operation fails.
3411 MH/01 Treat the transport option dkim_domain as a colon separated
3412 list, not as a single string, and sign the message with each element,
3413 omitting multiple occurences of the same signer.
3415 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3416 Bugzilla 985, 986. Patch by Simon Arlott
3418 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3419 Patch by Simon Arlott
3421 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3422 MBX locking. Notification from Dan Rosenberg.
3428 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3430 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3432 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3434 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3436 NM/04 Bugzilla 915: Flag broken perl installation during build.
3442 TK/01 Added patch by Johannes Berg that expands the main option
3443 "spamd_address" if it starts with a dollar sign.
3445 TK/02 Write list of recipients to X-Envelope-Sender header when building
3446 the mbox-format spool file for content scanning (suggested by Jakob
3449 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3450 (http://www.dcc-servers.net/) support via dccifd. Activated by
3451 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3453 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3454 by Mark Daniel Reidel <mr@df.eu>.
3456 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3457 When building exim an external PCRE library is now needed -
3458 PCRE is a system library on the majority of modern systems.
3459 See entry on PCRE_LIBS in EDITME file.
3461 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3462 conversation. Added nologin parameter to request.
3463 Patch contributed by Kirill Miazine.
3465 TF/01 Do not log submission mode rewrites if they do not change the address.
3467 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3469 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3470 log files in place. Contributed by Roberto Lima.
3472 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3474 TF/03 Bugzilla 615: When checking the local_parts router precondition
3475 after a local_part_suffix or local_part_prefix option, Exim now
3476 does not use the address's named list lookup cache, since this
3477 contains cached lookups for the whole local part.
3479 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3480 Robert Millan. Documentation is in experimental-spec.txt.
3482 TF/04 Bugzilla 668: Fix parallel build (make -j).
3484 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3486 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3487 Patch provided by Jan Srzednicki.
3489 TF/05 Leading white space used to be stripped from $spam_report which
3490 wrecked the formatting. Now it is preserved.
3492 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3493 that they are available at delivery time.
3495 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3497 TF/08 TLS error reporting now respects the incoming_interface and
3498 incoming_port log selectors.
3500 TF/09 Produce a more useful error message if an SMTP transport's hosts
3501 setting expands to an empty string.
3503 NM/06 Bugzilla 744: EXPN did not work under TLS.
3504 Patch provided by Phil Pennock.
3506 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3507 Patch provided by Richard Godbee.
3509 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3510 acl_smtp_notquit, added index entry.
3512 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3513 Patch provided by Eugene Bujak.
3515 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3516 accept(). Patch provided by Maxim Dounin.
3518 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3519 Patch provided by Phil Pennock.
3521 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3523 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3524 Patch provided by Brad "anomie" Jorsch.
3526 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3527 Patch provided by Dean Brooks.
3529 TK/05 Add native DKIM support (does not depend on external libraries).
3531 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3532 Patch provided by Graeme Fowler.
3534 NM/16 Bugzilla 851: Documentation example syntax fix.
3536 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3538 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3541 NM/19 Bugzilla 745: TLS version reporting.
3542 Patch provided by Phil Pennock.
3544 NM/20 Bugzilla 167: bool: condition support.
3545 Patch provided by Phil Pennock.
3547 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3548 clients. Patch provided by Phil Pennock.
3550 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3551 Patch provided by Brad "anomie" Jorsch.
3553 NM/23 Bugzilla 687: Fix misparses in eximstats.
3554 Patch provided by Heiko Schlittermann.
3556 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3557 Patch provided by Heiko Schlittermann.
3559 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3560 plus update to original patch.
3562 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3564 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3565 Patch provided by David Brownlee.
3567 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3569 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3571 NM/30 Bugzilla 888: TLS documentation bugfixes.
3573 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3575 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3576 Unlike the original bugzilla I have changed all shell scripts in src tree.
3578 NM/33 Bugzilla 898: Transport filter timeout fix.
3579 Patch by Todd Rinaldo.
3581 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3582 Patch by Serge Demonchaux.
3584 NM/35 Bugzilla 39: Base64 decode bug fixes.
3585 Patch by Jakob Hirsch.
3587 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3589 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3591 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3593 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3599 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3600 ALT-N's libdkim that I have put here:
3601 http://duncanthrax.net/exim-experimental/
3603 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3604 were called 'true' and it seems that C99 defines that as a reserved
3605 keyword to be used with 'bool' variable types. That means you could
3606 not include C99-style headers which use bools without triggering
3607 build errors in sieve.c.
3609 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3610 as mailq or other aliases. Changed the --help handling significantly
3611 to do whats expected. exim_usage() emits usage/help information.
3613 SC/01 Added the -bylocaldomain option to eximstats.
3615 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3617 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3619 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3625 PH/01 Another patch from the Sieve maintainer.
3627 PH/02 When an IPv6 address is converted to a string for single-key lookup
3628 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3629 dots are used instead of colons so that keys in lsearch files need not
3630 contain colons. This was done some time before quoting was made available
3631 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3632 (notated using the quote facility) so as to distinguish them from IPv4
3633 keys. This meant that lookups for IP addresses in host lists did not work
3634 for iplsearch lookups.
3636 This has been fixed by arranging for IPv6 addresses to be expressed with
3637 colons if the lookup type is iplsearch. This is not incompatible, because
3638 previously such lookups could never work.
3640 The situation is now rather anomalous, since one *can* have colons in
3641 ordinary lsearch keys. However, making the change in all cases is
3642 incompatible and would probably break a number of configurations.
3644 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3647 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3648 conversion specification without a maximum field width, thereby enabling
3649 a rogue spamd server to cause a buffer overflow. While nobody in their
3650 right mind would setup Exim to query an untrusted spamd server, an
3651 attacker that gains access to a server running spamd could potentially
3652 exploit this vulnerability to run arbitrary code as the Exim user.
3654 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3655 $primary_hostname instead of what libspf2 thinks the hosts name is.
3657 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3658 a directory entry by the name of the lookup key. Previously, if a
3659 symlink pointed to a non-existing file or a file in a directory that
3660 Exim lacked permissions to read, a lookup for a key matching that
3661 symlink would fail. Now it is enough that a matching directory entry
3662 exists, symlink or not. (Bugzilla 503.)
3664 PH/03 The body_linecount and body_zerocount variables are now exported in the
3667 PH/04 Added the $dnslist_matched variable.
3669 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3670 This means they are set thereafter only if the connection becomes
3673 PH/06 Added the client_condition to authenticators so that some can be skipped
3674 by clients under certain conditions.
3676 PH/07 The error message for a badly-placed control=no_multiline_responses left
3677 "_responses" off the end of the name.
3679 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3681 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3682 (without spaces) instead of just copying the configuration text.
3684 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3686 PH/11 Added $max_received_linelength.
3688 PH/12 Added +ignore_defer and +include_defer to host lists.
3690 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3691 way in which PCRE > 7.0 is built.
3693 PH/14 Implemented queue_only_load_latch.
3695 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3696 MAIL command. The effect was to mangle the value on 64-bit systems.
3698 PH/16 Another patch from the Sieve maintainer.
3700 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3702 PH/18 If a system quota error occurred while trying to create the file for
3703 a maildir delivery, the message "Mailbox is full" was not appended to the
3704 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3705 only to a quota excession during the actual writing of the file.
3707 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3708 characters?) which causes problems in log lines. The DN values are now
3709 passed through string_printing() before being added to log lines.
3711 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3712 and InterBase are left for another time.)
3714 PH/21 Added message_body_newlines option.
3716 PH/22 Guard against possible overflow in moan_check_errorcopy().
3718 PH/23 POSIX allows open() to be a macro; guard against that.
3720 PH/24 If the recipient of an error message contained an @ in the local part
3721 (suitably quoted, of course), incorrect values were put in $domain and
3722 $local_part during the evaluation of errors_copy.
3728 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3729 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3732 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3733 issue a MAIL command.
3735 PH/02 In an ACL statement such as
3737 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3739 if a client was not listed at all, or was listed with a value other than
3740 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3741 the condition was not true (as it should be), so access was not denied.
3742 The bug was that the ! inversion was incorrectly passed on to the second
3743 item. This has been fixed.
3745 PH/03 Added additional dnslists conditions == and =& which are different from
3746 = and & when the dns lookup returns more than one IP address.
3748 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3749 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3751 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3752 FSYNC, which compiles an option called disable_fsync that allows for
3753 bypassing fsync(). The documentation is heavily laced with warnings.
3755 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3757 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3758 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3759 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3760 including adding "make clean"; (3) Added -fPIC when compiling the test
3761 dynamically loaded module, to get rid of a warning.
3763 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3764 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3765 = 0s. The bug is otherwise harmless.
3767 PH/07 There was a bug in the dovecot authenticator such that the value of
3768 $auth1 could be overwritten, and so not correctly preserved, after a
3769 successful authentication. This usually meant that the value preserved by
3770 the server_setid option was incorrect.
3772 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3774 PH/09 Installed PCRE release 7.0.
3776 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3777 run for batched SMTP input. It is now run at the start of every message
3778 in the batch. While fixing this I discovered that the process information
3779 (output by running exiwhat) was not always getting set for -bs and -bS
3780 input. This is fixed, and it now also says "batched" for BSMTP.
3782 PH/11 Added control=no_pipelining.
3784 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3785 patch, slightly modified), and move the expansion of helo_data till after
3786 the connection is made in the smtp transport (so it can use these
3789 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3791 PH/14 Added log_selector = +pid.
3793 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3795 PH/16 Add ${if forany and ${if forall.
3797 PH/17 Added dsn_from option to vary the From: line in DSNs.
3799 PH/18 Flush SMTP output before performing a callout, unless control =
3800 no_callout_flush is set.
3802 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3803 was true (the default) a successful delivery failed to delete the retry
3804 item, thus causing premature timeout of the address. The bug is now
3807 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3809 PH/21 Long custom messages for fakedefer and fakereject are now split up
3810 into multiline responses in the same way that messages for "deny" and
3811 other ACL rejections are.
3813 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3814 with slight modification.
3816 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3817 draft, changing the syntax and factoring some duplicate code".
3819 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3820 for deliveries of the second and subsequent messages over the same SMTP
3823 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3824 ${reduce, with only minor "tidies".
3826 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3828 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3829 expansion side effects.
3831 PH/27 When a message times out after an over-quota error from an Exim-imposed
3832 quota, the bounce message says "mailbox is full". This message was not
3833 being given when it was a system quota that was exceeded. It now should
3836 MH/03 Made $recipients available in local_scan(). local_scan() already has
3837 better access to the recipient list through recipients_list[], but
3838 $recipients can be useful in postmaster-provided expansion strings.
3840 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3841 in the case of a MAIL command with additional options following the
3842 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3843 were accidentally chopped off.
3845 PH/29 SMTP synchronization checks are implemented when a command is read -
3846 there is a check that no more input is waiting when there shouldn't be
3847 any. However, for some commands, a delay in an ACL can mean that it is
3848 some time before the response is written. In this time, more input might
3849 arrive, invalidly. So now there are extra checks after an ACL has run for
3850 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3851 pipelining has not been advertised.
3853 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3855 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3856 returned in the SMTP response when smtp_return_error_details was false.
3857 This has been fixed.
3859 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3860 instead of the C I/O that was originally supplied, because problems were
3861 reported on Solaris.
3863 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3864 Exim which did not show up earlier: it was assuming that a call to
3865 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3866 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3867 was a macro that became an assignment, so it seemed to work. This has
3868 changed to a proper function call with a void return, hence the compile
3869 error. Exim's code has been fixed.
3871 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3874 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3876 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3878 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3879 because it turns out to be too big or there is a timeout) there is no
3880 "Completed" line in the log. When some messages of this type were
3881 selected by exigrep, they were listed as "not completed". Others were
3882 picked up by some special patterns. I have improved the selection
3883 criteria to be more general.
3885 PH/38 The host_find_failed option in the manualroute router can now be set
3886 to "ignore", to completely ignore a host whose IP address cannot be
3887 found. If all hosts are ignored, the behaviour is controlled by the new
3888 host_all_ignored option.
3890 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3891 homing or because of multiple MX records with /mx) generated more than
3892 one IP address, and the following item turned out to be the local host,
3893 all the secondary addresses of the first item were incorrectly removed
3894 from the list, along with the local host and any following hosts (which
3895 is what is supposed to happen).
3897 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3898 whoever called Exim into the -H file. In the case of the daemon it was
3899 behaving confusingly. When first started, it used values for whoever
3900 started the daemon, but after a SIGHUP it used the Exim user (because it
3901 calls itself on a restart). I have changed the code so that it now always
3904 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3905 message are rejected with the same error (e.g. no authentication or bad
3906 sender address), and a DATA command is nevertheless sent (as can happen
3907 with PIPELINING or a stupid MUA), the error message that was given to the
3908 RCPT commands is included in the rejection of the DATA command. This is
3909 intended to be helpful for MUAs that show only the final error to their
3912 PH/42 Another patch from the Sieve maintainer.
3914 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3915 Eximstats - Fixed some broken HTML links and added missing column headers
3917 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3918 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3920 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3922 PH/43 Yet another patch from the Sieve maintainer.
3924 PH/44 I found a way to check for a TCP/IP connection going away before sending
3925 the response to the final '.' that terminates a message, but only in the
3926 case where the client has not sent further data following the '.'
3927 (unfortunately, this is allowed). However, in many cases there won't be
3928 any further data because there won't be any more messages to send. A call
3929 to select() can be used: if it shows that the input is "ready", there is
3930 either input waiting, or the socket has been closed. An attempt to read
3931 the next input character can distinguish the two cases. Previously, Exim
3932 would have sent an OK response which the client would never have see.
3933 This could lead to message repetition. This fix should cure that, at
3934 least in a lot of common cases.
3936 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3937 advertised in response to EHLO.
3943 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3944 fixed by 4.65/MH/01 (is this a record?) are fixed:
3946 (i) An empty string was always treated as zero by the numeric comparison
3947 operators. This behaviour has been restored.
3949 (ii) It is documented that the numeric comparison operators always treat
3950 their arguments as decimal numbers. This was broken in that numbers
3951 starting with 0 were being interpreted as octal.
3953 While fixing these problems I realized that there was another issue that
3954 hadn't been noticed. Values of message_size_limit (both the global option
3955 and the transport option) were treated as octal if they started with 0.
3956 The documentation was vague. These values are now always treated as
3957 decimal, and I will make that clear in the documentation.
3963 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3964 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3967 MH/01 Don't check that the operands of numeric comparison operators are
3968 integers when their expansion is in "skipping" mode (fixes bug
3969 introduced by 4.64-PH/07).
3971 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3972 child addresses, Exim now panics and dies. Previously, because the count
3973 is held in a short int, deliveries were likely to be lost. As such a
3974 large number of recipients for a single message is ridiculous
3975 (performance will be very, very poor), I have chosen to impose a limit
3976 rather than extend the field.
3982 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3983 leftover -K file (the existence of which was triggered by #402).
3984 While we were at it, introduced process PID as part of the -K
3985 filename. This should rule out race conditions when creating
3988 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3989 processing considerably. Previous code took too long for large mails,
3990 triggering a timeout which in turn triggers #401.
3992 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
3993 in the DK code in transports.c. sendfile() is not really portable,
3994 hence the _LINUX specificness.
3996 TF/01 In the add_headers option to the mail command in an Exim filter,
3997 there was a bug that Exim would claim a syntax error in any
3998 header after the first one which had an odd number of characters
4001 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4002 callout verification, Exim cached a "reject" for the entire domain. This
4003 is correct for most verifications, but it is not correct for a recipient
4004 verification with use_sender or use_postmaster set, because in that case
4005 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4006 case of MAIL FROM:<> rejection from other early rejections (e.g.
4007 rejection of HELO). When verifying a recipient using a non-null MAIL
4008 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4009 Whatever the result of the callout, the value of the domain cache is
4010 left unchanged (for any other kind of callout, getting as far as trying
4011 RCPT means that the domain itself is ok).
4013 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4016 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4017 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4018 session, the daemon ignores ECONNECT errors and logs others; it now
4019 ignores EPIPE as well.
4021 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4022 (quoted-printable decoding).
4024 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4025 later the small subsequent patch to fix an introduced bug.
4027 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4029 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4031 PH/08 An error is now given if message_size_limit is specified negative.
4033 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4034 to be given (somewhat) arbitrary names.
4036 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4039 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4040 miscellaneous code fixes
4042 PH/10 Added the log_reject_target ACL modifier to specify where to log
4045 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4046 hostname. This is wrong, because it relates to the incoming message (and
4047 probably the interface on which it is arriving) and not to the outgoing
4048 callout (which could be using a different interface). This has been
4049 changed to use the value of the helo_data option from the smtp transport
4050 instead - this is what is used when a message is actually being sent. If
4051 there is no remote transport (possible with a router that sets up host
4052 addresses), $smtp_active_hostname is used.
4054 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4055 tweaks were necessary in order to get it to work (see also 21 below):
4056 (a) The code assumed that strncpy() returns a negative number on buffer
4057 overflow, which isn't the case. Replaced with Exim's string_format()
4059 (b) There were several signed/unsigned issues. I just did the minimum
4060 hacking in of casts. There is scope for a larger refactoring.
4061 (c) The code used strcasecmp() which is not a standard C function.
4062 Replaced with Exim's strcmpic() function.
4063 (d) The code set only $1; it now sets $auth1 as well.
4064 (e) A simple test gave the error "authentication client didn't specify
4065 service in request". It would seem that Dovecot has changed its
4066 interface. Fortunately there's a specification; I followed it and
4067 changed what the client sends and it appears to be working now.
4069 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4072 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4073 address A is aliased to B and C, where B exists and C does not. Without
4074 -v the output is "A verified" because verification stops after a
4075 successful redirection if more than one address is generated. However,
4076 with -v the child addresses are also verified. Exim was outputting "A
4077 failed to verify" and then showing the successful verification for C,
4078 with its parentage. It now outputs "B failed to verify", showing B's
4079 parentage before showing the successful verification of C.
4081 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4082 look up a TXT record in a specific list after matching in a combined
4085 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4086 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4087 they consult the DNS. I had assumed they would set it the way they
4088 wanted; and indeed my experiments on Linux seem to show that in some
4089 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4090 To be on the safe side, however, I have now made the interface to
4091 host_find_byname() similar to host_find_bydns(), with an argument
4092 containing the DNS resolver options. The host_find_byname() function now
4093 sets these options at its start, just as host_find_bydns() does. The smtp
4094 transport options dns_qualify_single and dns_search_parents are passed to
4095 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4096 of host_find_byname() use the default settings of RES_DEFNAMES
4097 (qualify_single) but not RES_DNSRCH (search_parents).
4099 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4100 spool_read_header() do less string testing, by means of a preliminary
4101 switch on the second character of optional "-foo" lines. (This is
4102 overdue, caused by the large number of possibilities that now exist.
4103 Originally there were few.) While I was there, I also converted the
4104 str(n)cmp tests so they don't re-test the leading "-" and the first
4105 character, in the hope this might squeeze out yet more improvement.
4107 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4108 flag allowing group syntax was set by the header_syntax check but not
4109 turned off, possible causing trouble later; (2) The flag was not being
4110 set at all for the header_verify test, causing "group"-style headers to
4111 be rejected. I have now set it in this case, and also caused header_
4112 verify to ignore an empty address taken from a group. While doing this, I
4113 came across some other cases where the code for allowing group syntax
4114 while scanning a header line wasn't quite right (mostly, not resetting
4115 the flag correctly in the right place). These bugs could have caused
4116 trouble for malformed header lines. I hope it is now all correct.
4118 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4119 with the "reply" argument non-NULL. The code, however (which originally
4120 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4121 but it didn't always do it. This confused somebody who was copying the
4122 code for some other use. I have removed all the tests.
4124 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4125 feature that was used to support insecure browsers during the U.S. crypto
4126 embargo. It requires special client support, and Exim is probably the
4127 only MTA that supported it -- and would never use it because real RSA is
4128 always available. This code has been removed, because it had the bad
4129 effect of slowing Exim down by computing (never used) parameters for the
4130 RSA_EXPORT functionality.
4132 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4133 authenticator to fail if there's a tab character in the incoming data
4134 (there should never be unless someone is messing about, as it's supposed
4135 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4136 if the connection is using TLS or if the remote IP is the same as the
4137 local IP, and the "valid-client-cert option" if a client certificate has
4140 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4141 authenticators. This can be used for authorization after authentication
4142 succeeds. (In the case of plaintext, it servers for both authentication
4145 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4146 if any retry times were supplied.
4148 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4149 connection, where there is no client IP address to check. In this
4150 situation, the verify now always succeeds.
4152 PH/25 Applied John Jetmore's -Mset patch.
4154 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4156 PH/27 In a string expansion for a processed (not raw) header when multiple
4157 headers of the same name were present, leading whitespace was being
4158 removed from all of them, but trailing whitespace was being removed only
4159 from the last one. Now trailing whitespace is removed from each header
4160 before concatenation. Completely empty headers in a concatenation (as
4161 before) are ignored.
4163 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4164 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4166 PH/29 [Removed. This was a change that I later backed out, and forgot to
4167 correct the ChangeLog entry (that I had efficiently created) before
4168 committing the later change.]
4170 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4171 address errors (4xx response to RCPT) over the same connection as other
4172 messages routed to the same hosts. Such deliveries are always "forced",
4173 so retry times are not inspected. This resulted in far too many retries
4174 for the affected addresses. The effect occurred only when there were more
4175 hosts than the hosts_max_try setting in the smtp transport when it had
4176 the 4xx errors. Those hosts that it had tried were not added to the list
4177 of hosts for which the message was waiting, so if all were tried, there
4178 was no problem. Two fixes have been applied:
4180 (i) If there are any address or message errors in an SMTP delivery, none
4181 of the hosts (tried or untried) are now added to the list of hosts
4182 for which the message is waiting, so the message should not be a
4183 candidate for sending over the same connection that was used for a
4184 successful delivery of some other message. This seems entirely
4185 reasonable: after all the message is NOT "waiting for some host".
4186 This is so "obvious" that I'm not sure why it wasn't done
4187 previously. Hope I haven't missed anything, but it can't do any
4188 harm, as the worst effect is to miss an optimization.
4190 (ii) If, despite (i), such a delivery is accidentally attempted, the
4191 routing retry time is respected, so at least it doesn't keep
4192 hammering the server.
4194 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4195 in ${readsocket because some servers need this prod.
4197 PH/32 Added some extra debug output when updating a wait-xxx database.
4199 PH/33 The hint "could be header name not terminated by colon", which has been
4200 given for certain expansion errors for a long time, was not being given
4201 for the ${if def:h_colon_omitted{... case.
4203 PH/34 The spec says: "With one important exception, whenever a domain list is
4204 being scanned, $domain contains the subject domain." There was at least
4205 one case where this was not true.
4207 PH/35 The error "getsockname() failed: connection reset by peer" was being
4208 written to the panic log as well as the main log, but it isn't really
4209 panic-worthy as it just means the connection died rather early on. I have
4210 removed the panic log writing for the ECONNRESET error when getsockname()
4213 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4214 runs only) independently of the message's sender address. This meant
4215 that, if the 4xx error was in fact related to the sender, a different
4216 message to the same recipient with a different sender could confuse
4217 things. In particular, this can happen when sending to a greylisting
4218 server, but other circumstances could also provoke similar problems.
4219 I have changed the default so that the retry time for these errors is now
4220 based a combination of the sender and recipient addresses. This change
4221 can be overridden by setting address_retry_include_sender=false in the
4224 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4225 remote server are returned as part of bounce messages. This was not
4226 happening for LMTP over a pipe (the lmtp transport), but now it is the
4227 same for both kinds of LMTP.
4229 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4230 in header lines that were in fact CNAMEs. This is no longer the case.
4232 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4233 and queue runs started by the daemon processed all messages. This has
4234 been fixed so that -R and -S can now usefully be given with -q<time>.
4236 PH/40 Import PCRE release 6.7 (fixes some bugs).
4238 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4240 PH/42 Give an error if -q is specified more than once.
4242 PH/43 Renamed the variables $interface_address and $interface_port as
4243 $received_ip_address and $received_port, to make it clear that these
4244 values apply to message reception, and not to the outgoing interface when
4245 a message is delivered. (The old names remain recognized, of course.)
4247 PH/44 There was no timeout on the connect() call when using a Unix domain
4248 socket in the ${readsocket expansion. There now is.
4250 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4251 be meaningful with "accept".
4253 SC/01 Eximstats V1.43
4254 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4256 SC/02 Eximstats V1.44
4257 Use a glob alias rather than an array ref in the generated
4258 parser. This improves both readability and performance.
4260 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4261 Collect SpamAssassin and rejection statistics.
4262 Don't display local sender or destination tables unless
4263 there is data to show.
4264 Added average volumes into the top table text output.
4266 SC/04 Eximstats V1.46
4267 Collect data on the number of addresses (recipients)
4268 as well as the number of messages.
4270 SC/05 Eximstats V1.47
4271 Added 'Message too big' to the list of mail rejection
4272 reasons (thanks to Marco Gaiarin).
4274 SC/06 Eximstats V1.48
4275 Mainlog lines which have GMT offsets and are too short to
4276 have a flag are now skipped.
4278 SC/07 Eximstats V1.49 (Alain Williams)
4279 Added the -emptyok flag.
4281 SC/08 Eximstats V1.50
4282 Fixes for obtaining the IP address from reject messages.
4284 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4285 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4286 whitespace changes from 4.64-PH/27
4288 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4291 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4292 are found, allow negative numbers in numeric criteria)
4294 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4296 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4297 to match changes made in 4.64-PH/43
4299 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4301 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4302 "domain" is actually an IP address, and force a failure. This locks out
4303 those revolvers/nameservers that support "A-for-A" lookups, in
4304 contravention of the specifications.
4306 PH/48 When a host name was looked up from an IP address, and the subsequent
4307 forward lookup of the name timed out, the host name was left in
4308 $sender_host_name, contrary to the specification.
4310 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4311 restricted to single-key lookups, Exim was not diagnosing an error if
4312 * or *@ was used with a query-style lookup.
4314 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4316 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4317 long ago, but noone interested enough thought of it. Let's just say that
4318 the "1.1" means that there are some new functions that weren't there at
4319 some point in the past.
4321 PH/51 Error processing for expansion failure of helo_data from an smtp
4322 transport during callout processing was broken.
4324 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4325 tested/used via the -bh/-bhc/-bs options.
4327 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4328 bug, fixed in subsequent PCRE releases).
4330 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4331 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4333 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4339 SC/01 Use a glob alias rather than an array ref in eximstats generated
4340 parser. This improves both readability and performance.
4342 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4343 Don't display local sender or destination tables in eximstats unless
4344 there is data to show.
4345 Added average volumes into the eximstats top table text output.
4347 SC/03 Collect data on the number of addresses (recipients) as well
4348 as the number of messages in eximstats.
4350 TF/01 Correct an error in the documentation for the redirect router. Exim
4351 does (usually) call initgroups() when daemonizing.
4353 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4354 with consistent privilege compared to when running as a daemon.
4356 TF/03 Note in the spec that $authenticated_id is not set for local
4357 submissions from trusted users.
4359 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4360 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4362 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4363 by adding some example configuration directives to the default
4364 configuration file. A little bit of work is required to uncomment the
4365 directives and define how usernames and passwords are checked, but
4366 there is now a framework to start from.
4368 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4369 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4370 without this. I don't know how relevant this is to other LDAP libraries.
4372 PH/02 Add the verb name to the "unknown ACL verb" error.
4374 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4376 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4378 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4379 directory not expanded when it should be if an expanded home directory
4380 was set for the address (which is overridden by the transport).
4382 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4385 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4386 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4387 because it is too late at that time, and has no effect.
4389 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4390 security issue with \' (bugzilla #107). I could not use the
4391 PQescapeStringConn() function, because it needs a PGconn value as one of
4394 PH/08 When testing addresses using -bt, indicate those final addresses that
4395 are duplicates that would not cause an additional delivery. At least one
4396 person was confused, thinking that -bt output corresponded to deliveries.
4397 (Suppressing duplicates isn't a good idea as you lose the information
4398 about possibly different redirections that led to the duplicates.)
4400 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4401 systems where poll() doesn't work, in particular OS X.
4403 PH/10 Added more information to debugging output for retry time not reached.
4405 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4406 operations in malware.c.
4408 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4411 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4412 syslog_duplication set false, log lines that would normally be written
4413 both the the main log and to the reject log were not written to syslog at
4416 PH/14 In the default configuration, change the use of "message" in ACL warn
4417 statements to "add_header".
4419 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4420 not followed by a command (e.g. "seen endif").
4422 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4423 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4426 PH/17 Added extra conditions to the default value of delay_warning_condition
4430 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4431 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4432 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4435 The Auto-Submitted: and various List- headers are standardised, whereas I
4436 don't think Precedence: ever was.
4438 PH/18 Refactored debugging code in route_finduser() to show more information,
4439 in particular, the error code if getpwnam() issues one.
4441 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4442 This is apparently needed in addition to the PH/07 change above to avoid
4443 any possible encoding problems.
4445 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4446 but not after initializing Perl.
4448 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4449 output them only if debugging. By default they are written stderr,
4450 apparently, which is not desirable.
4452 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4455 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4458 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4460 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4461 authenticated or an ident call has been made. Suppress the default
4462 values for $authenticated_id and $authenticated_sender (but permit -oMai
4463 and -oMas) when testing with -bh.
4465 PH/24 Re-jigged the order of the tests in the default configuration so that the
4466 tests for valid domains and recipients precede the DNS black list and CSA
4467 tests, on the grounds that those ones are more expensive.
4469 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4470 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4471 This bug exists in every version of Exim that I still have, right back to
4474 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4475 However, an attempt to turn on case-sensitivity in a regex key by
4476 including (?-i) didn't work because the subject string was already
4477 lowercased, and the effects were non-intuitive. It turns out that a
4478 one-line patch can be used to allow (?-i) to work as expected.
4484 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4485 other effects) broke the use of negated acl sub-conditions.
4487 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4490 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4491 "Deny" causes Exim to reject the incoming connection with a 554 error.
4492 Unfortunately, if there is a major crisis, such as a disk failure,
4493 tcp-wrappers gives "deny", whereas what one would like would be some
4494 kind of temporary error. A kludge has been added to help with this.
4495 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4496 554 error is used if errno is still zero or contains ENOENT (which occurs
4497 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4500 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4502 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4503 errors. Otherwise a message that provokes a temporary error (when other
4504 messages do not) can cause a whole host to time out.
4506 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4507 addresses were routed directly to files or pipes from a redirect router.
4508 File deliveries just didn't batch; pipe deliveries might have suffered
4511 PH/06 A failure to get a lock for a hints database would erroneously always say
4512 "Failed to get write lock", even when it was really a read lock.
4514 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4515 of 0600. This has been changed to use the value of the lockfile_mode
4516 option (which defaults to 0600).
4518 PH/08 Applied small patch from the Sieve maintainer.
4520 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4521 folder from quota calculations, a direct delivery into this folder messed
4522 up the contents of the maildirsize file. This was because the regex was
4523 used only to exclude .Trash (or whatever) when the size of the mailbox
4524 was calculated. There was no check that a delivery was happening into an
4525 excluded directory. This bug has been fixed by ignoring all quota
4526 processing for deliveries into excluded directories.
4528 PH/10 Added the maildirfolder_create_regex option to appendfile.
4534 PH/01 The code for finding all the local interface addresses on a FreeBSD
4535 system running IPv6 was broken. This may well have applied to all BSD
4536 systems, as well as to others that have similar system calls. The broken
4537 code found IPv4 interfaces correctly, but gave incorrect values for the
4538 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4539 that it would not match correctly against @[] and not recognize the IPv6
4542 PH/02 The ipliteral router was not recognizing addresses of the form user@
4543 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4545 PH/03 Added disable_ipv6.
4547 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4548 decoded form, because it is most often used to construct To: headers
4549 lines in autoreplies, and the decoded form may well be syntactically
4550 invalid. However, $reply_address has leading white space removed, and all
4551 newlines turned into spaces so that the autoreply transport does not
4554 PH/05 If group was specified without a user on a router, and no group or user
4555 was specified on a transport, the group from the router was ignored.
4557 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4558 for visible compile-time settings that can be used to change these
4559 numbers, for those that want even more. Backwards compatibility with old
4560 spool files has been maintained. However, going back to a previous Exim
4561 release will lost any variables that are in spool files.
4563 PH/07 Two small changes when running in the test harness: increase delay when
4564 passing a TCP/IP connection to a new process, in case the original
4565 process has to generate a bounce, and remove special handling of
4566 127.0.0.2 (sic), which is no longer necessary.
4568 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4569 be the same on different OS.
4571 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4574 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4575 whether --show-vars was specified or not
4577 JJ/02 exipick: Added support for new ACL variable spool format introduced
4580 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4581 syntactically invalid From: or Reply-to: line, and a filter used this to
4582 generate an autoreply, and therefore failed to obtain an address for the
4583 autoreply, Exim could try to deliver to a non-existent relative file
4584 name, causing unrelated and misleading errors. What now happens is that
4585 it logs this as a hard delivery error, but does not attempt to create a
4588 PH/11 The exinext utility has a -C option for testing purposes, but although
4589 the given file was scanned by exinext itself; it wasn't being passed on
4590 when Exim was called.
4592 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4593 an end-of-file indication when reading a command response.
4595 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4596 compiled. In many other places in Exim, IPv6 addresses are always
4597 recognized, so I have changed this. It also means that IPv4 domain
4598 literals of the form [IPV4:n.n.n.n] are now always recognized.
4600 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4601 used if the router is not running as root, for example, when verifying at
4602 ACL time, or when using -bh. The debugging output from this situation was
4603 non-existent - all you got was a failure to exec. I have made two
4606 (a) Failures to set uid/gid, the current directory, or a process leader
4607 in a subprocess such as that created by queryprogram now generate
4608 suitable debugging output when -d is set.
4610 (b) The queryprogram router detects when it is not running as root,
4611 outputs suitable debugging information if -d is set, and then runs
4612 the subprocess without attempting to change uid/gid.
4614 PH/15 Minor change to Makefile for building test_host (undocumented testing
4617 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4618 additional section of a DNS packet that returns MX or SRV records.
4619 Instead, it always explicitly searches for A/AAAA records. This avoids
4620 major problems that occur when a DNS server includes only records of one
4621 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4622 fixed another bug: if SRV records were looked up and the corresponding
4623 address records were *not* found in the additional section, the port
4624 values from the SRV records were lost.
4626 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4627 using the correct key (the original address) when searching the retry
4628 rules in order to find which one to use for generating the retry hint.
4630 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4631 adding the default one. Similarly, if it contains a Reply-To: header, the
4632 errors_reply_to option, if set, is not used.
4634 PH/19 When calculating a retry time, Exim used to measure the "time since
4635 failure" by looking at the "first failed" field in the retry record. Now
4636 it does not use this if it is later than than the arrival time of the
4637 message. Instead it uses the arrival time. This makes for better
4638 behaviour in cases where some deliveries succeed, thus re-setting the
4639 "first failed" field. An example is a quota failure for a huge message
4640 when small messages continue to be delivered. Without this change, the
4641 "time since failure" will always be short, possible causing more frequent
4642 delivery attempts for the huge message than are intended.
4643 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4645 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4646 $1, $2, $3) because the numerical variables can be reset during some
4647 expansion items (e.g. "match"), thereby losing the authentication data.
4649 PH/21 Make -bV show the size of off_t variables so that the test suite can
4650 decide whether to run tests for quotas > 2G.
4652 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4653 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4654 filecount value is greater than 2G or if a quota value is greater than 2G
4655 on a system where the size of off_t is not greater than 4, a panic error
4658 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4659 never match. The debug and -bh output now contains an explicit error
4660 message indicating a malformed IPv4 address or mask.
4662 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4663 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4664 PH/23 above applies.
4666 PH/25 Do not write to syslog when running in the test harness. The only
4667 occasion when this arises is a failure to open the main or panic logs
4668 (for which there is an explicit test).
4670 PH/26 Added the /no_tell option to "control=freeze".
4672 PH/27 If a host name lookup failed very early in a connection, for example, if
4673 the IP address matched host_lookup and the reverse lookup yielded a name
4674 that did not have a forward lookup, an error message of the form "no IP
4675 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4676 could be logged. Now it outputs the IP address instead of "NULL".
4678 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4679 allows the sender and the authenticated sender to be set when
4680 submitting a message from within Exim. Since child_open_exim() is
4681 documented for local_scan(), the new function should be too.
4683 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4684 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4685 results in an empty string is now treated as unset.
4687 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4689 PH/31 Added sender_verify_fail logging option.
4691 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4692 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4693 tidied the source and removed it altogether.
4695 PH/33 When a queue run was abandoned because the load average was too high, a
4696 log line was always written; now it is written only if the queue_run log
4697 selector is set. In addition, the log line for abandonment now contains
4698 information about the queue run such as the pid. This is always present
4699 in "start" and "stop" lines but was omitted from the "abandon" line.
4701 PH/34 Omit spaces between a header name and the colon in the error message that
4702 is given when verify = headers_syntax fails (if there are lots of them,
4703 the message gets confusing).
4705 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4706 names, as there are now some PTR records that contain slashes. This check
4707 is only to protect against broken name servers that fall over on strange
4708 characters, so the fact that it applies to all lookups doesn't matter.
4710 PH/36 Now that the new test suite is complete, we can remove some of the
4711 special code in Exim that was needed for the old test suite. For example,
4712 sorting DNS records because real resolvers return them in an arbitrary
4713 order. The new test suite's fake resolver always returns records in the
4716 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4717 bounces) except when queue_only is set, to avoid logging races between
4718 the different processes.
4720 PH/38 Panic-die if .include specifies a non-absolute path.
4722 PH/39 A tweak to the "H" retry rule from its user.
4724 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4725 a label. They prevented compilation on older perls.
4727 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4728 a warning to be raised on newish perls.
4730 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4731 on queue. Changes to match documented behaviour of showing count of
4732 messages matching specified criteria.
4734 PH/40 Changed the default ident timeout from 30s to 5s.
4736 PH/41 Added support for the use of login_cap features, on those BSD systems
4737 that have them, for controlling the resources used by pipe deliveries.
4739 PH/42 The content-scanning code uses fopen() to create files in which to put
4740 message data. Previously it was not paying any attention to the mode of
4741 the files. Exim runs with umask(0) because the rest of the code creates
4742 files with open(), and sets the required mode explicitly. Thus, these
4743 files were ending up world-writeable. This was not a big issue, because,
4744 being within the spool directory, they were not world-accessible. I have
4745 created a function called modefopen, which takes an additional mode
4746 argument. It sets umask(777), creates the file, chmods it to the required
4747 mode, then resets the umask. All the relevant calls to fopen() in the
4748 content scanning code have been changed to use this function.
4750 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4751 to 24 hours. This avoids potential overflow problems when processing G
4752 and H retry rules. I suspect nobody ever tinkers with this value.
4754 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4756 PH/45 When the plaintext authenticator is running as a client, the server's
4757 challenges are checked to ensure they are valid base64 strings. By
4758 default, the authentication attempt is cancelled if an invalid string is
4759 received. Setting client_ignore_invalid_base64 true ignores these errors.
4760 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4761 they are received. Thus, the responses can be made to depend on the
4762 challenges. If an invalid string is ignored, an empty string is placed in
4765 PH/46 Messages that are created by the autoreply transport now contains a
4766 References: header, in accordance with RFCs 2822 and 3834.
4768 PH/47 Added authenticated_sender_force to the smtp transport.
4770 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4772 PH/49 Installed latest patch from the Sieve maintainer.
4774 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4775 was also set, the appendfile transport was unnecessarily scanning a
4776 directory of message files (e.g. for maildir delivery) to find the count
4777 of files (along with the size), even though it did not need this
4778 information. It now does the scan only if it needs to find either the
4779 size of the count of files.
4781 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4783 PH/52 Two bugs concerned with error handling when the smtp transport is
4786 (i) Exim was not creating retry information for temporary errors given
4787 for individual recipients after the DATA command when the smtp transport
4788 was used in LMTP mode. This meant that they could be retried too
4789 frequently, and not timed out correctly.
4791 (ii) Exim was setting the flag that allows error details to be returned
4792 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4793 recipients that were returned after the DATA command.
4795 PH/53 This is related to PH/52, but is more general: for any failing address,
4796 when detailed error information was permitted to be returned to the
4797 sender, but the error was temporary, then after the final timeout, only
4798 "retry timeout exceeded" was returned. Now it returns the full error as
4799 well as "retry timeout exceeded".
4801 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4802 do this, and (what is worse) MTAs that accept it.
4804 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4805 will now be deprecated.
4807 PH/56 New os.c-cygwin from the Cygwin maintainer.
4809 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4810 formats (previously only available in exim formats via -bpr, -bpru,
4811 and -bpra. Now also available in native and exiqgrep formats)
4813 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4814 with very large, slow to parse queues
4816 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4818 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4820 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4821 responses to authentication challenges, though it was showing the
4822 challenges; (ii) I've removed the CR characters from the debug output for
4825 PH/58 Allow for the insertion of a newline as well as a space when a string
4826 is turned into more than one encoded-word during RFC 2047 encoding. The
4827 Sieve code now uses this.
4829 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4830 data_4xx, lost_connection, tls_required.
4832 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4833 message was being sent as an SMTP response.
4835 PH/61 Add -l and -k options to exicyclog.
4837 PH/62 When verifying, if an address was redirected to one new address, so that
4838 verification continued, and the new address failed or deferred after
4839 having set something in $address_data, the value of $address_data was not
4840 passed back to the ACL. This was different to the case when no
4841 redirection occurred. The value is now passed back in both cases.
4843 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4844 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4845 use login_cap.h, so on its own it isn't the distinguishing feature. The
4846 new name refers directly to the setclassresources() function.
4848 PH/65 Added configuration files for NetBSD3.
4850 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4852 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4853 is preferred over IPv4.
4855 PH/68 The bounce_return_message and bounce_return_body options were not being
4856 honoured for bounces generated during the reception of non-SMTP messages.
4857 In particular, this applied to messages rejected by the ACL. This bug has
4858 been fixed. However, if bounce_return_message is true and bounce_return_
4859 body is false, the headers that are returned for a non-SMTP message
4860 include only those that have been read before the error was detected.
4861 (In the case of an ACL rejection, they have all been read.)
4863 PH/69 The HTML version of the specification is now built in a directory called
4864 spec_html instead of spec.html, because the latter looks like a path with
4865 a MIME-type, and this confuses some software.
4867 PH/70 Catch two compiler warnings in sieve.c.
4869 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4870 function verify_get_ident() calls ip_connect() to connect a socket, but
4871 if the "connect()" function timed out, ip_connect() used to close the
4872 socket. However, verify_get_ident() also closes the socket later, and in
4873 between Exim writes to the log, which may get opened at this point. When
4874 the socket was closed in ip_connect(), the log could get the same file
4875 descriptor number as the socket. This naturally causes chaos. The fix is
4876 not to close the socket in ip_connect(); the socket should be closed by
4877 the function that creates it. There was only one place in the code where
4878 this was missing, in the iplookup router, which I don't think anybody now
4879 uses, but I've fixed it anyway.
4881 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4882 well as to direct DNS lookups. Otherwise the handling of names in host
4883 lists is inconsistent and therefore confusing.
4889 PH/01 Two changes to the default runtime configuration:
4891 (1) Move the checks for relay_from_hosts and authenticated clients from
4892 after to before the (commented out) DNS black list checks.
4894 (2) Add control=submission to the relay_from_hosts and authenticated
4895 clients checks, on the grounds that messages accepted by these
4896 statements are most likely to be submissions.
4898 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4900 (1) Generate an error if the third argument for the ${prvs expansion is
4903 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4906 (3) Reset the variables that are obtained from the first argument of
4907 ${prvscheck and used in the second argument before leaving the code,
4908 because their memory is reclaimed, so using them afterwards may do
4911 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4912 one (it's much easier than Tom thought :-).
4914 (5) Because of (4), we can now allow for the use of $prvscheck_result
4915 inside the third argument.
4917 PH/03 For some reason, the default setting of PATH when running a command from
4918 a pipe transport was just "/usr/bin". I have changed it to
4921 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4922 anything to be listed in the output from -bV.
4924 PH/05 When a filter generated an autoreply, the entire To: header line was
4925 quoted in the delivery log line, like this:
4927 => >A.N.Other <ano@some.domain> <original@ddress> ...
4929 This has been changed so that it extracts the operative address. There
4930 may be more than one such address. If so, they are comma-separated, like
4933 => >ano@some.domain,ona@other.domain <original@ddress> ...
4935 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4936 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4937 not being looked up in the rDNS to get a host name, Exim was showing the
4938 IP address twice in Received: lines, even though the IP addresses were
4939 identical. For example:
4941 Received: from [1.2.3.4] (helo=[1.2.3.4])
4943 However, if the real host name was known, it was omitting the HELO data
4944 if it matched the actual IP address. This has been tidied up so that it
4945 doesn't show the same IP address twice.
4947 PH/07 When both +timestamp and +memory debugging was on, the value given by
4948 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4949 called by the debug printing, thereby overwriting the timestamp buffer.
4950 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4952 PH/08 When the original message was included in an autoreply transport, it
4953 always said "this is a copy of the message, including all the headers",
4954 even if body_only or headers_only was set. It now gives an appropriate
4957 PH/09 Applied a patch from the Sieve maintainer which:
4959 o fixes some comments
4960 o adds the (disabled) notify extension core
4961 o adds some debug output for the result of if/elsif tests
4962 o points to the current vacation draft in the documentation
4963 and documents the missing references header update
4967 o fixes a bug in processing the envelope test (when testing
4968 multiple envelope elements, the last element determined the
4971 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4972 Electronic Mail") by including:
4974 Auto-submitted: auto-generated
4976 in the messages that it generates (bounce messages and others, such as
4977 warnings). In the case of bounce messages for non-SMTP messages, there was
4978 also a typo: it was using "Auto_submitted" (underscore instead of
4979 hyphen). Since every message generated by Exim is necessarily in response
4980 to another message, thes have all been changed to:
4982 Auto-Submitted: auto-replied
4984 in accordance with these statements in the RFC:
4986 The auto-replied keyword:
4988 - SHOULD be used on messages sent in direct response to another
4989 message by an automatic process,
4991 - MUST NOT be used on manually-generated messages,
4993 - MAY be used on Delivery Status Notifications (DSNs) and Message
4994 Disposition Notifications (MDNs),
4996 - MUST NOT be used on messages generated by automatic or periodic
4997 processes, except for messages which are automatic responses to
5000 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5001 to the default Received: header definition.
5003 PH/12 Added log selector acl_warn_skipped (default on).
5005 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5006 variables because (a) they are in the wrong storage pool and (b) even if
5007 they were copied, it wouldn't work properly because of the caching.
5009 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5010 checking when decoding. Apparently there are clients that generate
5011 overlong encoded strings. Why am I not surprised?
5013 PH/15 If the first argument of "${if match_address" was not empty, but did not
5014 contain an "@" character, Exim crashed. Now it writes a panic log message
5015 and treats the condition as false.
5017 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5019 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5020 extension "envelope-auth". The code is finished and in agreement with
5021 other implementations, but there is no documentation so far and in fact,
5022 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5023 not changing the active code.
5025 Print executed "if" and "elsif" statements when debugging is used. This
5026 helps a great deal to understand what a filter does.
5028 Document more things not specified clearly in RFC3028. I had all this
5029 sorted out, when out of a sudden new issues came to my mind. Oops."
5031 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5034 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5035 canonical form (as documented). However, after a host name lookup from
5036 the IP address, check_host() was doing a simple string comparison with
5037 addresses acquired from the DNS when checking that the found name did
5038 have the original IP as one of its addresses. Since any found IPv6
5039 addresses are likely to be in abbreviated form, the comparison could
5040 fail. Luckily, there already exists a function for doing the comparison
5041 by converting both addresses to binary, so now that is used instead of
5042 the text comparison.
5044 PH/20 There was another similar case to PH/19, when a complete host name was
5045 given in a host list; looking up its IP address could give an abbreviated
5046 form, whereas the current host's name might or might not be abbreviated.
5047 The same fix has been applied.
5053 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5054 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5057 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5058 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5060 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5062 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5063 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5064 Also, for "Auto-Submitted", treat anything other than "no" as
5065 non-personal, in accordance with RFC 3834. (Previously it treated
5066 anything starting "auto-" as non-personal.)
5068 TF/01 The control=submission/name=... option had a problem with syntax
5069 errors if the name included a slash character. The /name= option
5070 now slurps the rest of the string, so it can include any characters
5071 but it must come last in the list of options (after /sender_retain
5074 PH/05 Some modifications to the interface to the fake nameserver for the new
5082 TK/01 Added the "success_on_redirect" address verification option. See
5083 NewStuff for rationale and an example.
5085 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5087 PH/02 Patch to exigrep to allow it to work on syslog lines.
5089 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5090 fread() to skip over the body file's header line, because in Cygwin the
5091 header line is locked and is inaccessible.
5093 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5094 co-exist for some time) to make it clear that it is the Exim ID that is
5095 referenced, not the Message-ID: header line.
5097 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5098 string_format() function, because snprintf() does not exist on all
5101 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5102 ${stat: expansion item.
5104 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5105 protocol synchronization error", to keep the pedants happy.
5107 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5108 well as for IRIX systems, when gcc is being used. See the host.c source
5111 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5113 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5116 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5117 transport and to the smtp transport in LMTP mode.
5119 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5121 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5122 run a filter in a subprocess. This could lead to confusion in subsequent
5123 lookups in the parent process. There should also be a search_tidyup() at
5124 the end of the subprocess.
5126 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5127 only if the host matched helo_try_verify_hosts, which caused the
5128 verification to occur when the EHLO/HELO command was issued. The ACL just
5129 tested the remembered result. Now, if a previous verification attempt has
5130 not happened, "verify = helo" does it there and then.
5132 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5134 TK/03 Fix log output including CR from clamd.
5136 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5137 exist provoked a memory error which could cause a segfault.
5139 PH/15 Installed PCRE 6.2
5141 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5143 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5144 of the problem. Specifically, suggested +O2 rather than +O1 for the
5147 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5149 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5150 record, the port was not correctly logged when the outgoing_port log
5151 selector was set (it logged the transort's default port).
5153 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5154 fallback_hosts, and "hosts" in the smtp transport.
5156 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5157 host errors such as "Connection refused".
5159 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5160 authentication with radiusclient 0.4.9:
5162 - Error returned from rc_read_config was caught wrongly
5163 - Username/password not passed on to radius server due to wrong length.
5165 The presumption is that some radiusclient API changes for 4.51/PH/17
5166 were not taken care of correctly. The code is still untested by me (my
5167 Linux distribution still has 0.3.2 of radiusclient), but it was
5168 contributed by a Radius user.
5170 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5171 expanding the "port" option of the smtp transport.
5173 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5174 while reading a MIME header. Thanks to Tom Hughes for a patch.
5176 PH/24 Include config.h inside local_scan.h so that configuration settings are
5179 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5180 that in an ACL for RCPT (for example), you can examine exactly what was
5183 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5184 commands, but it was not correctly comparing the address with the actual
5185 client host address. Thus, it would show the EHLO address in Received:
5186 header lines when this was not necessary.
5188 PH/27 Added the % operator to ${eval:}.
5190 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5191 it should be ignoring failures (because with -C, for example, it has lost
5192 privilege). It wasn't ignoring creation failures other than "already
5195 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5198 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5199 because an input line was too long, either on its own, or by virtue of
5200 too many continuations, the temporary file was not being removed, and the
5201 return code was incorrect.
5203 PH/31 Missing "BOOL" in function definition in filtertest.c.
5205 PH/32 Applied Sieve patches from the maintainer.
5207 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5209 PH/33 Added "verify = not_blind".
5211 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5212 Local/Makefile (with some defaults set). These are used in built scripts
5213 such as exicyclog, but they have never been used in the exim_install
5214 script (though there are many overriding facilities there). I have
5215 arranged that the exim_install script now takes note of these two
5218 PH/35 Installed configuration files for Dragonfly.
5220 PH/36 When a locally submitted message by a trusted user did not contain a
5221 From: header, and the sender address was obtained from -f or from an SMTP
5222 MAIL command, and the trusted user did not use -F to supply a sender
5223 name, $originator_name was incorrectly used when constructing a From:
5224 header. Furthermore, $originator_name was used for submission mode
5225 messages from external hosts without From: headers in a similar way,
5226 which is clearly wrong.
5228 PH/37 Added control=suppress_local_fixups.
5230 PH/38 When log_selector = +received_sender was set, and the addition of the
5231 sender made the log line's construction buffer exactly full, or one byte
5232 less than full, an overflow happened when the terminating "\n" was
5235 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5236 when the result of a list match is failure because a DNS lookup failed.
5238 PH/40 RM_COMMAND is now used in the building process.
5240 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5241 the "build-* directories that it finds.
5243 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5244 address in a domain literal was a prefix of an interface address.
5246 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5247 when verifying a sender address, unless rewrite_headers is false.
5249 PH/44 Wrote a long comment about why errors_to addresses are verified as
5250 recipients, not senders.
5252 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5253 the ratelimit ACL was added.
5255 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5257 PH/46 Added extra information about PostgreSQL errors to the error string.
5259 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5260 avoiding the need to install special zones in a real server. This is
5261 backwards compatible; if it can't find the fake resolver, it drops back.
5262 Thus, both old and new test suites can be run.
5264 TF/02 Added util/ratelimit.pl
5266 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5267 clock is set back in time.
5269 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5270 Candler <B.Candler@pobox.com>.
5272 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5273 OK for addresses that are the result of redirections.
5275 PH/48 A number of further additions for the benefit of the new test suite,
5276 including a fake gethostbyname() that interfaces to the fake DNS resolver
5279 TF/06 The fix for widen_domains has also been applied to qualify_single and
5280 search_parents which are the other dnslookup options that can cause
5283 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5286 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5288 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5289 service" if it encountered an MX record with an empty target hostname.
5290 The message is now "an MX or SRV record indicated no SMTP service".
5292 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5293 if the DNS of the sending site is misconfigured. This is quite a
5294 common situation. This change restores the behaviour of treating a
5295 helo verification defer as a failure.
5297 PH/51 If self=fail was set on a router, the bounce message did not include the
5298 actual error message.
5304 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5306 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5307 command itself ended in error, the underlying message about the transport
5308 filter timeout was being overwritten with the pipe command error. Now the
5309 underlying error message should be appended to the second error message.
5311 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5313 PH/02 Reduce the amount of output that "make" produces by default. Full output
5314 can still be requested.
5316 PH/03 The warning log line about a condition test deferring for a "warn" verb
5317 was being output only once per connection, rather than after each
5318 occurrence (because it was using the same function as for successful
5319 "warn" verbs). This seems wrong, so I have changed it.
5321 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5322 it should not have, which might have caused a crash in the right
5323 circumstances, but probably never did.
5325 PH/04 Installed a modified version of Tony Finch's patch to make submission
5326 mode fix the return path as well as the Sender: header line, and to
5327 add a /name= option so that you can make the user's friendly name appear
5330 TF/03 Added the control = fakedefer ACL modifier.
5332 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5333 Mark Lowes for thorough testing.
5335 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5337 TK/03 Merged latest SRS patch from Miles Wilton.
5339 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5340 with the definition in sysexits.h (which is #included earlier).
5341 Fortunately, Exim does not actually use EX_OK. The code used to try to
5342 preserve the sysexits.h value, by assuming that macro definitions were
5343 scanned for macro replacements. I have been disabused of this notion,
5344 so now the code just undefines EX_OK before #including unistd.h.
5346 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5347 in the smtp transport. When a block could not be written in a single
5348 write() function, the timeout was being re-applied to each part-write.
5349 This seems wrong - if the receiver was accepting one byte at a time it
5350 would take for ever. The timeout is now adjusted when this happens. It
5351 doesn't have to be particularly precise.
5353 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5354 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5356 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5357 without a domain if the check to <postmaster@domain> fails.
5359 SC/01 Eximstats: added -xls and the ability to specify output files
5360 (patch written by Frank Heydlauf).
5362 SC/02 Eximstats: use FileHandles for outputting results.
5364 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5366 SC/04 Eximstats: fixed display of large numbers with -nvr option
5368 SC/05 Eximstats: fixed merging of reports with empty tables.
5370 SC/06 Eximstats: added the -include_original_destination flag
5372 SC/07 Eximstats: removed tabs and trailing whitespace.
5374 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5376 TK/06 MBOX spool code: Add real "From " MBOX separator line
5377 so the .eml file is really in mbox format (even though
5378 most programs do not really care). Patch from Alex Miller.
5380 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5381 The latter is generated from $received_to and is only set if the
5382 message has one envelope recipient. SA can use these headers,
5383 obviously out-of-the-box. Patch from Alex Miller.
5385 PH/08 The ${def test on a variable was returning false if the variable's
5386 value was "0", contrary to what the specification has always said!
5387 The result should be true unless the variable is empty.
5389 PH/09 The syntax error of a character other than { following "${if
5390 def:variable_name" (after optional whitespace) was not being diagnosed.
5391 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5392 accidental colon was present, for example, could give incorrect results.
5394 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5395 result is used (not including appendfile, where other changes are about
5398 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5399 This involved changing a lot of size variables from int to off_t. It
5400 should work with maildirs and everything.
5402 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5403 spamd dying while we are connected to it.
5405 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5408 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5409 function for BDB 4.3.
5411 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5413 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5414 the sources for PCRE have been re-arranged and more files are now
5417 PH/15 The code I had for printing potentially long long variables in PH/11
5418 above was not the best (it lost precision). The length of off_t variables
5419 is now inspected at build time, and an appropriate printing format (%ld
5420 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5421 to be "long long int" or "long int". This is needed for the internal
5422 formatting function string_vformat().
5424 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5425 the configuration file to be ":syslog", then the script "guesses" where
5426 the logs files are, rather than using the compiled in default. In our
5427 case the guess is not the same as the compiled default, so the script
5428 suddenly stopped working when I started to use syslog. The patch checks
5429 to see if log_file_path is "". If so, it attempts to read it from exim
5430 with no configuration file to get the compiled in version, before it
5431 falls back to the previous guessing code."
5433 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5434 implementing BATV in an Exim configuration. See NewStuff for the gory
5437 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5438 Makefile that are specific to HP-UX.
5440 PH/18 If the "use_postmaster" option was set for a recipient callout together
5441 with the "random" option, the postmaster address was used as the MAIL
5442 FROM address for the random test, but not for the subsequent recipient
5443 test. It is now used for both.
5445 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5446 patch removes a few documentation additions to RFC 3028, because the
5447 latest draft now contains them. It adds the new en;ascii-case comparator
5448 and a new error check for 8bit text in MIME parts. Comparator and
5449 require names are now matched exactly. I enabled the subaddress
5450 extension, but it is not well tested yet (read: it works for me)."
5452 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5453 rework some of the code of TK/09 above to avoid the hardwired use of
5454 "%lld" and "long long". Replaced the call to snprintf() with a call to
5457 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5458 records point to non-existent hosts", "retry timeout exceeded", and
5459 "retry time not reached for any host after a long failure period".
5461 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5462 experimental DomainKeys support:
5464 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5465 (2) On an error such as an illegally used "control", the wrong name for
5466 the control was given.
5468 These problems did NOT occur unless DomainKeys support was compiled.
5470 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5472 PH/24 Added ${if match_ip condition.
5474 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5475 databases so that it will be absolutely obvious if a crash occurs in the
5476 DB library. This is a regular occurrence (often caused by mis-matched
5479 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5480 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5481 picked up on a user's system that detects such things. There doesn't seem
5482 to be a gcc warning option for this - only an attribute that has to be
5483 put on the function's prototype. It seems that in Fedora Core 4 they have
5484 set this on a number of new functions. No doubt there will be more in due
5487 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5488 specify a transport. However, if an address that was verified by such a
5489 router was the subject of a callout, Exim crashed because it tried to
5490 read the rcpt_include_affixes from the non-existent transport. Now it
5491 just assumes that the setting of that option is false. This bug was
5492 introduced by 4.51/PH/31.
5494 PH/28 Changed -d+all to exclude +memory, because that information is very
5495 rarely of interest, but it makes the output a lot bigger. People tend to
5496 do -d+all out of habit.
5498 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5499 code in os-type was giving problems when libc.so lives in lib64, like on
5502 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5503 aren't the modern standard, and it seems that some systems' include files
5504 don't always have them. Exim was already checking for some of the newer
5505 ones like T_AAAA, and defining it itself. I've added checks for all the
5506 record types that Exim uses.
5508 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5509 not automatically generating a new one, as it is supposed to. This
5510 prevented TLS from working. If the file did exist, but contained invalid
5511 data, a new version was generated, as expected. It was only the case of a
5512 non-existent file that was broken.
5514 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5515 with a change in libdomainkeys > 0.64.
5517 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5518 from DNS. If the selector record carries the flag, it now has
5519 precedence over the domain-wide flag.
5521 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5523 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5524 the use of a port name that isn't defined in /etc/services) occurred, the
5525 message was deferred as in a normal delivery, and thus remained on the
5526 spool, instead of being failed because of the mua_wrapper setting. This
5527 is now fixed, and I tidied up some of the mua_wrapper messages at the
5530 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5531 the messages in a hash of arrays rather than using individual hashes.
5532 This is a bit cleaner and results in dramatic memory savings, albeit
5533 at a slight CPU cost.
5535 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5536 as requested by Marc Sherman.
5538 SC/10 Eximstats: added histograms for user specified patterns as requested
5541 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5543 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5544 fopen() in the content-scanning modules that did not already have it.
5550 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5551 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5553 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5555 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5557 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5558 to test Sieve filters that use "vacation".
5560 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5561 that changes the way the GnuTLS parameters are stored in the cache file.
5562 The new format can be generated externally. For backward compatibility,
5563 if the data in the cache doesn't make sense, Exim assumes it has read an
5564 old-format file, and it generates new data and writes a new file. This
5565 means that you can't go back to an older release without removing the
5568 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5569 work if there are any delivery delays because "one_time" forces the
5570 parent to be marked "delivered", so its unseen clone is never tried
5571 again. For this reason, Exim now forbids the simultaneous setting of
5574 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5575 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5576 entry if you want to know the details. The fix, however, neglected to
5577 consider the case when local delivery batching is involved. The test for
5578 "previously delivered" was not happening when checking to see if an
5579 address could be batched with a previous (undelivered) one; under
5580 certain circumstances this could lead to multiple deliveries to the same
5583 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5584 in its include files, and this causes problems building Exim.
5586 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5587 header_syntax) but Exim was just ignoring anything given after a slash.
5588 In particular, this caused confusion with an attempt to use "verify =
5589 reverse_host_lookup/defer_ok". An error is now given when options are
5590 supplied for verify items that do not have them. (Maybe reverse_host_
5591 lookup should have a defer_ok option, but that's a different point.)
5593 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5594 defined by RFC 821) to 2048, because there were problems with some AUTH
5595 commands, and RFC 1869 says the size should be increased for extended
5596 SMTP commands that take arguments.
5598 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5601 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5602 "unknown" error; now it says that the functionality isn't in the binary.
5604 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5605 an address' error message when a string expansion fails (syntax or
5606 whatever). Otherwise the password may appear in the log. Following change
5607 PH/42 below, there is no longer a chance of it appearing in a bounce
5610 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5612 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5613 of its addresses was ever tried. (Bugzilla bug #2.)
5615 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5616 the result incorrectly in the debug output. (It correctly added a newline
5617 to what was transported.)
5619 TF/01 Added $received_time.
5621 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5622 commented out examples of how to interface to a virus scanner and to
5623 SpamAssassin. Also added commented examples of av_scanner and
5624 spamd_address settings.
5626 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5627 and controls are allowed in which ACLs. There were a couple of minor
5628 errors. Some of the entries in the conditions table (which is a table of
5629 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5630 negation of where the condition IS allowed.
5632 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5634 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5635 header file does not have a version number, so I've had to invent a new
5636 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5637 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5638 radiusclient), but it was contributed by a Radius user.
5640 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5641 files or directories, for OpenSSL.
5643 PH/19 When an Exim process that is running as root has to create an Exim log
5644 file, it does so in a subprocess that runs as exim:exim so as to get the
5645 ownership right at creation (otherwise, other Exim processes might see
5646 the file with the wrong ownership). There was no test for failure of this
5647 fork() call, which would lead to the process getting stuck as it waited
5648 for a non-existent subprocess. Forks do occasionally fail when resources
5649 run out. I reviewed all the other calls to fork(); they all seem to check
5652 PH/20 When checking for unexpected SMTP input at connect time (before writing
5653 the banner), Exim was not dealing correctly with a non-positive return
5654 from the read() function. If the client had disconnected by this time,
5655 the result was a log entry for a synchronization error with an empty
5656 string after "input=" when read() returned zero. If read() returned -1
5657 (an event I could not check), uninitialized data bytes were printed.
5658 There were reports of junk text (parts of files, etc) appearing after
5661 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5663 PH/22 Added support for macro redefinition, and (re)definition in between
5664 driver and ACL definitions.
5666 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5667 forgetting to use the resulting value; it was using the unexpanded value.
5669 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5670 hadn't been configured. The fix is from Juergen Kreileder, who
5671 understands it better than I do:
5673 "Here's what I see happening with three configured cyrus_sasl
5674 authenticators configured (plain, login, cram-md5):
5676 On startup auth_cyrus_sasl_init() gets called for each of these.
5677 This means three calls to sasl_listmech() without a specified mech_list.
5678 => SASL tests which mechs of all available mechs actually work
5679 => three warnings about OTP not working
5680 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5682 With the patch, sasl_listmech() also gets called three times. But now
5683 SASL's mech_list option is set to the server_mech specified in the the
5684 authenticator. Or in other words, the answer from sasl_listmech()
5685 gets limited to just the mech you're testing for (which is different
5687 => the return list contains just 'plain' or 'login', 'cram-md5' or
5688 nothing depending on the value of ob->server_mech.
5690 I've just tested the patch: Authentication still works fine,
5691 unavailable mechs specified in the exim configuration are still
5692 caught, and the auth.log warnings about OTP are gone."
5694 PH/25 When debugging is enabled, the contents of the command line are added
5695 to the debugging output, even when log_selector=+arguments is not
5698 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5699 answer is "GNU", and only if the return is "GNU/something" is the answer
5702 PH/27 $acl_verify_message is now set immediately after the failure of a
5703 verification in an ACL, and so is available in subsequent modifiers. In
5704 particular, the message can be preserved by coding like this:
5706 warn !verify = sender
5707 set acl_m0 = $acl_verify_message
5709 Previously, $acl_verify_message was set only while expanding "message"
5710 and "log_message" when a very denied access.
5712 PH/28 Modified OS/os.c-Linux with
5714 -#ifndef OS_LOAD_AVERAGE
5715 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5717 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5718 nomenclature these days.)
5720 PH/29 Installed patch from the Sieve maintainer that adds the options
5721 sieve_useraddress and sieve_subaddress to the redirect router.
5723 PH/30 In these circumstances:
5724 . Two addresses routed to the same list of hosts;
5725 . First host does not offer TLS;
5726 . First host accepts first address;
5727 . First host gives temporary error to second address;
5728 . Second host offers TLS and a TLS session is established;
5729 . Second host accepts second address.
5730 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5731 and peerdn, if requested) that were in fact used only for the second
5734 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5735 attention to any local part prefix or suffix that was matched by the
5736 router that accepted the address. It now behaves in the same way as it
5737 does for delivery: the affixes are removed from the local part unless
5738 rcpt_include_affixes is set on the transport.
5740 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5741 timeout during the DATA phase of an incoming message.
5743 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5744 applied a patch sanctioned by the Sieve maintainer.
5746 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5747 the uid or gid is negative. A case of a negative gid caused this to be
5748 noticed. The fix allows for either to be negative.
5750 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5751 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5754 PH/36 The change PH/12 above was broken. Fixed it.
5756 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5757 the grounds that routing the same address twice would always produce the
5758 same answer. This might have been true once, but it is certainly no
5759 longer true now. Routing a child address may depend on the previous
5760 routing that produced that child. Some complicated redirection strategies
5761 went wrong when messages had multiple recipients, and made Exim's
5762 behaviour dependent on the order in which the addresses were given.
5764 I have moved the duplicate checking until after the routing is complete.
5765 Exim scans the addresses that are assigned to local and remote
5766 transports, and removes any duplicates. This means that more work will be
5767 done, as duplicates will always all be routed, but duplicates are
5768 presumably rare, so I don't expect this is of any significance.
5770 For deliveries to pipes, files, and autoreplies, the duplicate checking
5771 still happens during the routing process, since they are not going to be
5774 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5775 It corrects a timeout issue with spamd. This is Ian's comment: "The
5776 background is that sometimes spamd either never reads data from a
5777 connection it has accepted, or it never writes response data. The exiscan
5778 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5779 blindly assumes that writes won't block so it may never time out."
5781 PH/39 Allow G after quota size as well as K and M.
5783 PH/40 The value set for $authenticated_id in an authenticator may not contain
5784 binary zeroes or newlines because the value is written to log lines and
5785 to spool files. There was no check on this. Now the value is run through
5786 the string_printing() function so that such characters are converted to
5787 printable escape sequences.
5789 PH/41 $message_linecount is a new variable that contains the total number of
5790 lines in the message. Compare $body_linecount, which is the count for the
5793 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5794 bounce and delay warning messages, except in certain special cases, which
5797 (a) An SMTP error message from a remote host;
5798 (b) A message specified in a :fail: redirection;
5799 (c) A message specified in a "fail" command in a system filter;
5800 (d) A message specified in a FAIL return from the queryprogram router;
5801 (e) A message specified by the cannot_route_message router option.
5803 In these cases only, Exim does include the error details in bounce and
5804 warning messages. There are also a few cases where bland messages such
5805 as "unrouteable address" or "local delivery error" are given.
5807 PH/43 $value is now also set for the "else" part of a ${run expansion.
5809 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5810 being worked on, but at least Exim now implements the latest version to
5813 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5814 process to complete was treated as a delivery failure, a timeout while
5815 writing the message to the pipe was logged, but erroneously treated as a
5816 successful delivery. Such timeouts include transport filter timeouts. For
5817 consistency with the overall process timeout, these timeouts are now
5818 treated as errors, giving rise to delivery failures by default. However,
5819 there is now a new Boolean option for the pipe transport called
5820 timeout_defer, which, if set TRUE, converts the failures into defers for
5821 both kinds of timeout. A transport filter timeout is now identified in
5824 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5825 systems where "make" and "gmake" are different, calling "gmake" at top
5826 level broke things. I've arranged for the value of $(MAKE) to be passed
5827 from the Makefile to this script so that it can call the same version of
5831 A note about Exim versions 4.44 and 4.50
5832 ----------------------------------------
5834 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5835 changes of various kinds. As a consequence, a big documentation update was
5836 needed. This delayed the release for rather longer than seemed good, especially
5837 in the light of a couple of (minor) security issues. Therefore, the changes
5838 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5839 release. So 4.44 and 4.50 are in effect two different branches that both start
5842 I have left the 4.50 change log unchanged; it contains all the changes since
5843 4.43. The change log for 4.44 is below; many of its items are identical to
5844 those for 4.50. This seems to be the most sensible way to preserve the
5845 historical information.
5851 1. Minor wording change to the doc/README.SIEVE file.
5853 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5854 computation of the current number of files was incorrect.
5856 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5857 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5860 4. Give more explanation in the error message when the command for a transport
5861 filter fails to execute.
5863 5. There are several places where Exim runs a non-Exim command in a
5864 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5865 was being done only for the command run by the queryprogram router. It is
5866 now done for all such subprocesses. The other cases are: ${run, transport
5867 filters, and the commands run by the lmtp and pipe transports.
5869 6. Added CONFIGURE_GROUP build-time option.
5871 7. Some older OS have a limit of 256 on the maximum number of file
5872 descriptors. Exim was using setrlimit() to set 1000 as a large value
5873 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5874 systems. I've change it so that if it can't get 1000, it tries for 256.
5876 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5877 was an oversight, and furthermore, ever since the addition of extra
5878 controls (e.g. 4.43/32), the checks on when to allow different forms of
5879 "control" were broken. There should now be diagnostics for all cases when a
5880 control that does not make sense is encountered.
5882 9. Added the /retain_sender option to "control=submission".
5884 10. $recipients is now available in the predata ACL (oversight).
5886 11. Tidy the search cache before the fork to do a delivery from a message
5887 received from the command line. Otherwise the child will trigger a lookup
5888 failure and thereby defer the delivery if it tries to use (for example) a
5889 cached ldap connection that the parent has called unbind on.
5891 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5892 of $address_data from the recipient verification was clobbered by the
5893 sender verification.
5895 13. The value of address_data from a sender verification is now available in
5896 $sender_address_data in subsequent conditions in the ACL statement.
5898 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5900 15. Added a new option "connect=<time>" to callout options, to set a different
5903 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5904 was its contents. (It was OK if the option was not defined at all.)
5906 17. A "Completed" log line is now written for messages that are removed from
5907 the spool by the -Mrm option.
5909 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5910 information about exactly what failed.
5912 19. Added -dd to debug only the daemon process.
5914 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5915 handles timeouts, both on the server side and network timeouts. Renamed the
5916 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5918 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5919 It is now set to "smtps".
5921 22. $host_address is now set to the target address during the checking of
5922 ignore_target_hosts.
5924 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5925 being passed; this would have caused $sender_host_name to have been used if
5926 matching the list had actually called for a host name (not very likely,
5927 since this list is usually IP addresses). A host name is now passed as
5930 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5931 code that specifies a non-restarting handler (typically sigaction() in
5932 modern systems) in an attempt to fix a rare and obscure crash bug.
5934 25. Narrowed the window for a race in the daemon that could cause it to ignore
5935 SIGCHLD signals. This is not a major problem, because they are used only to
5936 wake it up if nothing else does.
5938 26. A malformed maildirsize file could cause Exim to calculate negative values
5939 for the mailbox size or file count. Odd effects could occur as a result.
5940 The maildirsize information is now recalculated if the size or filecount
5943 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5944 support for a long time. Removed HAVE_SYS_VFS_H.
5946 28. Installed the latest version of exipick from John Jetmore.
5948 29. In an address list, if the pattern was not a regular expression, an empty
5949 subject address (from a bounce message) matched only if the pattern was an
5950 empty string. Non-empty patterns were not even tested. This was the wrong
5951 because it is perfectly reasonable to use an empty address as part of a
5952 database query. An empty address is now tested by patterns that are
5953 lookups. However, all the other forms of pattern expect the subject to
5954 contain a local part and a domain, and therefore, for them, an empty
5955 address still always fails if the pattern is not itself empty.
5957 30. Exim went into a mad DNS loop when attempting to do a callout where the
5958 host was specified on an smtp transport, and looking it up yielded more
5959 than one IP address.
5961 31. Re-factored the code for checking spool and log partition space into a
5962 function that finds that data and another that does the check. The former
5963 is then used to implement four new variables: $spool_space, $log_space,
5964 $spool_inodes, and $log_inodes.
5966 32. The RFC2047 encoding function was originally intended for short strings
5967 such as real names; it was not keeping to the 75-character limit for
5968 encoded words that the RFC imposes. It now respects the limit, and
5969 generates multiple encoded words if necessary. To be on the safe side, I
5970 have increased the buffer size for the ${rfc2047: expansion operator from
5973 33. It is now permitted to omit both strings after an "if" condition; if the
5974 condition is true, the result is "true". As before, when the second string
5975 is omitted, a false condition yields an empty string. This makes it less
5976 cumbersome to write custom ACL and router conditions.
5978 34. Failure to deliver a bounce message always caused it to be frozen, even if
5979 there was an errors_to setting on the router. The errors_to setting is now
5982 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5983 canonical form (fully expanded) before being placed in
5984 $sender_host_address.
5986 36. The table in the code that translates DNS record types into text (T_A to
5987 "A" for instance) was missing entries for NS and CNAME. It is just possible
5988 that this could have caused confusion if both these types were looked up
5989 for the same domain, because the text type is used as part of Exim's
5990 per-process caching. But the chance of anyone hitting this buglet seems
5993 37. The dnsdb lookup has been extended in a number of ways.
5995 (1) There is a new type, "zns", which walks up the domain tree until it
5996 finds some nameserver records. It should be used with care.
5998 (2) There is a new type, "mxh", which is like "mx" except that it returns
5999 just the host names, not the priorities.
6001 (3) It is now possible to give a list of domains (or IP addresses) to be
6002 looked up. The behaviour when one of the lookups defers can be
6003 controlled by a keyword.
6005 (4) It is now possible to specify the separator character for use when
6006 multiple records are returned.
6008 38. The dnslists ACL condition has been extended: it is now possible to supply
6009 a list of IP addresses and/or domains to be looked up in a particular DNS
6012 39. Added log_selector=+queue_time_overall.
6014 40. When running the queue in the test harness, wait just a tad after forking a
6015 delivery process, to get repeatability of debugging output.
6017 41. Include certificate and key file names in error message when GnuTLS fails
6018 to set them up, because the GnuTLS error message doesn't include the name
6019 of the failing file when there is a problem reading it.
6021 42. Allow both -bf and -bF in the same test run.
6023 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6025 44. The "Exiscan patch" is now merged into the mainline Exim source.
6027 45. Sometimes the final signoff response after QUIT could fail to get
6028 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6029 < 0 before doing a fflush(). This bug looks as though it goes back to the
6030 introduction of TLS in release 3.20, but "sometimes" must have been rare
6031 because the tests only now provoked it.
6033 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6034 (this can affect the format of dates).
6036 47. exim_tidydb, when checking for the continued existence of a message for
6037 which it has found a message-specific retry record, was not finding
6038 messages that were in split spool directories. Consequently, it was
6039 deleting retry records that should have stayed in existence.
6041 48. Steve fixed some bugs in eximstats.
6043 49. The SPA authentication driver was not abandoning authentication and moving
6044 on to the next authenticator when an expansion was forced to fail,
6045 contradicting the general specification for all authenticators. Instead it
6046 was generating a temporary error. It now behaves as specified.
6048 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6049 (the order specifies the preference for clients). The order is now AES256,
6050 AES128, 3DES, ARCFOUR128.
6052 51. Small patch to Sieve code - explicitly set From: when generating an
6055 52. Exim crashed if a remote delivery caused a very long error message to be
6056 recorded - for instance if somebody sent an entire SpamAssassin report back
6057 as a large number of 550 error lines. This bug was coincidentally fixed by
6058 increasing the size of one of Exim's internal buffers (big_buffer) that
6059 happened as part of the Exiscan merge. However, to be on the safe side, I
6060 have made the code more robust (and fixed the comments that describe what
6063 53. Now that there can be additional text after "Completed" in log lines (if
6064 the queue_time_overall log selector is set), a one-byte patch to exigrep
6065 was needed to allow it to recognize "Completed" as not the last thing in
6068 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6069 patch that reportedly fixes this has been added. I am not expert enough to
6070 create a test for it. This is what the patch creator wrote:
6072 "I found a little strange behaviour of ldap code when working with
6073 Windows 2003 AD Domain, where users was placed in more than one
6074 Organization Units. When I tried to give exim partial DN, the exit code
6075 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6076 But simultaneously result of request was absolutely normal ldap result,
6077 so I produce this patch..."
6079 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6080 REFERENCE, so I have modified the code to exclude the patch when that macro
6083 55. Some experimental protocols are using DNS PTR records for new purposes. The
6084 keys for these records are domain names, not reversed IP addresses. The
6085 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6086 leaves it alone. Component reversal etc. now happens only for IP addresses.
6089 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6091 57. Double the size of the debug message buffer (to 2048) so that more of very
6092 long debug lines gets shown.
6094 58. The exicyclog utility now does better if the number of log files to keep
6095 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6097 59. Two changes related to the smtp_active_hostname option:
6099 (1) $smtp_active_hostname is now available as a variable.
6100 (2) The default for smtp_banner uses $smtp_active_hostname instead
6101 of $primary_hostname.
6103 60. The host_aton() function is supposed to be passed a string that is known
6104 to be a valid IP address. However, in the case of IPv6 addresses, it was
6105 not checking this. This is a hostage to fortune. Exim now panics and dies
6106 if the condition is not met. A case was found where this could be provoked
6107 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6108 components; fortuitously, this particular loophole had already been fixed
6109 by change 4.50/55 above.
6111 If there are any other similar loopholes, the new check in host_aton()
6112 itself should stop them being exploited. The report I received stated that
6113 data on the command line could provoke the exploit when Exim was running as
6114 exim, but did not say which command line option was involved. All I could
6115 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6116 running as the user.
6119 61. There was a buffer overflow vulnerability in the SPA authentication code
6120 (which came originally from the Samba project). I have added a test to the
6121 spa_base64_to_bits() function which I hope fixes it.
6124 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6125 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6127 63. The daemon start-up calls getloadavg() while still root for those OS that
6128 need the first call to be done as root, but it missed one case: when
6129 deliver_queue_load_max is set with deliver_drop_privilege. This is
6130 necessary for the benefit of the queue runner, because there is no re-exec
6131 when deliver_drop_privilege is set.
6133 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6134 This has been fixed.
6136 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6137 was in use, was not putting the data itself into the right store pool;
6138 consequently, it could be overwritten for a subsequent message in the same
6139 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6142 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6144 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6145 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6146 the function were treating the return as a boolean value, which happened to
6147 work because 0=false and not-0=true, but is not correct code.
6149 68. The host_aton() function was not handling scoped IPv6 addresses (those
6150 with, for example, "%eth0" on the end) correctly.
6152 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6153 negated items (that is, ~something) in unsigned ints. Some compilers
6154 apparently mutter when there is no cast.
6156 70. If an address verification called from an ACL failed, and did not produce a
6157 user-specific message (i.e. there was only a "system" message), nothing was
6158 put in $acl_verify_message. In this situation, it now puts the system
6161 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6162 session; change 4.31/43 added the unwanted input to the log line - except
6163 that it did not do this in the start of session case. It now does.
6165 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6166 This is wrong and can cause the other end to generate a synchronization
6167 error if it is another Exim or anything else that does the synchronization
6168 check. A QUIT command is no longer sent after a timeout.
6170 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6171 during host lookups.
6173 74. The defer_ok option of callout verification was not working if it was used
6174 when verifying addresses in header lines, that is, for this case:
6176 verify = header_sender/callout=defer_ok
6178 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6179 those file descriptors could be used for SMTP connections. If anything
6180 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6181 could be sent to the SMTP client, causing chaos. The daemon now opens
6182 stdin, stdout, and stderr to /dev/null when it puts itself into the
6185 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6186 log by default. The user can override this with suitable Perl magic.
6188 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6189 the log message when discard triggers, was not working for the DATA ACL or
6190 for the non-SMTP ACL.
6192 78. Error message wording change in sieve.c.
6194 79. If smtp_accept_max_per_host was set, the number of connections could be
6195 restricted to fewer than expected, because the daemon was trying to set up
6196 a new connection before checking whether the processes handling previous
6197 connections had finished. The check for completed processes is now done
6198 earlier. On busy systems, this bug wouldn't be noticed because something
6199 else would have woken the daemon, and it would have reaped the completed
6202 80. If a message was submitted locally by a user whose login name contained one
6203 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6204 It caused a spool format error. I have fixed the spool reading code. A
6205 related problem was that the "from" clause in the Received: line became
6206 illegal because of the space(s). It is now covered by ${quote_local_part.
6208 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6211 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6213 83. Patch from Sieve maintainer to fix unterminated string problem in
6214 "vacation" handling.
6216 84. Some minor changes to the Linux configuration files to help with other
6217 OS variants using glibc.
6219 85. One more patch for Sieve to update vacation handling to latest spec.
6222 ----------------------------------------------------
6223 See the note above about the 4.44 and 4.50 releases.
6224 ----------------------------------------------------
6230 1. Change 4.43/35 introduced a bug that caused file counts to be
6231 incorrectly computed when quota_filecount was set in an appendfile
6234 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6235 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6238 3. Give more explanation in the error message when the command for a transport
6239 filter fails to execute.
6241 4. There are several places where Exim runs a non-Exim command in a
6242 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6243 was being done only for the command run by the queryprogram router. It is
6244 now done for all such subprocesses. The other cases are: ${run, transport
6245 filters, and the commands run by the lmtp and pipe transports.
6247 5. Some older OS have a limit of 256 on the maximum number of file
6248 descriptors. Exim was using setrlimit() to set 1000 as a large value
6249 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6250 systems. I've change it so that if it can't get 1000, it tries for 256.
6252 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6253 was an oversight, and furthermore, ever since the addition of extra
6254 controls (e.g. 4.43/32), the checks on when to allow different forms of
6255 "control" were broken. There should now be diagnostics for all cases when a
6256 control that does not make sense is encountered.
6258 7. $recipients is now available in the predata ACL (oversight).
6260 8. Tidy the search cache before the fork to do a delivery from a message
6261 received from the command line. Otherwise the child will trigger a lookup
6262 failure and thereby defer the delivery if it tries to use (for example) a
6263 cached ldap connection that the parent has called unbind on.
6265 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6266 of $address_data from the recipient verification was clobbered by the
6267 sender verification.
6269 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6270 was its contents. (It was OK if the option was not defined at all.)
6272 11. A "Completed" log line is now written for messages that are removed from
6273 the spool by the -Mrm option.
6275 12. $host_address is now set to the target address during the checking of
6276 ignore_target_hosts.
6278 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6279 being passed; this would have caused $sender_host_name to have been used if
6280 matching the list had actually called for a host name (not very likely,
6281 since this list is usually IP addresses). A host name is now passed as
6284 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6285 code that specifies a non-restarting handler (typically sigaction() in
6286 modern systems) in an attempt to fix a rare and obscure crash bug.
6288 15. Narrowed the window for a race in the daemon that could cause it to ignore
6289 SIGCHLD signals. This is not a major problem, because they are used only to
6290 wake it up if nothing else does.
6292 16. A malformed maildirsize file could cause Exim to calculate negative values
6293 for the mailbox size or file count. Odd effects could occur as a result.
6294 The maildirsize information is now recalculated if the size or filecount
6297 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6298 support for a long time. Removed HAVE_SYS_VFS_H.
6300 18. Updated exipick to current release from John Jetmore.
6302 19. Allow an empty sender to be matched against a lookup in an address list.
6303 Previously the only cases considered were a regular expression, or an
6306 20. Exim went into a mad DNS lookup loop when doing a callout where the
6307 host was specified on the transport, if the DNS lookup yielded more than
6310 21. The RFC2047 encoding function was originally intended for short strings
6311 such as real names; it was not keeping to the 75-character limit for
6312 encoded words that the RFC imposes. It now respects the limit, and
6313 generates multiple encoded words if necessary. To be on the safe side, I
6314 have increased the buffer size for the ${rfc2047: expansion operator from
6317 22. Failure to deliver a bounce message always caused it to be frozen, even if
6318 there was an errors_to setting on the router. The errors_to setting is now
6321 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6322 canonical form (fully expanded) before being placed in
6323 $sender_host_address.
6325 24. Updated eximstats to version 1.33
6327 25. Include certificate and key file names in error message when GnuTLS fails
6328 to set them up, because the GnuTLS error message doesn't include the name
6329 of the failing file when there is a problem reading it.
6331 26. Expand error message when OpenSSL has problems setting up cert/key files.
6334 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6335 (this can affect the format of dates).
6337 28. exim_tidydb, when checking for the continued existence of a message for
6338 which it has found a message-specific retry record, was not finding
6339 messages that were in split spool directories. Consequently, it was
6340 deleting retry records that should have stayed in existence.
6342 29. eximstats updated to version 1.35
6343 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6344 1.35 - bugfix such that pie charts by volume are generated correctly
6346 30. The SPA authentication driver was not abandoning authentication and moving
6347 on to the next authenticator when an expansion was forced to fail,
6348 contradicting the general specification for all authenticators. Instead it
6349 was generating a temporary error. It now behaves as specified.
6351 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6352 (the order specifies the preference for clients). The order is now AES256,
6353 AES128, 3DES, ARCFOUR128.
6355 31. Small patch to Sieve code - explicitly set From: when generating an
6358 32. Exim crashed if a remote delivery caused a very long error message to be
6359 recorded - for instance if somebody sent an entire SpamAssassin report back
6360 as a large number of 550 error lines. This bug was coincidentally fixed by
6361 increasing the size of one of Exim's internal buffers (big_buffer) that
6362 happened as part of the Exiscan merge. However, to be on the safe side, I
6363 have made the code more robust (and fixed the comments that describe what
6366 33. Some experimental protocols are using DNS PTR records for new purposes. The
6367 keys for these records are domain names, not reversed IP addresses. The
6368 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6369 leaves it alone. Component reversal etc. now happens only for IP addresses.
6372 34. The host_aton() function is supposed to be passed a string that is known
6373 to be a valid IP address. However, in the case of IPv6 addresses, it was
6374 not checking this. This is a hostage to fortune. Exim now panics and dies
6375 if the condition is not met. A case was found where this could be provoked
6376 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6377 components; fortuitously, this particular loophole had already been fixed
6378 by change 4.50/55 or 4.44/33 above.
6380 If there are any other similar loopholes, the new check in host_aton()
6381 itself should stop them being exploited. The report I received stated that
6382 data on the command line could provoke the exploit when Exim was running as
6383 exim, but did not say which command line option was involved. All I could
6384 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6385 running as the user.
6388 35. There was a buffer overflow vulnerability in the SPA authentication code
6389 (which came originally from the Samba project). I have added a test to the
6390 spa_base64_to_bits() function which I hope fixes it.
6393 36. The daemon start-up calls getloadavg() while still root for those OS that
6394 need the first call to be done as root, but it missed one case: when
6395 deliver_queue_load_max is set with deliver_drop_privilege. This is
6396 necessary for the benefit of the queue runner, because there is no re-exec
6397 when deliver_drop_privilege is set.
6399 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6400 was in use, was not putting the data itself into the right store pool;
6401 consequently, it could be overwritten for a subsequent message in the same
6402 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6405 38. Sometimes the final signoff response after QUIT could fail to get
6406 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6407 < 0 before doing a fflush(). This bug looks as though it goes back to the
6408 introduction of TLS in release 3.20, but "sometimes" must have been rare
6409 because the tests only now provoked it.
6415 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6416 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6417 Now it returns an integer. A number of calls were still expecting a T/F
6418 return. Fortuitously, in all cases, the tests worked in OK situations,
6419 which is the norm. However, things would have gone wrong on any write
6420 failures on the smtp file descriptor. This function is used when sending
6421 messages over SMTP and also when doing verify callouts.
6423 2. When Exim is called to do synchronous delivery of a locally submitted
6424 message (the -odf or -odi options), it no longer closes stderr before doing
6427 3. Implemented the mua_wrapper option.
6429 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6431 5. Implemented the functions header_remove(), header_testname(),
6432 header_add_at_position(), and receive_remove_recipient(), and exported them
6435 6. If an ACL "warn" statement specified the addition of headers, Exim already
6436 inserted X-ACL-Warn: at the start if there was no header name. However, it
6437 was not making this test for the second and subsequent header lines if
6438 there were newlines in the string. This meant that an invalid header could
6439 be inserted if Exim was badly configured.
6441 7. Allow an ACL "warn" statement to add header lines at the start or after all
6442 the Received: headers, as well as at the end.
6444 8. Added the rcpt_4xx retry error code.
6446 9. Added postmaster_mailfrom=xxx to callout verification option.
6448 10. Added mailfrom=xxxx to the callout verification option, for verify=
6451 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6452 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6454 12. Inserted some casts to stop certain compilers warning when using pointer
6455 differences as field lengths or precisions in printf-type calls (mostly
6456 affecting debugging statements).
6458 13. Added optional readline() support for -be (dynamically loaded).
6460 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6461 same clock tick as a message's arrival, so that its received time was the
6462 same as the "first fail" time on the retry record, and that message
6463 remained on the queue past the ultimate address timeout, every queue runner
6464 would try a delivery (because it was past the ultimate address timeout) but
6465 after another failure, the ultimate address timeout, which should have then
6466 bounced the address, did not kick in. This was a "< instead of <=" error;
6467 in most cases the first failure would have been in the next clock tick
6468 after the received time, and all would be well.
6470 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6471 being recognized when the domain list was tested by the match_domain
6472 condition in an expansion string.
6474 16. Added the ${str2b64: operator.
6476 17. Exim was always calling setrlimit() to set a large limit for the number of
6477 processes, without checking whether the existing limit was already
6478 adequate. (It did check for the limit on file descriptors.) Furthermore,
6479 errors from getrlimit() and setrlimit() were being ignored. Now they are
6480 logged to the main and panic logs, but Exim does carry on, to try to do its
6481 job under whatever limits there are.
6483 18. Imported PCRE 5.0.
6485 19. Trivial typo in log message " temporarily refused connection" (the leading
6488 20. If the log selector return_path_on_delivery was set and an address was
6489 redirected to /dev/null, the delivery process crashed because it assumed
6490 that a return path would always be set for a "successful" delivery. In this
6491 case, the whole delivery is bypassed as an optimization, and therefore no
6494 21. Internal re-arrangement: the function for sending a challenge and reading
6495 a response while authentication was assuming a zero-terminated challenge
6496 string. It's now changed to take a pointer and a length, to allow for
6497 binary data in such strings.
6499 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6501 23. Exim was not respecting finduser_retries when seeking the login of the
6502 uid under which it was called; it was always trying 10 times. (The default
6503 setting of finduser_retries is zero.) Also, it was sleeping after the final
6504 failure, which is pointless.
6506 24. Implemented tls_on_connect_ports.
6508 25. Implemented acl_smtp_predata.
6510 26. If the domain in control=submission is set empty, Exim assumes that the
6511 authenticated id is a complete email address when it generates From: or
6512 Sender: header lines.
6514 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6515 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6516 chown and chgrp in /bin and hostname in /usr/bin.
6518 28. Exim was keeping the "process log" file open after each use, just as it
6519 does for the main log. This opens the possibility of it remaining open for
6520 long periods when the USR1 signal hits a daemon. Occasional processlog
6521 errors were reported, that could have been caused by this. Anyway, it seems
6522 much more sensible not to leave this file open at all, so that is what now
6525 29. The long-running daemon process does not normally write to the log once it
6526 has entered its main loop, and it closes the log before doing so. This is
6527 so that log files can straightforwardly be renamed and moved. However,
6528 there are a couple of unusual error situations where the daemon does write
6529 log entries, and I had neglected to close the log afterwards.
6531 30. The text of an SMTP error response that was received during a remote
6532 delivery was being truncated at 512 bytes. This is too short for some of
6533 the long messages that one sometimes sees. I've increased the limit to
6536 31. It is now possible to make retry rules that apply only when a message has a
6537 specific sender, in particular, an empty sender.
6539 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6540 it possible to be selective about when SMTP synchronization is enforced.
6542 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6544 32. Implemented hosts_connection_nolog.
6546 33. Added an ACL for QUIT.
6548 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6551 35. Added mailbox_size and mailbox_filecount to appendfile.
6553 36. Added control = no_multiline_responses to ACLs.
6555 37. There was a bug in the logic of the code that waits for the clock to tick
6556 in the case where the clock went backwards by a substantial amount such
6557 that the microsecond fraction of "now" was more than the microsecond
6558 fraction of "then" (but the whole seconds number was less).
6560 38. Added support for the libradius Radius client library this is found on
6561 FreeBSD (previously only the radiusclient library was supported).
6567 1. When certain lookups returned multiple values in the form name=value, the
6568 quoting of the values was not always being done properly. Specifically:
6569 (a) If the value started with a double quote, but contained no whitespace,
6571 (b) If the value contained whitespace other than a space character (i.e.
6572 tabs or newlines or carriage returns) it was not quoted.
6573 This fix has been applied to the mysql and pgsql lookups by writing a
6574 separate quoting function and calling it from the lookup code. The fix
6575 should probably also be applied to nisplus, ibase and oracle lookups, but
6576 since I cannot test any of those, I have not disturbed their existing code.
6578 2. A hit in the callout cache for a specific address caused a log line with no
6579 reason for rejecting RCPT. Now it says "Previous (cached) callout
6580 verification failure".
6582 3. There was an off-by-one bug in the queryprogram router. An over-long
6583 return line was truncated at 256 instead of 255 characters, thereby
6584 overflowing its buffer with the terminating zero. As well as fixing this, I
6585 have increased the buffer size to 1024 (and made a note to document this).
6587 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6588 when Exim is waiting for an SMTP response from a remote server, Exim
6589 restarts its select() call on the socket, thereby resetting its timeout.
6590 This is not a problem when such interrupts are rare. Somebody set up a cron
6591 job to run exiwhat every 2 minutes, which is less than the normal select()
6592 timeout (5 or 10 minutes). This meant that the select() timeout never
6593 kicked in because it was always reset. I have fixed this by comparing the
6594 time when an interrupt arrives with the time at the start of the first call
6595 to select(). If more time than the timeout has elapsed, the interrupt is
6596 treated as a timeout.
6598 5. Some internal re-factoring in preparation for the addition of Sieve
6599 extensions (by MH). In particular, the "personal" test is moved to a
6600 separate function, and given an option for scanning Cc: and Bcc: (which is
6601 not set for Exim filters).
6603 6. When Exim created an email address using the login of the caller as the
6604 local part (e.g. when creating a From: or Sender: header line), it was not
6605 quoting the local part when it contained special characters such as @.
6607 7. Installed new OpenBSD configuration files.
6609 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6610 try to make them clearer.
6612 9. Callout options, other than the timeout value, were being ignored when
6613 verifying sender addresses in header lines. For example, when using
6615 verify = header_sender/callout=no_cache
6617 the cache was (incorrectly) being used.
6619 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6620 only the Cygwin environment.
6622 11. When return_path_on_delivery was set as a log selector, if different remote
6623 addresses in the same message used different return paths and parallel
6624 remote delivery occurred, the wrong values would sometimes be logged.
6625 (Whenever a remote delivery process finished, the return path value from
6626 the most recently started remote delivery process was logged.)
6628 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6629 lines when AUTH and/or TLS are in use. This is the "received protocol"
6630 field. Exim used to use "asmtp" for authenticated SMTP, without any
6631 indication (in the protocol name) for TLS use. Now it follows the RFC and
6632 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6633 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6634 names appear in log lines as well as in Received: header lines.
6636 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6637 extensions, and comparison tests, and to fix some bugs.
6639 14. Changes to the "personal" filter test:
6641 (1) The test was buggy in that it was just doing the equivalent of
6642 "contains" tests on header lines. For example, if a user's address was
6643 anne@some.where, the "personal" test would incorrectly be true for
6645 To: susanne@some.where
6647 This test is now done by extracting each address from the header in turn,
6648 and checking the entire address. Other tests that are part of "personal"
6649 are now done using regular expressions (for example, to check local parts
6650 of addresses in From: header lines).
6652 (2) The list of non-personal local parts in From: addresses has been
6653 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6654 taken from the Sieve specification recommendations.
6656 (3) If the message contains any header line starting with "List-" it is
6657 treated as non-personal.
6659 (4) The test for "circular" in the Subject: header line has been removed
6660 because it now seems ill-conceived.
6662 15. Minor typos in src/EDITME comments corrected.
6664 16. Installed latest exipick from John Jetmore.
6666 17. If headers_add on a router specified a text string that was too long for
6667 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6668 of string_sprintf() is now avoided.
6670 18. $message_body_size was not set (it was always zero) when running the DATA
6671 ACL and the local_scan() function.
6673 19. For the "mail" command in an Exim filter, no default was being set for
6674 the once_repeat time, causing a random time value to be used if "once" was
6675 specified. (If the value happened to be <= 0, no repeat happened.) The
6676 default is now 0s, meaning "never repeat". The "vacation" command was OK
6677 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6678 (I found it when inspecting the code).
6680 20. There is now an overall timeout for performing a callout verification. It
6681 defaults to 4 times the callout timeout, which applies to individual SMTP
6682 commands during the callout. The overall timeout applies when there is more
6683 than one host that can be tried. The timeout is checked before trying the
6684 next host. This prevents very long delays if there are a large number of
6685 hosts and all are timing out (e.g. when the network connections are timing
6686 out). The value of the overall timeout can be changed by specifying an
6687 additional sub-option for "callout", called "maxwait". For example:
6689 verify = sender/callout=5s,maxwait=20s
6691 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6692 to the end before writing, but this should make it even safer).
6694 22. Exim was forgetting that it had advertised PIPELINING for the second and
6695 subsequent messages on an SMTP connection. It was also not resetting its
6696 memory on STARTTLS and an internal HELO.
6698 23. When Exim logs an SMTP synchronization error within a session, it now
6699 records whether PIPELINING has been advertised or not.
6701 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6702 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6703 rather than long int.
6705 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6707 26. Added the never_mail option to autoreply.
6713 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6714 crash if the getsockname() call failed; this can happen if a connection is
6715 closed very soon after it is established. The problem was simply in the
6716 order in which certain operations were done, causing Exim to try to write
6717 to the SMTP stream before it had set up the file descriptor. The bug has
6718 been fixed by making things happen in the correct order.
6724 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6725 before the connection was closed, thus losing the rejection response.
6727 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6728 some early Solaris releases, but causes trouble in current releases where
6729 socklen_t is defined.
6731 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6734 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6737 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6738 that releasing the top of it at the end releases what was used for sub-
6739 expansions (unless the block got too big). However, discard this block if
6740 the first thing is a variable or header, so that we can use its block when
6741 it is dynamic (useful for very large $message_headers, for example).
6743 6. Lookups now cache *every* query, not just the most recent. A new, separate
6744 store pool is used for this. It can be recovered when all lookup caches are
6745 flushed. Lookups now release memory at the end of their result strings.
6746 This has involved some general refactoring of the lookup sources.
6748 7. Some code has been added to the store_xxx() functions to reduce the amount
6749 of flapping under certain conditions.
6751 8. log_incoming_interface used to affect only the <= reception log lines. Now
6752 it causes the local interface and port to be added to several more SMTP log
6753 lines, for example "SMTP connection from", and rejection lines.
6755 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6757 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6759 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6760 could be overwritten at the end of the current message (or the start of a
6761 new message if it was set in a HELO ACL). The value is now preserved for
6762 the duration of the SMTP connection.
6764 12. If a transport had a headers_rewrite setting, and a matching header line
6765 contained an unqualified address, that address was qualified, even if it
6766 did not match any rewriting rules. The underlying bug was that the values
6767 of the flags that permit the existence of unqualified sender and recipient
6768 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6769 non-local messages, and by -bnq for local messages) were not being
6770 preserved with the message after it was received.
6772 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6773 "next input=" as part of the text comprising the host identity instead of
6774 the correct text. The code was using the same buffer for two different
6775 strings. However, depending on which order the printing function evaluated
6776 its arguments, the bug did not always show up. Under Linux, for example, my
6777 test suite worked just fine.
6779 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6780 doesn't work with some older versions of Perl. It has been changed to "my",
6781 which in any case is probably the better facility to use.
6783 15. A really picky compiler found some instances of statements for creating
6784 error messages that either had too many or two few arguments for the format
6787 16. The size of the buffer for calls to the DNS resolver has been increased
6788 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6789 for addresses that have a lot of PTR records. This alleviates a problem; it
6790 does not fully solve it.
6792 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6793 buffer now truncates the list and logs the incident, which is the same
6794 action as happens when Exim is looking up a host name and its aliases.
6795 Previously in this situation something unpredictable would happen;
6796 sometimes it was "internal error: store_reset failed".
6798 18. If a server dropped the connection unexpectedly when an Exim client was
6799 using GnuTLS and trying to read a response, the client delivery process
6800 crashed while trying to generate an error log message.
6802 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6803 string, for example:
6805 warn message = H1: something\nH2: something
6807 the text was added as a single header line from Exim's point of view
6808 though it ended up OK in the delivered message. However, searching for the
6809 second and subsequent header lines using $h_h2: did not work. This has been
6810 fixed. Similarly, if a system filter added multiple headers in this way,
6811 the routers could not see them.
6813 20. Expanded the error message when iplsearch is called with an invalid key to
6814 suggest using net-iplsearch in a host list.
6816 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6817 ACLs are no longer actually imposed (and a message to that effect is
6820 22. If a "gecos" field in a passwd entry contained escaped characters, in
6821 particular, if it contained a \" sequence, Exim got it wrong when building
6822 a From: or a Sender: header from that name. A second bug also caused
6823 incorrect handling when an unquoted " was present following a character
6824 that needed quoting.
6826 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6827 was not being matched caselessly.
6829 24. Arranged for all hyphens in the exim.8 source to be escaped with
6832 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6833 properly. Recipient callout cache records were still being keyed to include
6834 the sender, even when use_sender was set false. This led to far more
6835 callouts that were necessary. The sender is no longer included in the key
6836 when use_sender is false.
6838 26. Added "control = submission" modifier to ACLs.
6840 27. Added the ${base62d: operator to decode base 62 numbers.
6842 28. dnsdb lookups can now access SRV records.
6844 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6845 the configuration file.
6847 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6848 (-v) mode. This makes the output for a verbose queue run more intelligible.
6850 31. Added a use_postmaster feature to recipient callouts.
6852 32. Added the $body_zerocount variable, containing the number of binary zero
6853 bytes in the message body.
6855 33. The time of last modification of the "new" subdirectory is now used as the
6856 "mailbox time last read" when there is a quota error for a maildir
6859 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6861 35. Added +ignore_unknown as a special item in host lists.
6863 36. Code for decoding IPv6 addresses in host lists is now included, even if
6864 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6865 address was recognized as an IP address, but was then not correctly decoded
6866 into binary, causing unexpected and incorrect effects when compared with
6873 1. Very minor rewording of debugging text in manualroute to say "list of
6874 hosts" instead of "hostlist".
6876 2. If verify=header_syntax was set, and a header line with an unqualified
6877 address (no domain) and a large number of spaces between the end of the
6878 name and the colon was received, the reception process suffered a buffer
6879 overflow, and (when I tested it) crashed. This was caused by some obsolete
6880 code that should have been removed. The fix is to remove it!
6882 3. When running in the test harness, delay a bit after writing a bounce
6883 message to get a bit more predictability in the log output.
6885 4. Added a call to search_tidyup() just before forking a reception process. In
6886 theory, someone could use a lookup in the expansion of smtp_accept_max_
6887 per_host which, without the tidyup, could leave open a database connection.
6889 5. Added the variables $recipient_data and $sender_data which get set from a
6890 lookup success in an ACL "recipients" or "senders" condition, or a router
6891 "senders" option, similar to $domain_data and $local_part_data.
6893 6. Moved the writing of debug_print from before to after the "senders" test
6896 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6897 problems for message scanning, either using a data ACL, or using
6898 local_scan() because the Received: header was not generated till after they
6899 were called (in order to set the time as the time of reception completion).
6900 I have revised the way this works. The header is now generated after the
6901 body is received, but before the ACL or local_scan() are called. After they
6902 are run, the timestamp in the header is updated.
6908 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6909 before starting a queue runner without re-exec. This happened only when
6910 deliver_drop_privilege was set or when the Exim user was set to root. The
6911 effect of the bug was that timeouts during subsequent deliveries caused
6912 crashes instead of being properly handled. The handler is now left at its
6913 default (and expected) setting.
6915 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6916 message, again when deliver_drop_privilege is set or Exim is run as root.
6917 The bug described in (1) was not present in this case, but the tidying up
6918 of the other signals was missing. I have made the two cases consistent.
6920 3. The ignore_target_hosts setting on a manualroute router was being ignored
6921 for hosts that were looked up using the /MX notation.
6923 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6926 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6927 operated on the sender address. After changing the $sender_address to <>
6928 for the sender address verify, Exim was re-instated it as the original
6929 (before rewriting) address, but remembering that it had rewritten it, so it
6930 wasn't rewriting it again. This bug also had the effect of breaking the
6931 sender address verification caching when the sender address was rewritten.
6933 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6934 This has been changed so that if the ip literal address matches
6935 ignore_target_hosts, the router declines.
6937 7. Added expansion conditions match_domain, match_address, and match_local_
6938 part (NOT match_host).
6940 8. The placeholder for the Received: header didn't have a length field set.
6942 9. Added code to Exim itself and to exim_lock to test for a specific race
6943 condition that could lead to file corruption when using MBX delivery. The
6944 issue is with the lockfile that is created in /tmp. If this file is removed
6945 after a process has opened it but before that process has acquired a lock,
6946 there is the potential for a second process to recreate the file and also
6947 acquire a lock. This could lead to two Exim processes writing to the file
6948 at the same time. The added code performs the same test as UW imapd; it
6949 checks after acquiring the lock that its file descriptor still refers to
6950 the same named file.
6952 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6953 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6956 11. Added the smtp_active_hostname option. If used, this will typically be made
6957 to depend on the incoming interface address. Because $interface_address is
6958 not set up until the daemon has forked a reception process, error responses
6959 that can happen earlier (such as "too many connections") no longer contain
6962 12. If an expansion in a condition on a "warn" statement fails because a lookup
6963 defers, the "warn" statement is abandoned, and the next ACL statement is
6964 processed. Previously this caused the whole ACL to be aborted.
6966 13. Added the iplsearch lookup type.
6968 14. Added ident_timeout as a log selector.
6970 15. Added tls_certificate_verified as a log selector.
6972 16. Added a global option tls_require_ciphers (compare the smtp transport
6973 option of the same name). This controls incoming TLS connections.
6975 17. I finally figured out how to make tls_require_ciphers do a similar thing
6976 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6977 before starting the TLS session.
6979 18. Tabs are now shown as \t in -bP output.
6981 19. If the log selector return_path_on_delivery was set, Exim crashed when
6982 bouncing a message because it had too many Received: header lines.
6984 20. If two routers both had headers_remove settings, and the first one included
6985 a superfluous trailing colon, the final name in the first list and the
6986 first name in the second list were incorrectly joined into one item (with a
6987 colon in the middle).
6993 1. Added -C and -D options to the exinext utility, mainly to make it easier
6994 to include in the automated testing, but these could be helpful when
6995 multiple configurations are in use.
6997 2. The exinext utility was not formatting the output nicely when there was
6998 an alternate port involved in the retry record key, nor when there was a
6999 message id as well (for retries that were specific to a specific message
7000 and a specific host). It was also confused by IPv6 addresses, because of
7001 the additional colons they contain. I have fixed the IPv4 problem, and
7002 patched it up to do a reasonable job for IPv6.
7004 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7005 delivery, the log line now contains "pipelined" if PIPELINING was used.
7007 4. An SMTP transport process used to panic and die if the bind() call to set
7008 an explicit outgoing interface failed. This has been changed; it is now
7009 treated in the same way as a connect() failure.
7011 5. A reference to $sender_host_name in the part of a conditional expansion
7012 that was being skipped was still causing a DNS lookup. This no longer
7015 6. The def: expansion condition was not recognizing references to header lines
7016 that used bh_ and bheader_.
7018 7. Added the _cache feature to named lists.
7020 8. The code for checking quota_filecount in the appendfile transport was
7021 allowing one more file than it should have been.
7023 9. For compatibility with Sendmail, the command line option
7031 and sets the incoming protocol and host name (for trusted callers). The
7032 host name and its colon can be omitted when only the protocol is to be set.
7033 Note the Exim already has two private options, -pd and -ps, that refer to
7034 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7035 "s", but I don't think that's a major issue.
7037 10. A number of refactoring changes to the code, none of which should affect
7040 (a) The number of logging options was getting close to filling up the
7041 32-bit word that was used as a bit map. I have split them into two classes:
7042 those that are passed in the argument to log_write(), and those that are
7043 only ever tested independently outside of that function. These are now in
7044 separate 32-bit words, so there is plenty of room for expansion again.
7045 There is no change in the user interface or the logging behaviour.
7047 (b) When building, for example, log lines, the code previously used a
7048 macro that called string_cat() twice, in order to add two strings. This is
7049 not really sufficiently general. Furthermore, there was one instance where
7050 it was actually wrong because one of the argument was used twice, and in
7051 one call a function was used. (As it happened, calling the function twice
7052 did not affect the overall behaviour.) The macro has been replaced by a
7053 function that can join an arbitrary number of extra strings onto a growing
7056 (c) The code for expansion conditions now uses a table and a binary chop
7057 instead of a serial search (which was left over from when there were very
7058 few conditions). Also, it now recognizes conditions like "pam" even when
7059 the relevant support is not compiled in: a suitably worded error message is
7060 given if an attempt is made to use such a condition.
7062 11. Added ${time_interval:xxxxx}.
7064 12. A bug was causing one of the ddress fields not to be passed back correctly
7065 from remote delivery subprocesses. The field in question was not being
7066 subsequently used, so this caused to problems in practice.
7068 13. Added new log selectors queue_time and deliver_time.
7070 14. Might have fixed a bug in maildirsizefile handling that threw up
7071 "unexpected character" debug warnings, and recalculated the data
7072 unnecessarily. In any case, I expanded the warning message to give more
7075 15. Added the message "Restricted characters in address" to the statements in
7076 the default ACL that block characters like @ and % in local parts.
7078 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7079 Three changes have been made:
7081 (a) There was a serious bug; a negative response to MAIL caused the whole
7082 recipient domain to be cached as invalid, thereby blocking all messages
7083 to all local parts at the same domain, from all senders. This bug has
7084 been fixed. The domain is no longer cached after a negative response to
7085 MAIL if the sender used is not empty.
7087 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7090 (c) A new callout option, "use_sender" has been added for people who want
7091 the modified behaviour.
7097 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7100 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7101 indeed breaks things for older releases.
7103 3. Added additional logging to the case where there is a problem reading data
7104 from a filter that is running in a subprocess using a pipe, in order to
7105 try to track down a specific problem.
7107 4. Testing facility fudge: when running in the test harness and attempting
7108 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7109 getting "No route to host". Convert this to a timeout.
7111 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7114 6. Some OS don't have socklen_t but use size_t instead. This affects the
7115 fifth argument of getsockopt() amongst other things. This is now
7116 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7117 can be set for individual OS. I have set it for SunOS5, OSF1, and
7118 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7119 some earlier ones do not.
7121 7. Change 4.30/15 was not doing the test caselessly.
7123 8. The standard form for an IPv6 address literal was being rejected by address
7124 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7125 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7126 this, as well as the form without the "IPv6" on the front (but only when
7127 address literals are enabled, of course).
7129 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7131 10. Exim crashed if a message with an empty sender address specified by -f
7132 encountered a router with an errors_to setting. This could be provoked only
7133 by a command such as
7137 where an empty string was supplied; "<>" did not hit this bug.
7139 11. Installed PCRE release 4.5.
7141 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7142 remained set. It is now erased.
7144 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7145 times from message ids (which are base 36 rather than the normal 62).
7147 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7148 were being counted as actual protocol errors, and logged if the log
7149 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7150 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7151 and DATA following a set of rejected RCPTs do not count as protocol errors.
7152 In other words, Exim assumes they were pipelined, though this may not
7153 actually be the case. Of course, in all cases the client gets an
7154 appropriate error code.
7156 15. If a lookup fails in an ACL condition, a message about the failure may
7157 be available; it is used if testing the ACL cannot continue, because most
7158 such messages specify what the cause of the deferral is. However, some
7159 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7160 that caused an old message to be retained and used if a later statement
7161 caused a defer, replacing the real cause of the deferral.
7163 16. If an IP address had so many PTR records that the DNS lookup buffer
7164 was not large enough to hold them, Exim could crash while trying to process
7165 the truncated data. It now detects and logs this case.
7167 17. Further to 4.21/58, another change has been made: if (and only if) the
7168 first line of a message (the first header line) ends with CRLF, a bare LF
7169 in a subsequent header line has a space inserted after it, so as not to
7170 terminate the header.
7172 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7173 unnecessarily, used atoi() instead of strtol(), and didn't check the
7174 termination when getting file sizes from file names by regex.
7176 19. Completely re-implemented the support for maildirsize files, in the light
7177 of a number of problems with the previous contributed implementation
7178 (4.30/29). In particular:
7180 . If the quota is zero, the maildirsize file is maintained, but no quota is
7183 . If the maildir directory does not exist, it is created before any attempt
7184 to write a maildirsize file.
7186 . The quota value in the file is just a cache; if the quota is changed in
7187 the transport, the new value overrides.
7189 . A regular expression is available for excluding directories from the
7192 20. The autoreply transport checks the characters in options that define the
7193 message's headers; it allows continued headers, but it was checking with
7194 isspace() after an embedded newline instead of explicitly looking for a
7197 21. If all the "regular" hosts to which an address was routed had passed their
7198 expiry times, and had not reached their retry times, the address was
7199 bounced, even if fallback hosts were defined. Now Exim should go on to try
7202 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7203 equivalent code in the SMTP transport. Some hosts send humungous responses
7204 to HELO/EHLO, more than 1024 it seems.
7206 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7207 gives compiler warnings in some environments. I've now done it "properly",
7210 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7211 (because of problems with the built-in one) was declared to return uschar *
7212 instead of char *, causing compiler failure.
7214 25. Fixed a file descriptor leak when processing alias/forward files.
7216 26. Fixed a minor format string issue in dbfn.c.
7218 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7220 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7221 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7224 29. When a sender address is verified, it is cached, to save repeating the test
7225 when there is more than one recipient in a message. However, when the
7226 verification involves a callout, it is possible for different callout
7227 options to be set for different recipients. It is too complicated to keep
7228 track of this in the cache, so now Exim always runs a verification when a
7229 callout is required, relying on the callout cache for the optimization.
7230 The overhead is duplication of the address routing, but this should not be
7233 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7234 to be verified with callout=postmaster, and the main callout worked but the
7235 postmaster check failed, the verification correctly failed. However, if a
7236 subsequent RCPT command asked for sender verification *without* the
7237 postmaster check, incorrect caching caused this verification also to fail,
7240 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7241 it was not caching the DNS options (qualify_single, search_parents) that
7242 were used when the lookup failed. A subsequent lookup with different
7243 options therefore always gave the same answer, though there were cases
7244 where it should not have. (Example: a "domains = !$mx_any" option on a
7245 dnslookup router: the "domains" option is always processed without any
7246 widening, but the router might have qualify_single set.) Now Exim uses the
7247 cached value only when the same options are set.
7249 32. Added John Jetmore's "exipick" utility to the distribution.
7251 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7252 than a timeout (e.g. a certificate is required, and is not provided), an
7253 Exim server now closes the connection immediately. Previously it waited for
7254 the client to close - but if the client is SSL, it seems that they each
7255 wait for each other, leading to a delay before one of them times out.
7257 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7258 maintained 0.8.x compatibility because I don't think many are using it, and
7259 it is clearly obsolete.
7261 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7264 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7265 client certificate was expired. A simple patch fixes this, though I don't
7266 understand the full logic of why the verify callback is called multiple
7269 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7270 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7271 which causes problems with some clients (such as the Certicom SSL Plus
7272 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7273 disables the coutermeasure allowing Eudora to connect."
7275 38. Exim was not checking that a write() to a log file succeeded. This could
7276 lead to Bad Things if a log got too big, in particular if it hit a file
7277 size limit. Exim now panics and dies if it cannot write to a log file, just
7278 as it does if it cannot open a log file.
7280 39. Modified OS/Makefile-Linux so that it now contains
7282 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7284 The two -D definitions ensure that Exim is compiled with large file
7285 support, which makes it possible to handle log files that are bigger than
7288 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7289 instance) a domain was checked against a named list that involved a lookup,
7290 causing $domain_data to be set, then another domain was checked against the
7291 same list, then the first domain was re-checked, the value of $domain_data
7292 after the final check could be wrong. In particular, if the second check
7293 failed, it could be set empty. This bug probably also applied to
7296 41. The strip_trailing_dot option was not being applied to the address given
7297 with the -f command-line option.
7299 42. The code for reading a message's header from the spool was incrementing
7300 $received_count, but never initializing it. This meant that the value was
7301 incorrect (doubled) while delivering a message in the same process in which
7302 it was received. In the most common configuration of Exim, this never
7303 happens - a fresh exec is done - but it can happen when
7304 deliver_drop_privilege is set.
7306 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7307 it now includes up to 150 characters of the unexpected data in the log
7310 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7311 and building data strings. The size of both of these buffers was 10 000
7312 bytes - far larger than anybody would *ever* want, thought I. Needless to
7313 say, somebody hit the limit. I have increased the maximum line length to
7314 20 000 and the maximum data length of concatenated lines to 100 000. I have
7315 also fixed two bugs, because there was no checking on these buffers. Tsk,
7316 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7317 buffer is too small.
7319 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7320 lsearch lookups. Now it does.
7322 46. When parsing a route_list item in a manualroute router, a fixed-length
7323 buffer was used for the list of hosts. I made this 1024 bytes long,
7324 thinking that nobody would ever have a list of hosts that long. Wrong.
7325 Somebody had a whole pile of complicated expansion conditions, and the
7326 string was silently truncated, leading to an expansion error. It turns out
7327 that it is easier to change to an unlimited length (owing to other changes
7328 that have happened since this code was originally written) than to build
7329 structure for giving a limitation error. The length of the item that
7330 expands into the list of hosts is now unlimited.
7332 47. The lsearch lookup could not handle data where the length of text line was
7333 more than 4095 characters. Such lines were truncated, leading to shortened
7334 data being returned. It should now handle lines of any length.
7336 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7337 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7340 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7341 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7342 no longer happens in comment lines. A list of replacements is now placed
7343 at the head of all of the source files, except those whose only change is
7344 to replace PERL_COMMAND in the very first #! line.
7346 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7347 messages on the queue, with a bottom-up merge sort, using code contributed
7348 by Michael Haardt. This should make operations like -bp somewhat faster on
7349 large queues. It won't affect queue runners, except when queue_run_in_order
7352 51. Installed eximstats 1.31 in the distribution.
7354 52. Added support for SRV lookups to the dnslookup router.
7356 53. If an ACL referred to $message_body or $message_body_end, the value was not
7357 reset for any messages that followed in the same SMTP session.
7359 54. The store-handling optimization for building very long strings was not
7360 differentiating between the different store pools. I don't think this
7361 actually made any difference in practice, but I've tidied it.
7363 55. While running the routers to verify a sender address, $sender_address
7364 was still set to the sender address. This is wrong, because when routing to
7365 send a bounce to the sender, it would be empty. Therefore, I have changed
7366 it so that, while verifying a sender address, $sender_address is set to <>.
7367 (There is no change to what happens when verifying a recipient address.)
7369 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7370 target A or AAAA records (if not already returned) without resetting the
7371 qualify_single or search_parents options of the DNS resolver. These are
7372 inappropriate in this case because the targets of MX and SRV records must
7373 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7374 target that, when qualified, matched something in the local domain. These
7375 two options are now turned off when doing these lookups.
7377 57. It seems that at least some releases of Reiserfs (which does not have the
7378 concept of a fixed number of inodes) returns zero and not -1 for the
7379 number of available inodes. This interacted badly with check_spool_inodes,
7380 which assumed that -1 was the "no such thing" setting. What I have done is
7381 to check that the total number of inodes is greater than zero before doing
7382 the test of how many are available.
7384 58. When a "warn" ACL statement has a log_message modifier, the message is
7385 remembered, and not repeated. This is to avoid a lot of repetition when a
7386 message has many recipients that cause the same warning to be written.
7387 However, Exim was preserving the list of already written lines for an
7388 entire SMTP session, which doesn't seem right. The memory is now reset if a
7389 new message is started.
7391 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7392 debugging output unless log_rewrite was also set.
7394 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7395 of (int)(handle) when we know that handle contains (void *)(-1).
7397 61. The Exim daemon panic-logs an error return when it closes the incoming
7398 connection. However "connection reset by peer" seems to be common, and
7399 isn't really an error worthy of noting specially, so that particular error
7402 62. When Exim is trying to find all the local interfaces, it used to panic and
7403 die if the ioctl to get the interface flags failed. However, it seems that
7404 on at least one OS (Solaris 9) it is possible to have an interface that is
7405 included in the list of interfaces, but for which you get a failure error
7406 for this call. This happens when the interface is not "plumbed" into a
7407 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7408 failure of the "get flags" call assumes that the interface is down.
7410 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7411 makes life easier for people who are doing arithmetic on fields extracted
7412 from dates, where you often get leading zeros that should not be
7413 interpreted as octal.
7415 64. Added qualify_domain to the redirect router, to override the global
7418 65. If a pathologically long header line contained very many addresses (the
7419 report of this problem mentioned 10 000) and each of them was rewritten,
7420 Exim could use up a very large amount of memory. (It kept on making new
7421 copies of the header line as it rewrote, and never released the old ones.)
7422 At the expense of a bit more processing, the header rewriting function has
7423 been changed so that it no longer eats memory in this way.
7425 66. The generation of the Received: header has been moved from the time that a
7426 message starts to be received, to the time that it finishes. The timestamp
7427 in the Received: header should now be very close to that of the <= log
7428 line. There are two side-effects of this change:
7430 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7431 logged header lines no longer include the local Received: line, because
7432 it has not yet been created. The same applies to a copy of the message
7433 that is returned to a non-SMTP sender when a message is rejected.
7435 (b) When a filter file is tested using -bf, no additional Received: header
7436 is added to the test message. After some thought, I decided that this
7439 This change does not affect the value of $received_for. It is still set
7440 after address rewriting, but before local_scan() is called.
7442 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7444 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7445 gave an unhelpful panic error message, and a defer error. I have managed to
7446 change this behaviour so that it now rejects any supplied certificate,
7447 which seems right, as the list of acceptable certificates is empty.
7449 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7450 gave an unhelpful defer error. I have not managed to make this reject any
7451 supplied certificates, but the error message it gives is "no certificate
7452 supplied", which is not helpful.
7454 70. exigrep's output now also includes lines that are not associated with any
7455 message, but which match the given pattern. Implemented by a patch from
7456 Martin Sluka, which also tidied up the Perl a bit.
7458 71. Recipient callout verification, like sender verification, was using <> in
7459 the MAIL FROM command. This isn't really the right thing, since the actual
7460 sender may affect whether the remote host accepts the recipient or not. I
7461 have changed it to use the actual sender in the callout; this means that
7462 the cache record is now keyed on a recipient/sender pair, not just the
7463 recipient address. There doesn't seem to be a real danger of callout loops,
7464 since a callout by the remote host to check the sender would use <>.
7465 [SEE ABOVE: changed after hitting problems.]
7467 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7468 temporary errors. However, in the case of such a code being given after
7469 the end of a data transmission (i.e. after ".") Exim was failing to write
7470 a retry record for the message. (Yes, there was some broken host that was
7471 actually sending 8xx at this point.)
7473 73. An unknown lookup type in a host list could cause Exim to panic-die when
7474 the list was checked. (An example that provoked this was putting <; in the
7475 middle of a list instead of at the start.) If this happened during a DATA
7476 ACL check, a -D file could be left lying around. This kind of configuration
7477 error no longer causes Exim to die; instead it causes a defer error. The
7478 incident is still logged to the main and panic logs.
7480 74. Buglet left over from Exim 3 conversion. The message "too many messages
7481 in one connection" was written to the rejectlog but not the mainlog, except
7482 when address rewriting (yes!) was being logged.
7484 75. Added write_rejectlog option.
7486 76. When a system filter was run not as root (that is, when system_filter_user
7487 was set), the values of the $n variables were not being returned to the
7488 main process; thus, they were not subsequently available in the $sn
7491 77. Added +return_path_on_delivery log selector.
7493 78. A connection timeout was being treated differently from recipients deferred
7494 when testing hosts_max_try with a message that was older than the host's
7495 retry timeout. (The host should not be counted, thus allowing all hosts to
7496 be tried at least once before bouncing.) This may have been the cause of an
7497 occasionally reported bug whereby a message would remain on the queue
7498 longer than the retry timeout, but would be bounced if a delivery was
7499 forced. I say "may" because I never totally pinned down the problem;
7500 setting up timeout/retry tests is difficult. See also the next item.
7502 79. The ultimate address timeout was not being applied to errors that involved
7503 a combination of host plus message (for example, a timeout on a MAIL
7504 command). When an address resolved to a number of possible hosts, and they
7505 were not all tried for each delivery (e.g. because of hosts_max_try), a
7506 message could remain on the queue longer than the retry timeout.
7508 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7511 81. Fixed an obscure SMTP outgoing bug which required at least the following
7512 conditions: (a) there was another message waiting for the same server;
7513 (b) the server returned 5xx to all RCPT commands in the first message so
7514 that the message was not completed; (c) the server dropped the connection
7515 or gave a negative response to the RSET that Exim sends to abort the
7516 transaction. The observed case was a dropped connection after DATA that had
7517 been sent in pipelining mode. That is, the server had advertised PIPELINING
7518 but was not implementing it correctly. The effect of the bug was incorrect
7519 behaviour, such as trying another host, and this could lead to a crash.
7525 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7526 and daemon.c were passed as pointers to ints; they should have been
7527 pointers to socklen_t variables (which are typically unsigned ints).
7529 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7532 3. Fixed a really odd bug that affected only the testing scheme; patching a
7533 certain fixed string in the binary changed the value of another string that
7534 happened to be identical to the end of the original first string.
7536 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7537 name", it returns that address as the IP address. On some operating
7538 systems (e.g. Solaris), it also passes back the IP address string as the
7539 "host name". However, on others (e.g. Linux), it passes back an empty
7540 string. Exim wasn't checking for this, and was changing the host name to an
7541 empty string, assuming it had been canonicalized.
7543 5. Although rare, it is permitted to have more than one PTR record for a given
7544 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7545 all the names associated with an address, because they do in Solaris.
7546 However, it seems that they do not in Linux for data that comes from the
7547 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7548 I found this out when I moved to a new Linux workstation and tried to run
7549 the Exim test suite.
7551 To get round this problem I have changed the code so that it now does its
7552 own call to the DNS to look up PTR records when searching for a host name.
7553 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7554 addresses that are only in /etc/hosts are still found.
7556 This behaviour is, however, controlled by an option called host_lookup_
7557 order, which defaults to "bydns:byaddr". If people want to use the other
7558 order, or indeed, just use one or the other means of lookup, they can
7559 specify it in this variable.
7561 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7562 some operating systems, this comes back from gethostbyaddr() as an empty
7563 string, and this is what Exim used to test for. However, it seems that in
7564 other systems, "." is yielded. Exim now tests for this case too.
7566 7. The values of check_spool_space and check_log_space are now held internally
7567 as a number of kilobytes instead of an absolute number of bytes. If a
7568 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7569 kilobyte. This means that much larger values can be stored.
7571 8. Exim monitor: an attempt to get the action menu when not actually pointing
7572 at a message produces an empty menu entitled "No message selected". This
7573 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7574 no entries in it ("Shell widget menu has zero width and/or height"). So I
7575 have added a single, blank menu entry in this case.
7577 9. Added ${quote_local_part.
7579 10. MIME decoding is now applied to the contents of Subject: header lines when
7582 11. Now that a reference to $sender_host_address automatically causes a reverse
7583 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7584 host lookup before query-style lookups in lists that might use this
7585 variable. This has therefore been abolished, and the "net-" prefix is no
7586 longer necessary for query-style lookups.
7588 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7589 have been a typo for LFLAGS, so it has been changed.
7591 13. The install script calls Exim with "-C /dev/null" in order to find the
7592 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7593 to be output. However, since Exim outputs its version number before the
7594 error, it didn't break the script. It just looked ugly. I fixed this by
7595 always allowing "-C /dev/null" if the caller is root.
7597 14. Ignore overlarge ACL variable number when reading spool file - insurance
7598 against a later release with more variables having written the file.
7600 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7601 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7602 this, as well as the form without the "IPv6" on the front.
7604 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7605 OS/Makefile-Darwin file.
7607 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7608 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7610 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7611 variable. However, some OS use ints for inodes. I've added cast to long int
7612 to get rid of the compiler warning.
7614 19. I had forgotten to lock out "/../" in configuration file names when
7615 ALT_CONFIG_PREFIX was set.
7617 20. Routers used for verification do not need to specify transports. However,
7618 if such a router generated a host list, and callout was configured, Exim
7619 crashed, because it could not find a port number from the (non-existent)
7620 transport. It now assumes port 25 in this circumstance.
7622 21. Added the -t option to exigrep.
7624 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7625 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7626 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7627 them is set, LOOKUP_LSEARCH is forced.
7629 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7631 24. Added sender and host information to the "rejected by local_scan()" log
7632 line; previously there was no indication of these.
7634 25. Added .include_if_exists.
7636 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7637 Linux. It turns out that not all file systems support this. Apparently some
7638 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7639 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7640 not supported on the file descriptor, is now ignored when Exim is trying to
7641 sync a directory. This applies only to Linux.
7643 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7645 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7646 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7649 29. Added support for maildirsize files from supplied patch (modified a bit).
7651 30. The use of :fail: followed by an empty string could lead Exim to respond to
7652 sender verification failures with (e.g.):
7654 550 Verification failed for <xxx>
7655 550 Sender verify failed
7657 where the first response line was missing the '-' that indicates it is not
7658 the final line of the response.
7660 31. The loop for finding the name of the user that called Exim had a hardwired
7661 limit of 10; it now uses the value of finduser_retries, which is used for
7662 all other user lookups.
7664 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7667 33. Exim was neglecting to zero errno before one call of strtol() when
7668 expanding a string and expecting an integer value. On some systems this
7669 resulted in spurious "integer overflow" errors. Also, it was casting the
7670 result into an int without checking.
7672 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7673 did not work. The code looks as if it has *never* worked, though it appears
7674 to have been documented since at least release 1.62. I have made it work.
7676 35. The "timeout_DNS" error in retry rules, also documented since at least
7677 1.62, also never worked. As it isn't clear exactly what this means, and
7678 clearly it isn't a major issue, I have abolished the feature by treating it
7679 as "timeout", and writing a warning to the main and panic logs.
7681 36. The display of retry rules for -brt wasn't always showing the error code
7684 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7685 timeout_connect_A, timeout_connect_MX.
7687 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7688 to the empty sender.
7690 39. The daemon was not analysing the content of -oX till after it had closed
7691 stderr and disconnected from the controlling terminal. This meant that any
7692 syntax errors were only noted on the panic log, and the return code from
7693 the command was 0. By re-arranging the code a little, I've made the
7694 decoding happen first, so such errors now appear on stderr, and the return
7695 code is 1. However, the actual setting up of the sockets still happens in
7696 the disconnected process, so errors there are still only recorded on the
7699 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7700 connections (as happens on some IP stacks) was logged at start up time as
7701 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7702 it from "IPv6 and IPv4", which means that two separate sockets are being
7705 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7706 says whether AF_INET or AF_INET6 was passed as an argument.
7708 42. Exiwhat output was messed up when time zones were included in log
7711 43. Exiwhat now gives more information about the daemon's listening ports,
7712 and whether -tls-on-connect was used.
7714 44. The "port" option of the smtp transport is now expanded.
7716 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7717 silently ignored. Now an error message is written to the main and panic
7720 46. There's a new ACL modifier called "logwrite" which writes to a log file
7721 as soon as it is encountered.
7723 47. Added $local_user_uid and $local_user_gid at routing time.
7725 48. Exim crashed when trying to verify a sender address that was being
7728 49. Exim was recognizing only a space character after ".include". It now also
7729 recognizes a tab character.
7731 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7732 extracting the relevant information from the specification. The man page no
7733 longer contains scrambled data for the -d option, and I've added a section
7734 at the front about calling Exim under different names.
7736 51. Added "extra_headers" argument to the "mail" command in filter files.
7738 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7741 53. Installed eximstats 1.29.
7743 54. Added transport_filter_timeout as a generic transport option.
7745 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7746 Cc: header lines. This was required by RFC 822, but it not required by RFC
7749 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7750 incoming messages that did not have them. Now it does so only if the
7751 message originates locally, that is, if there is no associated remote host
7752 address. When Resent- header lines are present, this applies to the Resent-
7753 lines rather than the non-Resent- lines.
7755 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7756 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7758 58. Messages for configuration errors now include the name of the main
7759 configuration file - useful now that there may be more than one file in a
7760 list (.included file names were always shown).
7762 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7763 for those rare installations that do not start the daemon as root or run it
7764 setuid root. I've cut out the call to initgroups() if the daemon is not
7767 60. The Exim user and group can now be bound into the binary as text strings
7768 that are looked up at the start of Exim's processing.
7770 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7772 62. Added $mailstore_basename variable.
7774 63. Installed patch to sieve.c from Michael Haardt.
7776 64. When Exim failed to open the panic log after failing to open the main log,
7777 the original message it was trying to log was written to stderr and debug
7778 output, but if they were not available (the usual case in production), it
7779 was lost. Now it is written to syslog before the two lines that record the
7780 failures to open the logs.
7782 65. Users' Exim filters run in subprocesses under the user's uid. It is
7783 possible for a "deliver" command or an alias in a "personal" command to
7784 provoke an address rewrite. If logging of address rewriting is configured,
7785 this fails because the process is not running as root or exim. There may be
7786 a better way of dealing with this, but for the moment (because 4.30 needs
7787 to be released), I have disabled address rewrite logging when running a
7788 filter in a non-root, non-exim process.
7794 1. The buildconfig auxiliary program wasn't quoting the value set for
7795 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7796 not defined. This bug was masked in 4.22 by the effect that was fixed in
7799 2. Some messages that were rejected after a message id was allocated were
7800 shown as "incomplete" by exigrep. It no longer does this for messages that
7801 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7803 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7804 have allow_domain_literals set, the ID did not get logged in the <= line.
7805 Domain literals are now always recognized in Message-ID: header lines.
7807 4. The first argument for a ${extract expansion item is the key name or field
7808 number. Leading and trailing spaces in this item were not being ignored,
7809 causing some misleading effects.
7811 5. When deliver_drop_privilege was set, single queue runner processes started
7812 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7813 same command in the process it spins off) were not dropping privilege.
7815 6. When the daemon running as "exim" started a queue runner, it always
7816 re-executed Exim in the spun-off process. This is a waste of effort when
7817 deliver_drop_privilege is set. The new process now just calls the
7818 queue-runner function directly.
7824 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7827 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7828 ignored. Though the use of -oP was forcing the writing of a pid file, it
7829 was always written to the default place.
7831 3. If the message "no IP address found for host xxxx" is generated during
7832 incoming verification, it is now followed by identification of the incoming
7833 connection (so you can more easily find what provoked it).
7835 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7837 5. Added some features to "harden" Exim a bit more against certain attacks:
7839 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7840 be put in Local/Makefile. This is like the never_users runtime option,
7841 but it cannot be overridden. The default setting is "root".
7843 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7844 prefix string with which any file named in a -C command line option
7847 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7848 is retained for -C and -D only if the caller of Exim is root. Without
7849 it, the exim user may also use -C and -D and retain privilege.
7851 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7852 command line option is disabled.
7854 6. Macro names set by the -D option must start with an upper case letter, just
7855 like macro names defined in the configuration file.
7857 7. Added "dereference=" facility to LDAP.
7859 8. Two instances of the typo "uknown" in the source files are fixed.
7861 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7862 the Configure-Makefile script screwed up while processing it.
7864 10. Incorporated PCRE 4.4.
7866 11. The SMTP synchronization check was not operating right at the start of an
7867 SMTP session. For example, it could not catch a HELO sent before the client
7868 waited for the greeting. There is now a check for outstanding input at the
7869 point when the greeting is written. Because of the duplex, asynchronous
7870 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7871 way, but not yet received, when the check is performed.
7873 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7874 on TCP/IP sockets, because this apparently causes some broken clients to
7877 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7878 unchanged) from the Cygwin maintainer.
7880 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7881 was supported instead of testing for mbx. Effectively a typo.
7883 15. The spa authenticator server code was not checking that the input it
7884 received was valid base64.
7886 16. The debug output line for the "set" modifier in ACLs was not showing the
7887 name of the variable that was being set.
7889 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7891 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7892 lookup on its own. Something else was needed to trigger the lookup. For
7893 example, a match in host_lookup or the need for a host name in a host list.
7894 Now, if $sender_host_name is referenced and the host name has not yet been
7895 looked up, a lookup is performed. If the lookup fails, the variable remains
7896 empty, and $host_lookup_failed is set to "1".
7898 19. Added "eqi" as a case-independent comparison operator.
7900 20. The saslauthd authentication condition could segfault if neither service
7901 nor realm was specified.
7903 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7904 error message that was logged was misleading, and incoming SMTP
7905 connections were dropped. The message is now more accurate, and temporary
7906 errors are given to SMTP connections.
7908 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7909 (except RSET) with a 421 error, until QUIT is received. However, it was
7910 failing to send a response to QUIT.
7912 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7913 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7914 ACL is now run after the helo_try_verify_hosts code.
7916 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7917 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7918 case-independent, so other case variants are also recognized). Apparently
7919 some systems use these upper case variants.
7921 25. If more than two messages were waiting for the same host, and a transport
7922 filter was specified for the transport, Exim sent two messages over the
7923 same TCP/IP connection, and then failed with "socket operation on non-
7924 socket" when it tried to send the third.
7926 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7928 27. The extern definition of crypt16() in expand.c was not being excluded when
7929 the OS had its own crypt16() function.
7931 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7932 as a preferred synonym for return_size_limit, both as an option and as an
7935 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7937 30. Changed the default configuration ACL to relax the local part checking rule
7938 for addresses that are not in any local domains. For these addresses,
7939 slashes and pipe symbols are allowed within local parts, but the sequence
7940 /../ is explicitly forbidden.
7942 31. SPA server authentication was not clearing the challenge buffer before
7945 32. log_message in a "warn" ACL statement was writing to the reject log as
7946 well as to the main log, which contradicts the documentation and doesn't
7947 seem right (because no rejection is happening). So I have stopped it.
7949 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7950 However, I am unable to do any testing of this.
7952 34. Fixed an infelicity in the appendfile transport. When checking directories
7953 for a mailbox, to see if any needed to be created, it was accidentally
7954 using path names with one or more superfluous leading slashes; tracing
7955 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7957 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7958 contents are added to the log line that is written for every discarded
7959 recipient. (Previously a log_message setting was ignored.)
7961 36. The ${quote: operator now quotes the string if it is empty.
7963 37. The install script runs exim in order to find its version number. If for
7964 some reason other than non-existence or emptiness, which it checks, it
7965 could not run './exim', it was installing it with an empty version number,
7966 i.e. as "exim-". This error state is now caught, and the installation is
7969 38. An argument was missing from the function that creates an error message
7970 when Exim fails to connect to the socket for saslauthd authentication.
7971 This could cause Exim to crash, or give a corrupted message.
7973 39. Added isip, isip4, and isip6 to ${if conditions.
7975 40. The ACL variables $acl_xx are now saved with the message, and can be
7976 accessed later in routers, transports, and filters.
7978 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7979 strings in the file are not string-expanded.
7981 42. If a MAIL command specified a SIZE value that was too large to fit into an
7982 int variable, the check against message_size_limit failed. Such values are
7983 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7984 day this will have to be increased, but I don't think I want to be around
7985 when emails are that large.
7992 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
7993 iconv() is not standard in FreeBSD.
7995 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
7996 IPv6 enabled. The observed symptom was a segmentation fault on return from
7997 the function os_common_find_running_interfaces() in src/os.c.
7999 3. In the check_special_case() function in daemon.c I had used "errno" as an
8000 argument name, which causes warnings on some systems. This was basically a
8001 typo, since it was named "eno" in the comments!
8003 4. The code that waits for the clock to tick (at a resolution of some fraction
8004 of a second) so as to ensure message-id uniqueness was always waiting for
8005 at least one whole tick, when it could have waited for less. [This is
8006 almost certainly not relevant at current processor speeds, where it is
8007 unlikely to ever wait at all. But we try to future-proof.]
8009 5. The function that sleeps for a time interval that includes fractions of a
8010 second contained a race. It did not block SIGALRM between setting the
8011 timer, and suspending (a couple of lines later). If the interval was short
8012 and the sigsuspend() was delayed until after it had expired, the suspension
8013 never ended. On busy systems this could lead to processes getting stuck for
8016 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8017 process, before it forks any delivery processes. The open lookup caching
8018 mechanism meant that the open file or database connection was passed into
8019 the delivery process. The problem was that delivery processes always tidy
8020 up cached lookup data. This could cause a problem for the next delivery
8021 process started by the queue runner, because the external queue runner
8022 process does not know about the closure. So the next delivery process
8023 still has data in the lookup cache. In the case of a file lookup, there was
8024 no problem because closing a file descriptor in a subprocess doesn't affect
8025 the parent. However, if the lookup was caching a connection to a database,
8026 the connection was closed, and the second delivery process was likely to
8027 see errors such as "PGSQL: query failed: server closed the connection
8028 unexpectedly". The problem has been fixed by closing all cached lookups
8029 in a queue runner before running a delivery process.
8031 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8032 seem to have the "const" qualifier which it has on other OS. I've
8035 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8036 *of the same type* (client or server) with the same public name that an
8037 error should be diagnosed.
8039 9. When Exim looked up a host name for an IP address, but failed to find the
8040 original IP address when looking up the host name (a safety check), it
8041 output the message "<ip address> does not match any IP for NULL", which was
8042 confusing, to say the least. The bug was that the host name should have
8043 appeared instead of "NULL".
8045 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8046 user that is built into the binary, and the -C or -D options is used, root
8047 privilege is dropped before the configuration file is read. In addition,
8048 logging is switched to stderr instead of the normal log files. If the
8049 configuration then re-defines the Exim user, the unprivileged environment
8050 is probably not what is expected, so Exim logs a panic warning message (but
8053 However, if deliver_drop_privilege is set, the unprivileged state may well
8054 be exactly what is intended, so the warning has been cut out in that case,
8055 and Exim is allowed to try to write to its normal log files.
8061 1. smtp_return_error_details was not giving details for temporary sender
8062 or receiver verification errors.
8064 2. Diagnose a configuration error if two authenticators have the same public
8067 3. Exim used not to create the message log file for a message until the first
8068 delivery attempt. This could be confusing when incoming messages were held
8069 for policy or load reasons. The message log file is now created at the time
8070 the message is received, and an initial "Received" line is written to it.
8072 4. The automatically generated man page for command line options had a minor
8073 bug that caused no ill effects; however, a more serious problem was that
8074 the procedure for building the man page automatically didn't always
8075 operate. Consequently, release 4.20 contains an out-of-date version. This
8076 shouldn't happen again.
8078 5. When building Exim with embedded Perl support, the script that builds the
8079 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8080 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8082 6. The freeze_tell option was not being used for messages that were frozen on
8083 arrival, either by an ACL or by local_scan().
8085 7. Added the smtp_incomplete_transaction log selector.
8087 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8088 was accepting AUTH without a new EHLO.
8090 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8091 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8094 10. Logging of TCP/IP connections (when configured) now happens in the main
8095 daemon process instead of the child process, so that the TCP/IP connection
8096 count is more accurate (but it can never be perfect).
8098 11. The use of "drop" in a nested ACL was not being handled correctly in the
8099 outer ACL. Now, if condition failure induced by the nested "drop" causes
8100 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8101 or "require"), the connection is dropped.
8103 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8104 that yield "discard" can now be used with an "accept" or a "discard" verb,
8105 but an error is generated for any others (because I can't see a useful way
8106 to define what should happen).
8108 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8109 are now processed in the same way as lines in the Exim configuration file.
8110 In particular, continuation lines are supported.
8112 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8114 15. Added -ti meaning -t -i.
8116 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8117 domains, and warn by logging if others are found.
8119 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8120 returned from GIFCONF when Exim is trying to find the list of interfaces on
8121 a host. The code in os.c has been modified to copy each ifreq to an aligned
8122 structure in all cases.
8124 Also, in some cases, the returned ifreq's were being copied to a 'struct
8125 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8126 means the last couple of bytes of an IPv6 address could be chopped if the
8127 ifreq contained only a normal sockaddr (14 bytes storage).
8129 18. Named domain lists were not supported in the hosts_treat_as_local option.
8130 An entry such as +xxxx was not recognized, and was treated as a literal
8133 19. Ensure that header lines added by a DATA ACL are included in the reject log
8134 if the ACL subsequently rejects the message.
8136 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8137 MD5 (which is deprecated).
8139 21. When testing a filter file using -bf, Exim was writing a message when it
8140 took the sender from a "From " line in the message, but it was not doing so
8141 when it took $return_path from a Return-Path: header line. It now does.
8143 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8144 with a valid header line field name (a series of printing characters
8145 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8147 23. Changed "disc" in the source to "disk" to conform to the documentation and
8148 the book and for uniformity.
8150 24. Ignore Sendmail's -Ooption=value command line item.
8152 25. When execve() failed while trying to run a command in a pipe transport,
8153 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8154 could be confused with a return value of 69 from the command itself. This
8155 has been changed to 127, the value the shell returns if it is asked to run
8156 a non-existent command. The wording for the related log line suggests a
8157 non-existent command as the problem.
8159 26. If received_header_text expands to an empty string, do not add a Received:
8160 header line to the message. (Well, it adds a token one on the spool, but
8161 marks it "old" so that it doesn't get used or transmitted.)
8163 27. Installed eximstats 1.28 (addition of -nt option).
8165 28. There was no check for failure on the call to getsockname() in the daemon
8166 code. This can fail if there is a shortage of resources on the system, with
8167 ENOMEM, for example. A temporary error is now given on failure.
8169 29. Contrary to the C standard, it seems that in some environments, the
8170 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8171 program. Exim now does this explicitly; it affects the formatting of
8172 timestamps using strftime().
8174 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8175 complaints. I've now initialized all the variables, to avoid this.
8177 32. Header lines added by a system filter were not being "seen" during
8178 transport-time rewrites.
8180 33. The info_callback() function passed to OpenSSL is set up with type void
8181 (*)(SSL *, int, int), as described somewhere. However, when calling the
8182 function (actually a macro) that sets it up, the type void(*)() is
8183 expected. I've put in a cast to prevent warnings from picky compilers.
8185 34. If a DNS black list lookup found a CNAME record, but there were no A
8186 records associated with the domain it pointed at, Exim crashed.
8188 35. If a DNS black list lookup returned more than one A record, Exim ignored
8189 all but the first. It now scans all returned addresses if a particular IP
8190 value is being sought. In this situation, the contents of the
8191 $dnslist_value variable are a list of all the addresses, separated by a
8194 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8195 to accept a host name and all its aliases if the forward lookup for any of
8196 them yielded the IP address of the incoming connection. Now it accepts only
8197 those names whose forward lookup yields the correct IP address. Any other
8198 names are discarded. This closes a loophole whereby a rogue DNS
8199 administrator could create reverse DNS records to break through a
8200 wildcarded host restriction in an ACL.
8202 37. If a user filter or a system filter that ran in a subprocess used any of
8203 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8204 the wrong values were passed to the pipe command ($thisaddress had the
8205 value of $0, $0 had the value of $1, etc). This bug was introduced by
8206 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8208 38. Improved the line breaking for long SMTP error messages from ACLs.
8209 Previously, if there was no break point between 40 and 75 characters, Exim
8210 left the rest of the message alone. Two changes have been made: (a) I've
8211 reduced the minimum length to 35 characters; (b) if it can't find a break
8212 point between 35 and 75 characters, it looks ahead and uses the first one
8213 that it finds. This may give the occasional overlong line, but at least the
8214 remaining text gets split now.
8216 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8217 file descriptors might be low, and that setting 1000 would always raise it.
8218 It turns out that in some environments, the limit is already over 1000 and
8219 that lowering it causes trouble. So now Exim takes care not to decrease it.
8221 40. When delivering a message, the value of $return_path is set to $sender_
8222 address at the start of routing (routers may change the value). By an
8223 oversight, this default was not being set up when an address was tested by
8224 -bt or -bv, which affected the outcome if any router or filter referred to
8227 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8228 log only when "message" or "log_message" are set. However, if one of the
8229 conditions was an address verification, or a call to a nested ACL, the
8230 messages generated by the underlying test were being passed through. This
8231 no longer happens. The underlying message is available in $acl_verify_
8232 message for both "message" and "log_message" expansions, so it can be
8233 passed through if needed.
8235 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8236 new expansion $bh_ to give the encoded byte string without charset
8237 translation. Translation happens only if iconv() is available; HAVE_ICONV
8238 indicates this at build time. HEADERS_CHARSET gives the charset to
8239 translate to; headers_charset can change it in the configuration, and
8240 "headers charset" can change it in an individual filter file.
8242 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8243 that creates RFC 2047 encoded "words" labels them as that charset instead
8244 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8245 expansion operator; (ii) when Exim creates a From: line for a local
8246 message; (iii) when a header line is rewritten to include a "phrase" part.
8248 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8249 buggy, causing it to skip the first lines of messages whose message ID
8250 ended in 'D'. This would not have bitten before Exim release 4.14, because
8251 message IDs were unlikely to end in 'D' before then. The effect was to have
8252 incorrect size information for certain domains.
8254 45. #include "config.h" was missing at the start of the crypt16.c module. This
8255 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8258 46. If there was a timeout during a "random" callout check, Exim treated it as
8259 a failure of the random address, and carried on sending RSET and the real
8260 address. If the delay was just some slowness somewhere, the response to the
8261 original RCPT would be taken as a response to RSET and so on, causing
8262 mayhem of various kinds.
8264 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8265 when I implemented it. It didn't allow for the fact that some option values
8266 may legitimately be negative (e.g. size_addition), and it didn't even do
8267 the right test for positive values.
8269 48. Domain names in DNS records are case-independent. Exim always looks them up
8270 in lower case. Some resolvers return domain names in exactly the case they
8271 appear in the zone file, that is, they may contain uppercase letters. Not
8272 all resolvers do this - some return always lower case. Exim was treating a
8273 change of case by a resolver as a change of domain, similar to a widening
8274 of a domain abbreviation. This triggered its re-routing code and so it was
8275 trying to route what was effectively the same domain again. This normally
8276 caused routing to fail (because the router wouldn't handle the domain
8277 twice). Now Exim checks for this case specially, and just changes the
8278 casing of the domain that it ultimately uses when it transmits the message
8281 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8284 50. If a filter generated a file delivery with a non-absolute name (possible if
8285 no home directory exists for the router), the forbid_file option was not
8288 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8289 the existing equality matching.
8291 52. Exim was using ints instead of ino_t variables in some places where it was
8292 dealing with inode numbers.
8294 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8295 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8296 called TMPDIR, and if it finds it is different, it changes its value.
8298 54. The smtp_printf() function is now made available to local_scan() so
8299 additional output lines can be written before returning. There is also an
8300 smtp_fflush() function to enable the detection of a dropped connection.
8301 The variables smtp_input and smtp_batched_input are exported to
8304 55. Changed the default runtime configuration: the message "Unknown user"
8305 has been removed from the ACL, and instead placed on the localuser router,
8306 using the cannot_route_message feature. This means that any verification
8307 failures that generate their own messages won't get overridden. Similarly,
8308 the "Unrouteable address" message that was in the ACL for unverifiable
8309 relay addresses has also been removed.
8311 56. Added hosts_avoid_esmtp to the smtp transport.
8313 57. The exicyclog script was not checking for the esoteric option
8314 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8315 will work only if exicyclog is run under the appropriate euid.
8317 58. Following a discussion on the list, the rules by which Exim recognises line
8318 endings on incoming messages have been changed. The -dropcr and drop_cr
8319 options are now no-ops, retained only for backwards compatibility. The
8320 following line terminators are recognized: LF CRLF CR. However, special
8321 processing applies to CR:
8323 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8324 nor a local message in the state where . is a terminator.
8326 (ii) If a bare CR is encountered in a header line, an extra space is added
8327 after the line terminator so as not to end the header. The reasoning
8328 behind this is that bare CRs in header lines are most likely either
8329 to be mistakes, or people trying to play silly games.
8331 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8332 was being incorrectly given as 18 bytes larger than it should have been.
8333 This is a VOB (very old bug).
8335 60. This may never have affected anything current, but just in case it has:
8336 When the local host is found other than at the start of a list of hosts,
8337 the local host, those with the same MX, and any that follow, are discarded.
8338 When the list in question was part of a longer list of hosts, the following
8339 hosts (not currently being processed) were also being discarded. This no
8340 longer happens. I'm not sure if this situation could ever has previously
8343 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8346 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8347 header. This is something that is recommended in a new Internet Draft, and
8348 is something that is documented as being done by Sendmail. There are two
8349 possible values. For messages generated by the autoreply transport, Exim
8352 Auto-Submitted: auto-replied
8354 whereas for all other generated messages (e.g. bounces) it adds
8356 Auto-Submitted: auto-generated
8358 63. The "personal" condition in filters now includes a test for the
8359 Auto-Submitted: header. If it contains the string "auto-" the message it
8360 not considered personal.
8362 64. Added rcpt_include_affixes as a generic transport option.
8364 65. Added queue_only_override (default true).
8366 66. Added the syslog_duplication option.
8368 67. If what should have been the first header line of a message consisted of
8369 a space followed by a colon, Exim was mis-interpreting it as a header line.
8370 It isn't of course - it is syntactically invalid and should therefore be
8371 treated as the start of the message body. The misbehaviour could have
8372 caused a number of strange effects, including loss of data in subsequent
8373 header lines, and spool format errors.
8375 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8376 client host had authenticated. This control can now be exercised by an ACL
8377 for more flexibility.
8379 69. By default, callouts do not happen when testing with -bh. There is now a
8380 variant, -bhc, which does actually run the callout code, including
8381 consulting and updating the callout cache.
8383 70. Added support for saslauthd authentication, courtesy of Alexander
8386 71. If statvfs() failed on the spool or log directories while checking their
8387 size for availability, Exim confusingly gave the error "space shortage".
8388 Furthermore, in debugging mode it crashed with a floating point exception.
8389 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8390 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8391 serious problem, Exim now writes to the main and panic logs when this
8392 happens, with details of the failure. It then refuses to accept the
8393 incoming message, giving the message "spool directory problem" or "log
8394 directory problem" with a 421 code for SMTP messages.
8396 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8397 0, 1, and 2 exist, because some OS complain for execs without them (see
8398 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8399 descriptors. However, the code omitted to check that the open succeeded,
8400 causing mysterious errors if for some reason the permissions on /dev/null
8401 got screwed. Now Exim writes a message to the main and panic logs, and
8402 bombs out if it can't open /dev/null.
8404 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8405 interact so that it is all more flexible. It is supposed to remain
8406 backwards compatible. Also added extra_local_interfaces.
8408 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8409 to bomb out with an assertion failure - to the client this appears as a
8410 connection drop. This problem occurs in the part of the code that was taken
8411 from the Samba project. Fortunately, the assertion is in a very simple
8412 function, so I have fixed this by reproducing the function inline in the
8413 one place where it is called, and arranging for authentication to fail
8414 instead of killing the process with assert().
8416 75. The SPA client code was not working when the server requested OEM rather
8417 than Unicode encoding.
8419 76. Added code to make require_files with a specific uid setting more usable in
8420 the case where statting the file as root fails - usually a non-root-mounted
8421 NFS file system. When this happens and the failure is EACCES, Exim now
8422 forks a subprocess and does the per-uid checking as the relevant uid.
8424 77. Added process_log_path.
8426 78. If log_file_path was not explicitly set, a setting of check_log_space or
8427 check_log_inodes was ignored.
8429 79. If a space check for the spool or log partitions fails, the incident is now
8430 logged. Of course, in the latter case the data may get lost...
8432 80. Added the %p formatting code to string_format() so that it can be used to
8433 print addresses in debug_print(). Adjusted all the address printing in the
8434 debugging in store.c to use %p rather than %d.
8436 81. There was a concern that a line of code in smtp_in.c could overflow a
8437 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8438 initially expressed, the concern was not well-founded, because trailing
8439 spaces are removed early. However, if the trailing spaces were followed by
8440 a NULL, they did not get removed, so the overflow was possible. Two fixes
8443 (a) I re-wrote the offending code in a cleaner fashion.
8444 (b) If an incoming SMTP command contains a NULL character, it is rejected
8447 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8448 runs initgroups(), so that if the Exim user is in any additional groups,
8449 they will be used during message reception.
8455 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob