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 records.
184 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
185 or fakedefer. Previously the sender could discover that the message
186 had in fact been accepted.
188 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
189 sequences. Previously the input was not noted as being tainted; notably
190 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
191 bad coding of authenticators.
197 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
198 after reception to before a subsequent reception. This should
199 mean slightly faster delivery, and also confirmation of reception
202 JH/02 Move from using the pcre library to pcre2. The former is no longer
203 being developed or supported (by the original developer).
205 JH/03 Constification work in the filters module required a major version
206 bump for the local-scan API. Specifically, the "headers_charset"
207 global which is visible via the API is now const and may therefore
208 not be modified by local-scan code.
210 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
211 sendfile() didi not account for the way the ClamAV driver code called it.
213 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
214 time check was being done for every character; replace that with one
217 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
218 sent was prefixed with a length byte.
220 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
221 RFC 5321. Previously Dovecot (at least) would log errors during
224 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
225 in 4.61 without a reason logged. Binaries will be bigger, which might
226 matter on diskspace-constrained systems, but debug is easier.
228 JH/09 Fix macro-definition during "-be" expansion testing. The move to
229 write-protected store for macros had not accounted for these runtime
230 additions; fix by removing this protection for "-be" mode.
232 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
233 handing out large-numbered file descriptors, violating the usual Unix
234 assumption (and required by Posix) that the lowest possible number will be
235 allocated by the kernel when a new one is needed. In the daemon, and any
236 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
237 are not useable for FD_SET() [and hence select()] and overwrite the stack.
238 Assorted crashes happen.
240 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
241 main-section options or in a connect ACL, the value from the first ever
242 connection was never replaced for subsequent connections. Found by
245 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
246 though only once PCRE2 was introduced: the memory accounting used under
247 debug offset allocations by an int, giving a hard trap in early startup.
248 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
250 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
251 with underbars is given. The write-protection of configuration introduced
252 in 4.95 trapped when normalisation was applied to an option not needing
255 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
257 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
258 failing to close the certificates directory, every hour or any time it
261 JH/16 Debugging initiated by an ACL control now continues through into routing
262 and transport processes. Previously debugging stopped any time Exim
263 re-execs, or for processing a queued message.
265 JH/17 The "expand" debug selector now gives more detail, specifically on the
266 result of expansion operators and items.
268 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
269 bad comparison between the option value and the name of the file to
270 be included was done, and a mismatch was wrongly identified.
271 4.88 to 4.95 are affected.
273 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
275 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
276 supplied as the db file-pair basename. Previously, if a directory
277 path was given, for example via the autoreply "once" option, the DB
278 file.pag and file.dir files would be created in that directory's
281 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
282 "taint" log_selector. These were previously deprecated.
284 JH/22 Fix static address-list lookups to properly return the matched item.
285 Previously only the domain part was returned.
287 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
288 the call into OpenSSL to send a TLS Close was being repeated; this
289 resulted in the library waiting for the peer's Close. If that was never
290 sent we waited forever. Fix by tracking send calls.
292 JH/24 The ${run} expansion item now expands its command string elements after
293 splitting. Previously it was before; the new ordering makes handling
294 zero-length arguments simpler. The old ordering can be obtained by
295 appending a new option "preexpand", after a comma, to the "run".
297 JH/25 Taint-check exec arguments for transport-initiated external processes.
298 Previously, tainted values could be used. This affects "pipe", "lmtp" and
299 "queryprogram" transport, transport-filter, and ETRN commands.
300 The ${run} expansion is also affected: in "preexpand" mode no part of
301 the command line may be tainted, in default mode the executable name
304 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
305 the the facility was not passed across execs, and only the first message
306 passed over a connection could use BDAT; any further ones using DATA.
308 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
309 uses $sending_ip_address and an interface is specified.
310 Previously any use of the local address in the EHLO name disabled
311 PIPECONNECT, the common case being to use the rDNS of it.
313 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
314 resumption. Previously verify failed because no certificate status is
315 passed on the wire for the restarted session. Fix by using the recorded
316 ocsp status of the stored session for the new connection.
318 JH/29 TLS resumption: the key for session lookup in the client now includes
319 more info that a server could potentially use in configuring a TLS
320 session, avoiding oferring mismatching sessions to such a server.
321 Previously only the server IP was used.
323 JH/30 Fix string_copyn() for limit greater than actual string length.
324 Previously the copied amount was the limit, which could result in a
325 overlapping memcpy for newly allocated destination soon after a
326 source string shorter than the limit. Found/investigated by KM.
328 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
329 close; it may be needed for a subsequent connection. This caused a
330 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
332 JH/32 Fix CHUNKING for a second message on a connection when the first was
333 rejected. Previously we did not reset the chunking-offered state, and
334 erroneously rejected the BDAT command. Investigation help from
337 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
338 an empty address. Previously the expansion returned an error.
340 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
341 proxy. Previously these were misparsed, leading to paniclog entries.
347 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
348 related applications. Previously an "H" was used where available info
349 says that "M" should be, so change to match.
351 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
352 as arguments, so an implementation trying to copy these into a local
353 buffer was taking a taint-enforcement trap. Fix by using dynamically
354 created buffers. Similar fix for radius expansion condition.
356 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
357 reasonable, eg. to count headers. Fix by using dynamically created
358 buffers rather than a local. Do similar fixes for ACL actions "dcc",
359 "log_reject_target", "malware" and "spam"; the arguments are expanded
360 so could be handling tainted values.
362 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
363 broken the (no-op) support for this sendmail command. Restore it
364 to doing nothing, silently, and returning good status.
366 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
367 record path was given (or the default used) without a leading directory
368 path, an error occurred on trying to open it. Use the transport's working
371 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
372 transport. Previously it was the name on the DNS A-record; use instead
373 the head of the CNAME chain leading there (if there is one). This seems
374 to align better with RFC 6125.
376 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
377 smtp_accept_max_per_host allocated resources which were not released
378 when the limit was exceeded. This eventually crashed the daemon. Fix
379 by adding a release action in that path.
381 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
382 expanded; previously using tainted values was rejected. Fix by using
383 dynamically-created buffers.
385 JH/09 Relax restrictions on ACL verify condition needing access to message
386 headers. Previously they were only permitted in data and non-smtp ACLs;
387 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
388 not_blind, header_sender and header_names_ascii verification.
390 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
391 Previously a macro used one argument twice; when called with the
392 argument as an expression having side-effects, incorrect operation
393 resulted. Use an inlineable function.
395 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
396 held open for a verify callout. Previously this wan not accounted for
397 and a corrupt onward SMTP conversation resulted.
399 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
400 Previously they were embedded naively in the constructed address; when
401 needed, strip the quoting and quote the entire local_part.
402 Also make the inbound_srs expansion condition handle quoting.
404 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
405 excluded, not matching the documentation.
407 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
408 was given for the sqlite_dbfile a trap resulted.
410 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
411 "name" argument resulted in a trap. There is no reason to disallow such;
412 this was a coding error.
414 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
415 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
416 spent suspended, ignoring the POSIX definition. Previously we assumed
417 it did and a constant offset from real time could be used as a correction.
418 Change to using the same clock source for the start-of-message and the
419 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
420 exists, just to get a clock slightly more aligned to reality.
422 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
423 RFC says it is optional some validators care. The missing char was not
424 intended but triggered by a line-wrap alignment. Discovery and fix by
425 Guillaume Outters, hacked on by JH.
427 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
428 name being quoted was tainted a trap would be taken. Fix by using
429 dynamically created buffers. The routine could have been called by a
430 rewrite with the "h" flag, by using the "-F" command-line option, or
431 by using a "name=" option on a control=submission ACL modifier.
433 JH/19 SPF: change the Authentication-Results expansion component to give
434 smtp.helo when the sender domain is empty. Previously it gave
437 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
438 not in 127.0.0.0/8 to help in spotting list domains taken over by a
439 domain-parking registrar.
441 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
442 Previously when a whitespace character was specified it was not inserted
443 after removing the newline.
445 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
446 the domain part of the recipient address. This overrides any tls_sni
447 option set, which was previously used.
449 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
452 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
453 is_tainted() had an off-by-one error in the overenthusiastic direction.
454 Find and fix by Gavan. Although NetBSD is not a supported platform for
455 4.94 this bug could affect other platforms.
457 PP/01 Fix default prime selection to be consistent.
458 One path used ike23 still, instead of exim.dev.20160529.3; now both
459 execution flows will use the same DH primes (currently
460 exim.dev.20160529.3).
462 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
463 option in smtp transport, to match the documentation. Previously
464 verification was not being done in some cases where it should have been.
466 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
467 than one server was defined and depending on the platform memory layout
468 details, an internal consistency trap could be hit while walking the list
471 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
472 files. The value is used by the authresults expansion item. Previously
473 if this was used in a router or transport, a crash could result.
475 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
476 have a sufficient incidence of EINTR returns from select that an
477 interaction with other operations done by the main daemon loop exposed
478 a bug in the error-handling. This was benign apart from the log
481 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
482 for consistency with delivered "=>" and failed "**" lines. While we're
483 there, handle PRX and TFO.
485 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
486 applied. This resulted, if any header-line rewrite rules were configured,
487 in a panic-log triggerable by sending a message with a long address in
488 a header. Fix by increasing the arbitrary limit to larger than a single
489 (dewrapped) 5322 header line maximum size.
491 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
492 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
493 RFC 6648 which deprecates X- options in protocols as a general practice.
494 Changeover between the implementations is handled by the mechanisms
497 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
498 a remote which uses callout/hold. Previously the local was queued.
500 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
503 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
504 log line. The asterisk is now consistently appended to the remote IP
505 (and port, if given), and will also be provided on defer and fail log
506 lines. Previously it could be placed on the local IP if that was being
507 logged, and was only provided on delivery lines.
509 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
511 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
512 authenticator client_send option. Previously the next char, after a pair
513 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
514 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
515 documentation. There is still no way to get a leading ^ immediately
516 after a NUL (ie. for the password of a PLAIN method authenticator.
518 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
519 files. For bad sizes read, delete the record and whine to paniclog.
521 JH/38 When logging an AUTH failure, as server, do not include sensitive
522 information. Previously, the credentials would be included if given
523 as part of the AUTH command line and an ACL denied authentication.
525 JH/39 Bug 2691: fix $local_part_data. When the matching list element
526 referred to a file, bad data was returned. This likely also affected
529 JH/40 The gsasl authenticator now supports caching of the salted password
530 generated by the client-side implementation. This required the addition
531 of a new variable: $auth4.
533 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
534 left undeleted; the attempt to re-create it then failed - resulting in
535 the usual "SIGHUP tp have daemon reload configuration" to not work.
536 This affected any platform not supporting "abstract" Unix-domain
537 sockets (i.e. not Linux).
539 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
540 recipients" response to RCPT in a later response, with a 250. The
541 previous coding assumed this would not happen, and under PIPELINING
542 would result in both lost and duplicate recipients for a message.
544 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
545 Previously the weighting was incorrectly applied. Similar fix for socks
546 proxies. Found and fixed by Heiko Schlichting.
548 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
549 not handle sub-lists included using the +namedlist syntax. While
550 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
551 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
554 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
555 Previously the permanent pool was used, so the sore could not be freed.
556 This meant a connection with many messages would use continually-growing
559 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
560 per-pool so as not to waste too much space. Previously a constant size
561 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
562 DOS attacks harder. The cost is wasted memory use in the larger blocks.
564 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
565 same space repeatedly, and a smaller process footprint.
567 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
568 for GnuTLS. Previously the name started "%s" which, while not a bug,
569 looked as if if might be one.
571 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
572 "continued connection") make the $proxy_* variables available. Previously
573 the information was not passed across the exec() call for subsequent
574 transport executions. This also mean that the log lines for the
575 messages can show the proxy information.
577 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
578 receive time. With modern systems the difference is significant.
579 The historical behaviour can be restored by disabling (a new) log_selector
580 "queue_time_exclusive".
582 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
583 content) were specifically tested for. Now, also cover expansions
584 resulting in ACL names and inline ACL content.
586 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
587 making it unusable in complex expressions.
589 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
590 fail with a taint-check on the spoolfile name, and leave the message
593 HS/01 Enforce absolute PID file path name.
595 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
597 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
599 PP/02 Bug 2643: Correct TLS DH constants.
600 A missing NUL termination in our code-generation tool had led to some
601 incorrect Diffie-Hellman constants in the Exim source.
602 Reported by kylon94, code-gen tool fix by Simon Arlott.
604 PP/03 Impose security length checks on various command-line options.
605 Fixes CVE-2020-SPRSS reported by Qualys.
607 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
608 better. Reported by Qualys.
610 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
611 providing a particularly obnoxious sender full name.
614 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
616 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
619 PP/08 Change default for recipients_max from unlimited to 50,000.
621 PP/09 Fix security issue with too many recipients on a message (to remove a
622 known security problem if someone does set recipients_max to unlimited,
623 or if local additions add to the recipient list).
624 Fixes CVE-2020-RCPTL reported by Qualys.
626 PP/10 Fix security issue in SMTP verb option parsing
627 Fixes CVE-2020-EXOPT reported by Qualys.
629 PP/11 Fix security issue in BDAT state confusion.
630 Ensure we reset known-good where we know we need to not be reading BDAT
631 data, as a general case fix, and move the places where we switch to BDAT
632 mode until after various protocol state checks.
633 Fixes CVE-2020-BDATA reported by Qualys.
635 HS/03 Die on "/../" in msglog file names
637 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
638 the Exim runtime user are allowed to create files.
640 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
643 QS/03 When reading the output from interpreted forward files we do not
644 pass the pipe between the parent and the interpreting process to
645 executed child processes (if any).
647 QS/04 Always die if requested from internal logging, even is logging is
650 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
651 the API; compilation noo longer completes with DMARC support included.
652 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
653 been reported on other platforms.
655 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
657 JH/56 Make the majority of info read from config files readonly, for defence-in-
658 depth against exploits. Suggestion by Qualys.
659 Not supported on Solaris 10.
661 JH/57 Fix control=fakreject for a custom message containing tainted data.
662 Previously this resulted in a log complaint, due to a re-expansion present
663 since fakereject was originally introduced.
665 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
666 modifier was given, a loop resulted.
668 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
669 pipelined SMTP command followed the BDAT LAST then it would be
670 incorrectly treated as part of the message body, causing a verification
673 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
674 require looser validation rules than those for 821-level addresses,
675 which only permit IP addresses.
681 JH/01 Avoid costly startup code when not strictly needed. This reduces time
682 for some exim process initialisations. It does mean that the logging
683 of TLS configuration problems is only done for the daemon startup.
685 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
687 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
688 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
691 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
692 DKIM signing is being done. Previously a transport_filter would always
693 disable CHUNKING, falling back to traditional DATA.
695 JH/05 Regard command-line recipients as tainted.
697 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
699 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
700 PAM library frees one of the arguments given to it, despite the
701 documentation. Therefore a plain malloc must be used.
703 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
704 on-stack buffers were used, resulting in a taint trap when DSN information
705 copied from a received message was written into the buffer.
707 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
708 the ordering of its ARC headers. This caused a crash.
710 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
711 a new record was being constructed with information from the peer, a trap
714 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
715 installation would get error messages from DMARC verify, when it hit the
716 nonexistent file indicated by the default. Distros wanting DMARC enabled
717 should both provide the file and set the option.
718 Also enforce no DMARC verification for command-line sourced messages.
720 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
721 could, depending on the platform, hang at the STARTTLS response.
723 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
724 message on a connection. Previously if one message had ARC headers and
725 the following one did not, a crash could result when adding an
726 Authentication-Results: header.
728 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
729 Exim main code and Exim-related utities. The introduction of taint
730 tracking also did many adjustments to string handling. Since then, eximon
731 frequently terminated with an assert failure.
733 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
734 check for 452 responses. This slightly helps the inefficieny of doing
735 a large alias-expansion into a recipient-limited target. The max_rcpt
736 transport option still applies (and at the current default, will override
737 the new feature). The check is done for either cause of synch, and forces
738 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
739 connection. The new facility is not tunable at this time.
741 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
742 library live data was being used, so the results became garbage. Make
743 copies while it is still usable.
745 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
746 on delivery deferred (==) and failed (**) lines (if a delivery was
747 attemtped). Previously it was only on completion (=>) lines.
749 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
750 for the expansion of the server_scram_iter and server_scram_salt options.
752 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
753 are now specifically given a NO_DATA response without hitting the system
754 resolver. The library goes on to do the now-standard TXT lookup.
755 Use of dnsdb lookups is not affected.
757 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
758 only retrieve the errormessage once. Previously two calls to dlerror()
759 were used, and the second one (for mainlog/paniclog) retrieved null
762 JH/20 Taint checking: disallow use of tainted data for
763 - the appendfile transport file and directory options
764 - the pipe transport command
765 - the autoreply transport file, log and once options
766 - file names used by the redirect router (including filter files)
768 - paths used by single-key lookups
769 Previously this was permitted.
771 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
772 adjusted the size of a major service buffer; this failed because the
773 buffer was in use at the time. Change to a compile-time increase in the
774 buffer size, when this authenticator is compiled into exim.
776 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
777 previous fast-mode was untenable in the face of glibs using mmap to
778 support larger malloc requests.
780 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
781 New values supported, if defined on system where compiled:
782 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
783 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
785 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
786 running a limited number of proceses in parallel, a benefit is gained. The
787 amount varies with the platform hardware and load. The use of the option
788 queue_run_in_order means we cannot do this, as ordering becomes
791 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
792 had introduced a string-copy (for ensuring NUL-termination) which was not
793 appropriate for that case, which can include embedded NUL bytes in the
794 block of data. Investigation showed the copy to actually be needless, the
795 data being length-specified.
797 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
798 done during a receiving connection, and both used TLS, global info was
799 used rather than per-connection info for tracking the state of data
800 queued for transmission. This could result in a connection hang.
802 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
803 Previously, when delivering serveral messages down a single connection
804 only the first would provide a SIZE. This was due to the size information
805 not being properly tracked.
807 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
808 TAI (at 37 seconds currently), pretend to be in UTC for time-related
809 expansion and logging. Previously, spurious values such as a future
810 minute could be seen.
812 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
813 it could crash from a null-deref. This could also affect the
814 ${addresses: } operator and ${readsock } item.
816 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
817 message following a mime one, the variable was not reset.
819 JH/30 When an pipelined-connect fails at the first response, assume incorrect
820 cached capability (perhaps the peer reneged?) and immediately retry in
823 JH/31 Fix spurious detection of timeout while writing to transport filter.
825 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
826 an attempt to copy the string was made before checking it.
828 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
829 taint of the lookup key was maintained; we now regard the presence in the
830 filesystem as sufficient validation.
832 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
833 argument is supplied.
835 JH/35 The dsearch lookup now requires that the directory is an absolute path.
836 Previously this was not checked, and nonempty relative paths made an
837 access under Exim's current working directory.
839 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
840 Previously no event was raised.
842 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
843 parameter supplied by the sender MAIL FROM command. Previously it was
844 ignored, and only the check_spool_space option value for the required
847 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
848 the size of the signing public-key. Previously it was instead giving
849 the size of the signature hash.
851 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
852 the default. See the (new) dkim_verify_min_keysizes option.
854 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
855 and an ACL use a lookup for checking either the local_part or domain,
856 stale data could be accessed. Ensure that variable references are
857 dropped between messages.
859 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
860 by the client was not checked as pointing within response data before
861 being used. A malicious client could thus cause an out-of-bounds read and
862 possibly gain authentication. Fix by adding the check.
864 JH/42 Internationalisation: change the default for downconversion in the smtp
865 transport to be "if needed". Previously it was "as previously set" for
866 the message, which usually meant "if needed" for message-submission but
867 "no" for everything else. However, MTAs have been seen using SMTPUTF8
868 even when the envelope addresses did not need it, resulting in forwarding
869 failures to non-supporting MTAs. A downconvert in such cases will be
870 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
871 transport. The change does mean that addresses needing conversion will
872 be converted when previously a delivery failure would occur.
874 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
875 response was received it would be used unchecked in a fail-DSN, violating
876 standards on line-length limits. Truncate if needed.
878 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
879 doesn't work. (Additionally add a new main config option to configure the
886 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
887 side, to decode a TLS 1.3 packet capture.
889 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
890 Previously the default library behaviour applied, sending two, each in
893 JH/03 Debug output for ACL now gives the config file name and line number for
896 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
898 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
900 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
901 buffer overrun for (non-chunking) other transports.
903 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
904 TLS1.3, means that a server rejecting a client certificate is not visible
905 to the client until the first read of encrypted data (typically the
906 response to EHLO). Add detection for that case and treat it as a failed
907 TLS connection attempt, so that the normal retry-in-clear can work (if
908 suitably configured).
910 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
911 and/or domain. Found and fixed by Jason Betts.
913 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
914 configuration). If a CNAME target was not a wellformed name pattern, a
917 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
918 the OS reports them interleaved with other addresses.
920 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
921 used both for input and for a verify callout, both encrypted, SMTP
922 responses being sent by the server could be lost. This resulted in
923 dropped connections and sometimes bounces generated by a peer sending
926 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
927 string. Previously it was possible to cause undefined behaviour in a
928 library routine (usually a crash). Found by "zerons".
930 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
933 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
934 API was removed, so update to use the newer ones.
936 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
937 any timeout set, is taking a long time. Previously we would hang on to a
938 rotated logfile "forever" if the input was arriving with long gaps
939 (a previous attempt to fix addressed lack, for a long time, of initial
942 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
943 shared (NFS) environment. The length of the tempfile name is now
944 4 + 16 ("hdr.$message_exim_id") which might break on file
945 systems which restrict the file name length to lower values.
948 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
949 shared (NFS) environment.
951 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
952 did for all versions <4.90). Notably -M, -m, --invert, -I may be
955 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
956 on some platforms for bit 31.
958 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
959 to changes apparently associated with TLS1.3 handling some of the APIs
960 previously used were either nonfunctional or inappropriate. Strings
961 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
962 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
963 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
964 This affects log line X= elements, the $tls_{in,out}_cipher variables,
965 and the use of specific cipher names in the encrypted= ACL condition.
967 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
969 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
970 verification result was not updated unless hosts_require_ocsp applied.
972 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
973 queue_list_requires_admin set to false, non-admin users were denied the
976 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
977 directory-of-certs mode. Previously they were advertised despite the
980 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
981 A single TCP connection by a client will now hold a TLS connection open
982 for multiple message deliveries, by default. Previously the default was to
985 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
986 default. If built with the facility, DANE will be used. The facility
987 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
989 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
990 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
991 must be defined and you must still, unless you define DISABLE_TLS, manage
992 the the include-dir and library-file requirements that go with that
993 choice. Non-TLS builds are still supported.
995 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
998 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
999 default. If the platform supports and has the facility enabled, it will
1000 be requested on all coneections.
1002 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1003 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1005 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1007 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1008 success-DSN messages. Previously the From: header was always the default
1009 one for these; the option was ignored.
1011 JH/28 Fix the timeout on smtp response to apply to the whole response.
1012 Previously it was reset for every read, so a teergrubing peer sending
1013 single bytes within the time limit could extend the connection for a
1014 long time. Credit to Qualsys Security Advisory Team for the discovery.
1016 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1017 delivery address, which leaked information of the results of local
1018 forwarding. Change to the original envelope recipient address, per
1021 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1022 requested. Previously not bounce was generated and a log entry of
1023 error ignored was made.
1025 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1027 JH/32 Introduce a general tainting mechanism for values read from the input
1028 channel, and values derived from them. Refuse to expand any tainted
1029 values, to catch one form of exploit.
1031 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1032 was unused and the unexpanded text used for the test. Found and
1033 fixed by Ruben Jenster.
1035 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1036 an attempt to use a TLS library read routine dereffed a nul pointer,
1039 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1040 them in our smtp response.
1042 JH/36 Have the general router option retry_use_local_part default to true when
1043 any of the restrictive preconditions are set (to anything). Previously it
1044 was only for check_local user. The change removes one item of manual
1045 configuration which is required for proper retries when a remote router
1046 handles a subset of addresses for a domain.
1048 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1049 link count into consideration.
1051 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1052 caused the extension of big_buffer, the following lines were ignored.
1054 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1055 accordance with RFC 2308. Previously there was no expiry, so a longlived
1056 receive process (eg. due to ACL delays) versus a short SOA value could
1059 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1061 JH/39 Promote DMARC support to mainline.
1063 JH/40 Bug 2452: Add a References: header to DSNs.
1065 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1066 parameters. The relevant library call is documented as "Deprecated: This
1067 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1068 3.6.0, DH parameters are negotiated following RFC7919."
1070 HS/06 Change the default of dnssec_request_domains to "*"
1072 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1073 carried on and emitted a BDAT command, even when PIPELINING was not
1076 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1077 buffer was used for the filename, resulting in a trap when tainted
1078 arguments (eg. $domain) were used.
1080 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1081 recommended to avoid a possible server-load attack. The feature can be
1082 re-enabled via the openssl_options main cofiguration option.
1084 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1085 for version 4.90 - adding a "more data" boolean to the arguments.
1086 Bumped the ABI version number also, this having been missed previously;
1087 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1088 respect of usage of smtp_printf() by either local_scan code or libraries
1089 accessed via the ${dlfunc } expansion item. Both will need coding
1090 adjustment for any calls to smtp_printf() to match the new function
1091 signature; a FALSE value for the new argument is always safe.
1093 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1094 the file-offset (which the Linux syscall does, and exim expects); this
1095 resulted in an indefinite loop.
1097 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1098 to do ARC verification. The Authentication-Results: header line added
1099 by the configuration then had no ARC item.
1101 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1102 an incorrect mode was used for reading the result, resulting in it being
1109 JH/01 Remove code calling the customisable local_scan function, unless a new
1110 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1112 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1113 non-signal-safe functions being used.
1115 JH/03 Bug 2269: When presented with a received message having a stupidly large
1116 number of DKIM-Signature headers, disable DKIM verification to avoid
1117 a resource-consumption attack. The limit is set at twenty.
1119 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1120 report of oldest_pass in ${authres } in consequence, and separate out
1121 some descriptions of reasons for verification fail.
1123 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1124 files in the spool were present and unlocked. A queue-runner could spot
1125 them, resulting in a duplicate delivery. Fix that by doing the unlock
1126 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1127 add more error-checking on spoolfile handling while that code is being
1130 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1131 No known attacks, no CVE, this is defensive hardening.
1133 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1134 a queue-runner could start a delivery while other operations were ongoing.
1135 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1136 Found and investigated by Tim Stewart. Fix by using the open message data
1137 file handle rather than opening another, and not locally closing it (which
1138 releases a lock) for that case, while creating the temporary .eml format
1139 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1141 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1142 $sender_verify_failure/$recipient_verify_failure to "random".
1144 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1147 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1148 Previously this would segfault.
1150 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1153 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1154 like zero, since the resolver should be doing this for us, But we need one
1155 as a CNAME but no MX presence gets the CNAME returned; we need to check
1156 that doesn't point to an MX to declare it "no MX returned" rather than
1157 "error, loop". A new main option is added so the older capability of
1158 following some limited number of chain links is maintained.
1160 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1162 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1163 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1164 now-deprecated earlier definitions used only specified the range up to TLS
1165 1.2 (in the older-version library docs).
1167 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1169 JH/15 Rework TLS client-side context management. Stop using a global, and
1170 explicitly pass a context around. This enables future use of TLS for
1171 connections to service-daemons (eg. malware scanning) while a client smtp
1172 connection is using TLS; with cutthrough connections this is quite likely.
1174 JH/16 Fix ARC verification to do AS checks in reverse order.
1176 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1178 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1179 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1180 Previously the "utf8" would be re-prepended for every additional message.
1182 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1183 Previously thery were accepted, resulting in issues when attempting to
1184 forward messages to a non-supporting MTA.
1186 PP/02 Let -n work with printing macros too, not just options.
1188 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1189 one parent address was copied, and bogus data was used at delivery-logging
1190 time. Either a crash (after delivery) or bogus log data could result.
1191 Discovery and analysis by Tim Stewart.
1193 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1194 Previously if the string ended mid-character, we did not insert the
1195 promised '?' replacement.
1197 PP/04 Documentation: current string operators work on bytes, not codepoints.
1199 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1200 should pack well giving a smaller memory footprint so better caching and
1201 therefore performance. Group the declarations where this can't be done so
1202 that the byte-sized flag variables are not interspersed among pointer
1203 variables, giving a better chance of good packing by the compiler.
1205 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1206 non-null, to avoid issues with sites running BATV. Previously reports were
1207 sent with an empty envelope sender so looked like bounces.
1209 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1210 The ignore_error flag wasn't being returned from the filter subprocess so
1211 was not set for later routers. Investigation and fix by Matthias Kurz.
1213 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1214 and a msg:complete for the whole, when a message is manually removed using
1215 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1217 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1218 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1219 hope that is portable enough.
1221 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1222 requiring a known-CA anchor certificate; make it now rely entirely on the
1223 TLSA as an anchor. Checking the name on the leaf cert against the name
1224 on the A-record for the host is still done for TA (but not for EE mode).
1226 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1227 included in delivery lines for non-proxied connections, when compiled with
1228 SUPPORT_SOCKS and running with proxy logging enabled.
1230 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1231 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1232 move the existing event to fire before the normal logging of message
1233 failure so that custom logging is bracketed by normal logging.
1235 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1236 msg:fail:internal event. Developement by Matthias Kurz.
1238 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1239 far too small for todays use of crypto signatures stored there. Go all
1240 the way to the max DNS message size of 64kB, even though this might be
1241 overmuch for IOT constrained device use.
1243 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1244 copy a string over itself. The library routine is documented as not
1245 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1247 JH/32 For main options check_spool_space and check_inode_space, where the
1248 platform supports 64b integers, support more than the previous 2^31 kB
1249 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1250 the previous G, M, k.
1252 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1253 $authenticated_fail_id variable on authentication failure. Previously
1256 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1257 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1258 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1259 GNUTLS_SEC_PARAM_MEDIUM.
1261 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1262 side. Previously we would continue as if no SNI had been received.
1264 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1265 "<" character, which should be a list-separator specification, we walked
1266 off past the nul-terimation.
1268 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1269 causes) even when the retry time is not yet met. Previously they were
1270 not, meaning that when (say) an account was over-quota and temp-rejecting,
1271 and multiple senders' messages were queued, only one sender would get
1272 notified on each configured delay_warning cycle.
1274 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1276 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1277 cipher-suites, an error can be left on the stack even for a succeeding
1278 accept; this results in impossible error messages when a later operation
1281 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1282 return error codes indicating retry. Under TLS1.3 this becomes required.
1284 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1285 it only wrote the new authenticators, resulting in a lack of tracking of
1286 peer changes of ESMTP extensions until the next cache flush.
1288 JH/41 Fix the loop reading a message header line to check for integer overflow,
1289 and more-often against header_maxsize. Previously a crafted message could
1290 induce a crash of the recive process; now the message is cleanly rejected.
1292 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1293 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1299 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1300 When redis_servers is set to a list of > 1 element, and the Redis servers
1301 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1302 case of MOVED into a DEFER case instead, thus moving the query onto the
1303 next server in the list. For a cluster of N elements, all N servers must
1304 be defined in redis_servers.
1306 GF/02 Catch and remove uninitialized value warning in exiqsumm
1307 Check for existence of @ARGV before looking at $ARGV[0]
1309 JH/01 Replace the store_release() internal interface with store_newblock(),
1310 which internalises the check required to safely use the old one, plus
1311 the allocate and data copy operations duplicated in both (!) of the
1312 extant use locations.
1314 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1315 modifier. This matches the restriction on the commandline.
1317 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1318 Previously only the last row was returned.
1320 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1321 we assumed that tags in the header were well-formed, and parsed the
1322 element content after inspecting only the first char of the tag.
1323 Assumptions at that stage could crash the receive process on malformed
1326 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1327 While running the DKIM ACL we operate on the Permanent memory pool so that
1328 variables created with "set" persist to the DATA ACL. Also (at any time)
1329 DNS lookups that fail create cache records using the Permanent pool. But
1330 expansions release any allocations made on the current pool - so a dnsdb
1331 lookup expansion done in the DKIM ACL releases the memory used for the
1332 DNS negative-cache, and bad things result. Solution is to switch to the
1333 Main pool for expansions.
1334 While we're in that code, add checks on the DNS cache during store_reset,
1335 active in the testsuite.
1336 Problem spotted, and debugging aided, by Wolfgang Breyha.
1338 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1339 When none of the hosts presented to a transport match an already-open
1340 connection, close it and proceed with the list. Previously we would
1341 queue the message. Spotted by Lena with Yahoo, probably involving
1344 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1345 Previously a spurious "250 OK id=" response was appended to the proper
1348 JH/08 The "support for" informational output now, which built with Content
1349 Scanning support, has a line for the malware scanner interfaces compiled
1350 in. Interface can be individually included or not at build time.
1352 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1353 by the template makefile "src/EDITME". The "STREAM" support for an older
1354 ClamAV interface method is removed.
1356 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1357 rows affected is given instead).
1359 JH/11 The runtime Berkeley DB library version is now additionally output by
1360 "exim -d -bV". Previously only the compile-time version was shown.
1362 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1363 SMTP connection. Previously, when one had more recipients than the
1364 first, an abortive onward connection was made. Move to full support for
1365 multiple onward connections in sequence, handling cutthrough connection
1366 for all multi-message initiating connections.
1368 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1369 routers. Previously, a multi-recipient message would fail to match the
1370 onward-connection opened for the first recipient, and cause its closure.
1372 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1373 a timeout on read on a GnuTLS initiating connection, resulting in the
1374 initiating connection being dropped. This mattered most when the callout
1375 was marked defer_ok. Fix to keep the two timeout-detection methods
1378 JH/15 Relax results from ACL control request to enable cutthrough, in
1379 unsupported situations, from error to silently (except under debug)
1380 ignoring. This covers use with PRDR, frozen messages, queue-only and
1383 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1385 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1386 metadata, resulting in a crash in free().
1388 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1389 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1390 Broken also in d185889f4, with init system revamp.
1392 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1393 Previously we abruptly closed the connection after reading a malware-
1394 found indication; now we go on to read the "scan ok" response line,
1397 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1398 ACL. Previously, a crash would result.
1400 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1401 macro text after a replacement (previously it was only once per line) and
1402 by skipping builtin macros when searching for an uppercase lead character.
1404 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1405 for the build is renamed.
1407 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1408 was allocated for every new TLS startup, meaning one per message. Fix
1409 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1411 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1412 reported the original. Fix to report (as far as possible) the ACL
1413 result replacing the original.
1415 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1416 OpenSSL. Certificate information is loaded for every new TLS startup,
1417 and the resources needed to be freed.
1419 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1421 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1424 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1425 DATA response info to the (existing) per-recipient response info for
1426 the "C=" log element. It can have useful tracking info from the
1427 destination system. Patch from Simon Arlott.
1429 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1430 length value. Previously this would segfault.
1432 HS/02 Support Avast multiline protoocol, this allows passing flags to
1433 newer versions of the scanner.
1435 JH/28 Ensure that variables possibly set during message acceptance are marked
1436 dead before release of memory in the daemon loop. This stops complaints
1437 about them when the debug_store option is enabled. Discovered specifically
1438 for sender_rate_period, but applies to a whole set of variables.
1439 Do the same for the queue-runner and queue-list loops, for variables set
1440 from spool message files. Do the same for the SMTP per-message loop, for
1441 certain variables indirectly set in ACL operations.
1443 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1444 as a multi-recipient message from a mailinglist manager). The coding had
1445 an arbitrary cutoff number of characters while checking for more input;
1446 enforced by writing a NUL into the buffer. This corrupted long / fast
1447 input. The problem was exposed more widely when more pipelineing of SMTP
1448 responses was introduced, and one Exim system was feeding another.
1449 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1450 receiving system, and refused recipients seen by the sending system
1451 (propating to people being dropped from mailing lists).
1452 Discovered and pinpointed by David Carter.
1454 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1455 replaced by the ${authresults } expansion.
1457 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1459 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1460 allows proper process termination in container environments.
1462 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1463 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1465 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1466 and "err_perm", deprecated since 4.83 when the RFC-defined words
1467 "temperror" and "permerror" were introduced.
1469 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1470 transport-filters or DKIM-signing. The restriction was lost in the
1471 consolidation of verify-callout and delivery SMTP handling.
1472 Extend the restriction to also cover ARC-signing.
1474 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1475 in defer=pass mode supply a 450 to the initiator. Previously the message
1478 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1479 tls_require_ciphers is used as before.
1481 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1482 "pass_unscanned". Only tmpfails from the scanner are written to
1483 the paniclog, as they may require admin intervention (permission
1484 denied, license issues). Other scanner errors (like decompression
1485 bombs) do not cause a paniclog entry.
1487 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1488 Previously it was possible to log spurious information in receive log
1491 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1492 triggered odd behaviour from Outlook Express clients.
1494 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1497 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1498 since the IETF WG has not yet settled on that versus the original
1499 "bare" representation.
1501 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1502 Previously the millisecond value corrupted the output.
1503 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1504 corrupted the output.
1510 JH/01 Rework error string handling in TLS interface so that the caller in
1511 more cases is responsible for logging. This permits library-sourced
1512 string to be attached to addresses during delivery, and collapses
1513 pairs of long lines into single ones.
1515 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1516 during configuration. Wildcards are allowed and expanded.
1518 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1519 This permits better logging.
1521 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1522 do not close it down and have the child transport start it up again on
1523 the passed-on TCP connection. Instead, proxy the child (and any
1524 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1525 affected: the continued delivery log lines do not have any DNSSEC, TLS
1526 Certificate or OCSP information. TLS cipher information is still logged.
1528 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1529 identical IP addresses on different listening ports. Will also affect
1532 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1533 add noisy ifdef guards to special-case this sillyness.
1534 Patch from Bernd Kuhls.
1536 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1537 than 255 are no longer allowed.
1539 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1540 Disable the session-cache too, which might reduce our load. Since we
1541 currrectly use a new context for every connection, both as server and
1542 client, there is no benefit for these.
1543 GnuTLS appears to not support tickets server-side by default (we don't
1544 call gnutls_session_ticket_enable_server()) but client side is enabled
1545 by default on recent versions (3.1.3 +) unless the PFS priority string
1548 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1549 <https://reproducible-builds.org/specs/source-date-epoch/>.
1551 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1552 the check for any unsuccessful recipients did not notice the limit, and
1553 erroneously found still-pending ones.
1555 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1556 MSG_MORE. Only in-clear (not on TLS connections).
1558 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1559 transport-filter is involved we can buffer the headers in memory for
1560 creating the signature, and read the spool data file once for the
1561 signature and again for transmission.
1563 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1564 4.77 as the kernel support then wasn't solid, having issues in 64bit
1565 mode. Now, it's been long enough. Add support for FreeBSD also.
1567 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1568 case where the routing stage had gathered several addresses to send to
1569 a host before calling the transport for the first, we previously failed
1570 to close down TLS in the old transport process before passing the TCP
1571 connection to the new process. The new one sent a STARTTLS command
1572 which naturally failed, giving a failed delivery and bloating the retry
1573 database. Investigation and fix prototype from Wolfgang Breyha.
1575 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1576 false-negatives in the check that the sender had not preempted a response
1577 or prompt from Exim (running as a server), due to that code's lack of
1578 awareness of the SMTP input buffering.
1580 PP/04 Add commandline_checks_require_admin option.
1581 Exim drops privileges sanely, various checks such as -be aren't a
1582 security problem, as long as you trust local users with access to their
1583 own account. When invoked by services which pass untrusted data to
1584 Exim, this might be an issue. Set this option in main configuration
1585 AND make fixes to the calling application, such as using `--` to stop
1588 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1589 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1590 responses to those, into a single TLS record each way (this usually means
1591 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1594 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1595 affects you only if you're dancing at the edge of the param size limits.
1596 If you are, and this message makes sense to you, then: raise the
1597 configured limit or use OpenSSL 1.1. Nothing we can do for older
1600 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1601 cmdline element to get the documented default one. Previously it was
1604 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1605 are unsuable for non-message-specific reasons - eg. connection timeout,
1606 banner-time rejection.
1608 JH/16 Fix logging of delivery remote port, when specified by router, under
1611 PP/06 Repair manualroute's ability to take options in any order, even if one
1612 is the name of a transport.
1615 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1617 JH/17 Change the list-building routines interface to use the expanding-string
1618 triplet model, for better allocation and copying behaviour.
1620 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1621 Previously it was constructed the first time a possibly-matching string
1622 was met in the configuration file input during startup; now it is done
1625 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1626 compatible one, to avoid the (poorly documented) possibility of a config
1627 file in the working directory redirecting the DB files, possibly correpting
1628 some existing file. CVE-2017-10140 assigned for BDB.
1630 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1631 cache-hot. Previously, although the result was properly cached, the
1632 initial verify call returned a defer.
1634 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1635 the main verify for receipient in uncached-mode.
1637 JH/22 Retire historical build files to an "unsupported" subdir. These are
1638 defined as "ones for which we have no current evidence of testing".
1640 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1641 if present. Previously it was ignored.
1643 JH/24 Start using specified-initialisers in C structure init coding. This is
1644 a C99 feature (it's 2017, so now considered safe).
1646 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1647 if was a fixed-sized field and bitmask ops via macros; it is now more
1650 PP/07 GitHub PR 56: Apply MariaDB build fix.
1651 Patch provided by Jaroslav Škarvada.
1653 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1654 during Coverity cleanups [4.87 JH/47]
1655 Diagnosis and fix provided by Michael Fischer v. Mollard.
1657 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1658 the right size to place the terminating semicolon on its own folded
1659 line, the header hash was calculated to an incorrect value thanks to
1660 the (relaxed) space the fold became.
1662 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1663 and confused the parent.
1665 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1666 which could crash as a result. This could lead to undeliverable messages.
1668 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1671 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1672 matters on fast-turnover and PID-randomising systems, which were getting
1673 out-of-order delivery.
1675 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1676 a possibly-overlapping copy. The symptom was that "Remote host closed
1677 connection in response to HELO" was logged instead of the actual 4xx
1680 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1681 Previously only that bufferd was discarded, resulting in SYMTP command
1684 JH/32 DKIM: when a message has multiple signatures matching an identity given
1685 in dkim_verify_signers, run the dkim acl once for each. Previously only
1686 one run was done. Bug 2189.
1688 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1689 "panic the current process" to "deliberately defer". The panic log is
1690 still written with the problem list name; the mail and reject logs now
1691 get a temp-reject line for the message that was being handled, saying
1692 something like "domains check lookup or other defer". The SMTP 451
1693 message is still "Temporary local problem".
1695 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1696 A crafted sequence of BDAT commands could result in in-use memory beeing
1697 freed. CVE-2017-16943.
1699 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1700 from SMTP input. Previously it was always done; now only done for DATA
1701 and not BDAT commands. CVE-2017-16944.
1703 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1704 to the message (such as an overlong header line). Previously this was
1705 not done and we did not exit BDAT mode. Followon from the previous item
1706 though a different problem.
1712 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1713 than -2003 did; needs libidn2 in addition to libidn.
1715 JH/02 The path option on a pipe transport is now expanded before use.
1717 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1718 Patch provided by "Björn", documentation fix added too.
1720 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1721 missing a wire-to-host endian conversion.
1723 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1724 close after a BDAT command line could be taken as a following command,
1725 giving a synch failure. Fix by only checking for synch immediately
1726 before acknowledging the chunk.
1728 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1729 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1730 macro. Patches provided by Josh Soref.
1732 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1733 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1734 by IANA as a service extension. However, John Klensin suggests that we
1737 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1738 the dkim code may be unix-mode line endings rather than smtp wire-format
1739 CRLF, so prepend a CR to any bare LF.
1741 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1742 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1744 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1745 the first were themselves being wrongly included in the feed into dkim
1746 processing; with most chunk sizes in use this resulted in an incorrect
1747 body hash calculated value.
1749 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1750 DKIM signature block, for verification. Although advised against by
1751 standards it is specifically not ruled illegal.
1753 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1755 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1756 missing a body hash (the bh= tag).
1758 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1759 It seems that HAProxy sends the Proxy Protocol information in clear and
1760 only then does a TLS startup, so do the same.
1762 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1763 TCP connections (such as for Spamd) unless the daemon successfully set
1764 Fast Open mode on its listening sockets. This fixes breakage seen on
1765 too-old kernels or those not configured for Fast Open, at the cost of
1766 requiring both directions being enabled for TFO, and TFO never being used
1767 by non-daemon-related Exim processes.
1769 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1770 endings, at least on the first header line. Try to canonify any that get
1771 past that check, despite the cost.
1773 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1774 now limited to an arbitrary five deep, while parsing addresses with the
1775 strip_excess_angle_brackets option enabled.
1777 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1778 instead leave the unprompted TLS handshake in socket buffer for the
1779 TLS library to consume.
1781 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1783 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1785 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1786 region in one operation, for speed, and this leaves assigned pointers
1787 dangling. Add checks run only under the testsuite which checks all
1788 variables at a store-reset and panics on a dangling pointer; add code
1789 explicitly nulling out all the variables discovered. Fixes one known
1790 bug: a transport crash, where a dangling pointer for $sending_ip_address
1791 originally assigned in a verify callout, is re-used.
1793 PP/06 Drop '.' from @INC in various Perl scripts.
1795 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1797 PP/08 Reduce a number of compilation warnings under clang; building with
1798 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1799 should be warning-free.
1801 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1803 HS/01 Fix portability problems introduced by PP/08 for platforms where
1804 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1806 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1807 chunk. This allows us to accept broken chunked messages. We need a more
1808 general solution here.
1810 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1811 already-broken messages in the queue.
1813 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1815 JH/19 Fix reference counting bug in routing-generated-address tracking.
1821 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1822 supports it and a size is available (ie. the sending peer gave us one).
1824 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1825 years of being deprecated). The replacements are the ACLs
1826 acl_smtp_mime and acl_not_smtp_mime.
1828 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1829 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1830 or even an in-clear connection were permitted. Now, if the host lookup
1831 was dnssec and dane was requested then the host is only used if the
1832 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1833 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1834 if one fails this test.
1835 This means that a poorly-configured remote DNS will make it incommunicado;
1836 but it protects against a DNS-interception attack on it.
1838 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1839 non-noisy when a race steals the message being considered.
1841 JH/05 If main configuration option tls_certificate is unset, generate a
1842 self-signed certificate for inbound TLS connections.
1844 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1845 in rewrites and routers.
1847 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1848 and logged a warning sing 4.83; now they are a configuration file error.
1850 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1851 (lacking @domain). Apply the same qualification processing as RCPT.
1853 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1855 JH/10 Support ${sha256:} applied to a string (as well as the previous
1858 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1859 a cutthrough deliver is pending, as we always want to make a connection.
1860 This also avoids re-routing the message when later placing the cutthrough
1861 connection after a verify cache hit.
1862 Do not update it with the verify result either.
1864 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1865 when routing results in more than one destination address.
1867 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1868 signing (which inhibits the cutthrough capability). Previously only
1869 the presence of an option was tested; now an expansion evaluating as
1870 empty is permissible (obviously it should depend only on data available
1871 when the cutthrough connection is made).
1873 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1874 the relevant preceding SMTP command did not note the pipelining mode.
1876 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1877 Previously they were not counted.
1879 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1880 as one having no matching records. Previously we deferred the message
1881 that needed the lookup.
1883 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1884 distinguished as "(=".
1886 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1887 for missing MX records. Previously it only worked for missing A records.
1889 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1891 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1892 after the data-go-ahead and data-ack. Patch from Jason Betts.
1894 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1895 even for a "none" policy. Patch from Tony Meyer.
1897 JH/22 Fix continued use of a connection for further deliveries. If a port was
1898 specified by a router, it must also match for the delivery to be
1901 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1902 When one of the recipients of a message was unsuitable for the connection
1903 (has no matching addresses), we lost track of needing to mark it
1904 deferred. As a result mail would be lost.
1906 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1908 JH/25 Decoding ACL controls is now done using a binary search; the source code
1909 takes up less space and should be simpler to maintain. Merge the ACL
1910 condition decode tables also, with similar effect.
1912 JH/26 Fix problem with one_time used on a redirect router which returned the
1913 parent address unchanged. A retry would see the parent address marked as
1914 delivered, so not attempt the (identical) child. As a result mail would
1917 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1918 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1919 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1922 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1923 Main config options check_{log,spool}_{inodes,space} are now
1924 100 inodes, 10MB unless set otherwise in the configuration.
1926 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1927 Previously it only applied to the main-section connection policy
1930 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1932 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1933 by me. Added RFC7919 DH primes as an alternative.
1935 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1936 are not in the system include path.
1938 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1939 GnuTLS, when a session startup failed (eg because the client disconnected)
1940 Exim did stdio operations after fclose. This was exposed by a recent
1941 change which nulled out the file handle after the fclose.
1943 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1944 signed directly by the cert-signing cert, rather than an intermediate
1945 OCSP-signing cert. This is the model used by LetsEncrypt.
1947 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
1949 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
1950 an incoming connection.
1952 HS/02 Bug 1802: Do not half-close the connection after sending a request
1955 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
1956 fallback to "prime256v1".
1958 JH/34 SECURITY: Use proper copy of DATA command in error message.
1959 Could leak key material. Remotely exploitable. CVE-2016-9963.
1965 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
1966 and 3.4.4 - once the server is enabled to respond to an OCSP request
1967 it does even when not requested, resulting in a stapling non-aware
1968 client dropping the TLS connection.
1970 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
1971 support variable-length bit vectors. No functional change.
1973 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
1974 The I= interface field on outgoing lines is now after the H= remote
1975 host field, same as incoming lines. There is a separate
1976 outgoing_interface log selector which allows you to disable the
1979 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
1980 If not running log_selector +smtp_connection the mainlog would be held
1981 open indefinitely after a "too many connections" event, including to a
1982 deleted file after a log rotate. Leave the per net connection logging
1983 leaving it open for efficiency as that will be quickly detected by the
1984 check on the next write.
1986 HS/01 Bug 1671: Fix post transport crash.
1987 Processing the wait-<transport> messages could crash the delivery
1988 process if the message IDs didn't exist for some reason. When
1989 using 'split_spool_directory=yes' the construction of the spool
1990 file name failed already, exposing the same netto behaviour.
1992 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
1993 mime_regex ACL conditions.
1995 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
1996 to DSN fail messages (bounces): remote IP, remote greeting, remote response
1997 to HELO, local diagnostic string.
1999 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2000 log line to debug. Even when configured with a tls authenticator many
2001 client connections are expected to not authenticate in this way, so
2002 an authenticate fail is not an error.
2004 HS/02 Add the Exim version string to the process info. This way exiwhat
2005 gives some more detail about the running daemon.
2007 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2008 matter for fast-change records such as DNSBLs.
2010 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2011 retry record, even if constant. There may be multiple transports with
2012 different interface settings and the retry behaviour needs to be kept
2015 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2017 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2019 JH/10 Bug 840: fix log_defer_output option of pipe transport
2021 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2022 hosts_randomize. This matters a lot when combined with mua_wrapper.
2024 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2025 ${quote_pgsql:<string>} operator.
2027 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2029 JH/14 Change header file name for freeradius-client. Relevant if compiling
2030 with Radius support; from the Gentoo tree and checked under Fedora.
2032 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2034 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2035 option transport_filter, meaning no filtering.
2037 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2039 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2040 defaults to "*" (all hosts). The variable is now available when not built
2041 with TLS, default unset, mainly to enable keeping the testsuite sane.
2042 If a server certificate is not supplied (via tls_certificate) an error is
2043 logged, and clients will find TLS connections fail on startup. Presumably
2044 they will retry in-clear.
2045 Packagers of Exim are strongly encouraged to create a server certificate
2046 at installation time.
2048 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2049 with the $config_file variable.
2051 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2052 in transport context, after the attempt, and per-recipient. The latter type
2053 is per host attempted. The event data is the error message, and the errno
2054 information encodes the lookup type (A vs. MX) used for the (first) host,
2055 and the trailing two digits of the smtp 4xx response.
2057 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2058 to write to mainlog (or rejectlog, paniclog) in the window between file
2059 creation and permissions/ownership being changed. Particularly affects
2060 installations where exicyclog is run as root, rather than exim user;
2061 result is that the running daemon panics and dies.
2063 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2065 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2066 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2067 "pri" and "weight". Note that the previous implicit priority given by the
2068 list order is no longer honoured.
2070 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2071 for DKIM processing.
2073 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2074 by defining SUPPORT_SOCKS.
2076 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2077 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2078 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2079 variables are renamed to proxy_{local,external}_{address,port}.
2081 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2082 for a build by defining SUPPORT_I18N
2084 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2085 of the query string, and make ${quote_redis:} do that quoting.
2087 JH/29 Move Events support from Experimental to mainline, enabled by default
2088 and removable for a build by defining DISABLE_EVENT.
2090 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2092 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2093 cached by the daemon.
2095 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2096 by defining LOOKUP_REDIS. The libhiredis library is required.
2098 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2099 keys are given for lookup.
2101 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2102 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2103 only supported when built with TLS support. The PolarSSL SHA routines
2104 are still used when the TLS library is too old for convenient support.
2106 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2107 openssl_options), for security. OpenSSL forces this from version 1.1.0
2108 server-side so match that on older versions.
2110 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2111 allocation for $value could be released as the expansion processing
2112 concluded, but leaving the global pointer active for it.
2114 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2115 and to use the domains and local_parts ACL conditions.
2117 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2118 incorrectly not doubled on cutthrough transmission, hence seen as a
2119 body-termination at the receiving system - resulting in truncated mails.
2120 Commonly the sender saw a TCP-level error, and retransmitted the message
2121 via the normal store-and-forward channel. This could result in duplicates
2122 received - but deduplicating mailstores were liable to retain only the
2123 initial truncated version.
2125 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2127 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2129 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2130 we're in there, support oversigning also; bug 1309.
2132 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2134 HS/04 Add support for keep_environment and add_environment options.
2136 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2137 either intentional arithmetic overflow during PRNG, or testing config-
2140 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2141 delivery resulted in actual delivery. Cancel cutthrough before DATA
2144 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2145 rejects a recipient: pass the reject to the originator.
2147 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2148 Many were false-positives and ignorable, but it's worth fixing the
2151 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2152 for the new environment-manipulation done at startup. Move the routines
2153 from being local to tls.c to being global via the os.c file.
2155 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2156 an extract embedded as result-arg for a map, the first arg for extract
2157 is unavailable so we cannot tell if this is a numbered or keyed
2158 extraction. Accept either.
2164 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2167 JH/02 The smtp transport option "multi_domain" is now expanded.
2169 JH/03 The smtp transport now requests PRDR by default, if the server offers
2172 JH/04 Certificate name checking on server certificates, when exim is a client,
2173 is now done by default. The transport option tls_verify_cert_hostnames
2174 can be used to disable this per-host. The build option
2175 EXPERIMENTAL_CERTNAMES is withdrawn.
2177 JH/05 The value of the tls_verify_certificates smtp transport and main options
2178 default to the word "system" to access the system default CA bundle.
2179 For GnuTLS, only version 3.0.20 or later.
2181 JH/06 Verification of the server certificate for a TLS connection is now tried
2182 (but not required) by default. The verification status is now logged by
2183 default, for both outbound TLS and client-certificate supplying inbound
2186 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2189 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2190 Status Notification (bounce) messages are now MIME format per RFC 3464.
2191 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2192 under the control of the dsn_advertise_hosts option, and routers may
2193 have a dsn_lasthop option.
2195 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2196 default, modifiable by a malware= option. The list separator for
2197 the options can now be changed in the usual way. Bug 68.
2199 JH/10 The smtp_receive_timeout main option is now expanded before use.
2201 JH/11 The incoming_interface log option now also enables logging of the
2202 local interface on delivery outgoing connections.
2204 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2205 if the interface and destination host and port all match.
2207 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2210 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2211 Patch from Andrew Lewis.
2213 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2214 now supports optional time-restrictions, weighting, and priority
2215 modifiers per server. Patch originally by <rommer@active.by>.
2217 JH/16 The spamd_address main option now supports a mixed list of local
2218 and remote servers. Remote servers can be IPv6 addresses, and
2219 specify a port-range.
2221 JH/17 Bug 68: The spamd_address main option now supports an optional
2222 timeout value per server.
2224 JH/18 Bug 1581: Router and transport options headers_add/remove can
2225 now have the list separator specified.
2227 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2230 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2233 JH/21 Support for the A6 type of dns record is withdrawn.
2235 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2236 rather than the verbs used.
2238 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2239 from 255 to 1024 chars.
2241 JH/24 Verification callouts now attempt to use TLS by default.
2243 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2244 are generic router options now. The defaults didn't change.
2246 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2247 Original patch from Alexander Shikoff, worked over by JH.
2249 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2250 files if ZCAT_COMMAND is not executable.
2252 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2254 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2256 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2257 Normally benign, it bites when the pair was led to by a CNAME;
2258 modern usage is to not canonicalize the domain to a CNAME target
2259 (and we were inconsistent anyway for A-only vs AAAA+A).
2261 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2263 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2264 when evaluating $sender_host_dnssec.
2266 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2267 $sender_helo_dnssec variable.
2269 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2271 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2273 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2275 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2276 documented as working, but never had. Support all but $spam_report.
2278 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2279 added for tls authenticator.
2281 HS/03 Add perl_taintmode main config option
2287 TL/01 When running the test suite, the README says that variables such as
2288 no_msglog_check are global and can be placed anywhere in a specific
2289 test's script, however it was observed that placement needed to be near
2290 the beginning for it to behave that way. Changed the runtest perl
2291 script to read through the entire script once to detect and set these
2292 variables, reset to the beginning of the script, and then run through
2293 the script parsing/test process like normal.
2295 TL/02 The BSD's have an arc4random API. One of the functions to induce
2296 adding randomness was arc4random_stir(), but it has been removed in
2297 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2298 function when detected.
2300 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2301 cause callback expansion.
2303 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2304 syntax errors in an expansion can be treated as a string instead of
2305 logging or causing an error, due to the internal use of bool_lax
2306 instead of bool when processing it.
2308 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2309 server certificates when making smtp deliveries.
2311 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2313 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2315 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2317 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2318 Merged patch from Sebastian Wiedenroth.
2320 JH/05 Fix results-pipe from transport process. Several recipients, combined
2321 with certificate use, exposed issues where response data items split
2322 over buffer boundaries were not parsed properly. This eventually
2323 resulted in duplicates being sent. This issue only became common enough
2324 to notice due to the introduction of connection certificate information,
2325 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2327 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2328 size buffer was used, resulting in syntax errors when an expansion
2331 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2332 version 3.3.6 or later.
2334 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2335 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2336 both become "event_action", the variables become $event_name, $event_data
2337 and $event_defer_errno. There is a new variable $verify_mode, usable in
2338 routers, transports and related events. The tls:cert event is now also
2339 raised for inbound connections, if the main configuration event_action
2342 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2343 early OCSP support, but no stapling (appears to be less than 1.0.0).
2345 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2346 server certificate names available under the smtp transport option
2347 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2350 JH/10 Time-related extraction expansions from certificates now use the main
2351 option "timezone" setting for output formatting, and are consistent
2352 between OpenSSL and GnuTLS compilations. Bug 1541.
2354 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2355 encoded parameter in the incoming message. Bug 1558.
2357 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2358 include certificate info, eximon was claiming there were spoolfile
2361 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2363 JH/14 Log delivery-related information more consistently, using the sequence
2364 "H=<name> [<ip>]" wherever possible.
2366 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2367 are problematic for Debian distribution, omit them from the release
2370 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2372 JH/16 Fix string representation of time values on 64bit time_t architectures.
2375 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2376 output list separator was used.
2381 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2382 checkers that were complaining about end of non-void function with no
2385 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2386 This was a regression introduced in 4.83 by another bugfix.
2388 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2390 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2391 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2397 TF/01 Correctly close the server side of TLS when forking for delivery.
2399 When a message was received over SMTP with TLS, Exim failed to clear up
2400 the incoming connection properly after forking off the child process to
2401 deliver the message. In some situations the subsequent outgoing
2402 delivery connection happened to have the same fd number as the incoming
2403 connection previously had. Exim would try to use TLS and fail, logging
2404 a "Bad file descriptor" error.
2406 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2407 utilities have not been installed.
2409 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2410 temporary space as the ACL may create new global variables.
2412 TL/01 LDAP support uses per connection or global context settings, depending
2413 upon the detected version of the libraries at build time.
2415 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2416 to extract and use the src ip:port in logging and expansions as if it
2417 were a direct connection from the outside internet. PPv2 support was
2418 updated based on HAProxy spec change in May 2014.
2420 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2422 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2423 Properly escape header and check for NULL return.
2425 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2428 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2430 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2431 characters in header names, implemented as a verify condition.
2432 Contributed by Michael Fischer v. Mollard.
2434 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2435 results permerror and temperror. Previous values are deprecated but
2436 still accepted. In a future release, err_perm and err_temp will be
2437 completely removed, which will be a backward incompatibility if the
2438 ACL tests for either of these two old results. Patch contributed by
2439 user bes-internal on the mailing list.
2441 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2443 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2444 selectors, in both main and reject logs.
2446 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2449 JH/07 Add malware type "sock" for talking to simple daemon.
2451 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2453 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2454 routers/transports under cutthrough routing.
2456 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2457 numbers. Touch up "bool" conditional to keep the same definition.
2459 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2461 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2463 JH/12 Expand items in router/transport headers_add or headers_remove lists
2464 individually rather than the list as a whole. Bug 1452.
2466 Required for reasonable handling of multiple headers_ options when
2467 they may be empty; requires that headers_remove items with embedded
2468 colons must have them doubled (or the list-separator changed).
2470 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2471 view the policy declared in the DMARC record. Currently, $dmarc_status
2472 is a combined value of both the record presence and the result of the
2475 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2477 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2478 dnslookup router and the smtp transport (applying to the forward
2481 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2482 of ldap servers used for a specific lookup. Patch provided by Heiko
2485 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2486 New variable $lookup_dnssec_authenticated for observability.
2488 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2489 Patch submitted by Lars Timman.
2491 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2493 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2494 Requires trusted mode and valid format message id, aborts otherwise.
2495 Patch contributed by Heiko Schlichting.
2497 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2498 certextract with support for various fields. Bug 1358.
2500 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2501 is requested by default, modifiable by smtp transport option
2504 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2505 operate on certificate variables to give certificate fingerprints
2506 Also new ${sha256:cert_variable}.
2508 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2510 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2511 Christian Aistleitner.
2513 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2515 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2516 file. Patch from Wolfgang Breyha.
2518 JH/25 Expand the coverage of the delivery $host and $host_address to
2519 client authenticators run in verify callout. Bug 1476.
2521 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2522 align with daemon_smtp_ports. Bug 72.
2524 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2525 support and error reporting did not work properly.
2527 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2528 and is readable. Patch from Andrew Colin Kissa.
2530 TL/14 Enhance documentation of ${run expansion and how it parses the
2531 commandline after expansion, particularly in the case when an
2532 unquoted variable expansion results in an empty value.
2534 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2536 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2537 by renaming away from it. Addresses GH issue 15, reported by
2540 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2542 TL/15 SECURITY: prevent double expansion in math comparison functions
2543 (can expand unsanitized data). Not remotely exploitable.
2550 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2552 PP/02 Make -n do something, by making it not do something.
2553 When combined with -bP, the name of an option is not output.
2555 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2558 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2559 $sender_host_name and config options to manage this, and basic check
2562 PP/05 DSCP support for outbound connections and control modifier for inbound.
2564 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2565 (Only plugin which currently uses this is kerberos4, which nobody should
2566 be using, but we should make it available and other future plugins might
2567 conceivably use it, even though it would break NAT; stuff *should* be
2568 using channel bindings instead).
2570 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2571 name; added for Sendmail compatibility; requires admin caller.
2572 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2573 just ignore it); requires trusted caller.
2574 Also parse but ignore: -Ac -Am -X<logfile>
2577 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2579 TL/02 Add +smtp_confirmation as a default logging option.
2581 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2582 Patch by Magnus Holmgren from 2007-02-20.
2584 TL/04 Bugzilla 1281 - Spec typo.
2585 Bugzilla 1283 - Spec typo.
2586 Bugzilla 1290 - Spec grammar fixes.
2588 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2590 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2592 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2593 mailing list by by Dmitry Isaikin.
2595 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2597 JH/02 Support "G" suffix to numbers in ${if comparisons.
2599 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2601 NM/01 Bugzilla 1197 - Spec typo
2602 Bugzilla 1196 - Spec examples corrections
2604 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2606 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2607 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2610 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2611 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2613 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2614 "acl {{name}{arg}...}", and optional args on acl condition
2617 JH/05 Permit multiple router/transport headers_add/remove lines.
2619 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2621 JH/07 Avoid using a waiting database for a single-message-only transport.
2622 Performance patch from Paul Fisher. Bugzilla 1262.
2624 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2627 JH/09 Add $headers_added variable, with content from use of ACL modifier
2628 add_header (but not yet added to the message). Bugzilla 199.
2630 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2631 Pulled from Bugzilla 817 by Wolfgang Breyha.
2633 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2635 (nb: this is the same fix as in Exim 4.80.1)
2637 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2640 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2643 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2645 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2646 advertise SMTP AUTH mechanism to us, instead of a generic
2647 protocol violation error. Also, make Exim more robust to bad
2648 data from the Dovecot auth socket.
2650 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2652 When a queue runner is handling a message, Exim first routes the
2653 recipient addresses, during which it prunes them based on the retry
2654 hints database. After that it attempts to deliver the message to
2655 any remaining recipients. It then updates the hints database using
2658 So if a recipient address works intermittently, it can get repeatedly
2659 deferred at routing time. The retry hints record remains fresh so the
2660 address never reaches the final cutoff time.
2662 This is a fairly common occurrence when a user is bumping up against
2663 their storage quota. Exim had some logic in its local delivery code
2664 to deal with this. However it did not apply to per-recipient defers
2665 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2667 This change adds a proper retry rule check during routing so that the
2668 final cutoff time is checked against the message's age. We only do
2669 this check if there is an address retry record and there is not a
2670 domain retry record; this implies that previous attempts to handle
2671 the address had the retry_use_local_parts option turned on. We use
2672 this as an approximation for the destination being like a local
2673 delivery, as in LMTP.
2675 I suspect this new check makes the old local delivery cutoff check
2676 redundant, but I have not verified this so I left the code in place.
2678 TF/02 Correct gecos expansion when From: is a prefix of the username.
2680 Test 0254 submits a message to Exim with the header
2684 When I ran the test suite under the user fanf2, Exim expanded
2685 the header to contain my full name, whereas it should have added
2686 a Resent-Sender: header. It erroneously treats any prefix of the
2687 username as equal to the username.
2689 This change corrects that bug.
2691 GF/01 DCC debug and logging tidyup
2692 Error conditions log to paniclog rather than rejectlog.
2693 Debug lines prefixed by "DCC: " to remove any ambiguity.
2695 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2697 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2698 Bug spotted by Jeremy Harris; was flawed since initial commit.
2699 Would have resulted in OCSP responses post-SNI triggering an Exim
2700 NULL dereference and crash.
2702 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2704 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2705 Bug detection, analysis and fix by Samuel Thibault.
2706 Bugzilla 1331, Debian bug #698092.
2708 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2710 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2711 Server implementation by Todd Lyons, client by JH.
2712 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2713 config variable "prdr_enable" controls whether the server
2714 advertises the facility. If the client requests PRDR a new
2715 acl_data_smtp_prdr ACL is called once for each recipient, after
2716 the body content is received and before the acl_smtp_data ACL.
2717 The client is controlled by both of: a hosts_try_prdr option
2718 on the smtp transport, and the server advertisement.
2719 Default client logging of deliveries and rejections involving
2720 PRDR are flagged with the string "PRDR".
2722 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2723 fclose(). Diagnosis by Todd Lyons.
2725 PP/17 Update configure.default to handle IPv6 localhost better.
2726 Patch by Alain Williams (plus minor tweaks).
2729 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2730 This is now consistent with GnuTLS, and is now documented: the
2731 previous undocumented portable approach to treating the option as
2732 unset was to force an expansion failure. That still works, and
2733 an empty string is now equivalent.
2735 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2736 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2737 not performing validation itself.
2739 PP/20 Added force_command boolean option to pipe transport.
2740 Patch from Nick Koston, of cPanel Inc.
2742 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2745 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2747 PP/21 Fix eximon continuous updating with timestamped log-files.
2748 Broken in a format-string cleanup in 4.80, missed when I repaired the
2749 other false fix of the same issue.
2750 Report and fix from Heiko Schlichting.
2753 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2754 Report from Prashanth Katuri.
2756 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2757 It's SecureTransport, so affects any MacOS clients which use the
2758 system-integrated TLS libraries, including email clients.
2760 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2761 using a MIME ACL for non-SMTP local injection.
2762 Report and assistance in diagnosis by Warren Baker.
2764 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2766 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2768 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2769 last id that failed so it may be referenced in subsequent ACL's.
2771 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2774 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2775 ldap library initialization, allowing self-signed CA's to be
2776 used. Also properly sets require_cert option later in code by
2777 using NULL (global ldap config) instead of ldap handle (per
2778 session). Bug diagnosis and testing by alxgomz.
2780 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2781 the src/util/ subdirectory.
2783 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2784 renamed to Transport Post Delivery Action by Jeremy Harris, as
2787 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2788 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2789 redis_servers = needs to be configured which will be used by the redis
2790 lookup. Patch from Warren Baker, of The Packet Hub.
2792 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2794 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2795 hostname or reverse DNS when processing a host list. Used suggestions
2796 from multiple comments on this bug.
2798 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2800 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2801 Missed a few lines, added it to make the runtest require no keyboard
2804 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2805 contains upper case chars. Make router use caseful_local_part.
2807 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2808 support when GnuTLS has been built with p11-kit.
2814 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2816 This, or similar/improved, will also be change PP/11 of 4.82.
2822 PP/01 Handle short writes when writing local log-files.
2823 In practice, only affects FreeBSD (8 onwards).
2824 Bugzilla 1053, with thanks to Dmitry Isaikin.
2826 NM/01 Bugzilla 949 - Documentation tweak
2828 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2831 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2833 PP/02 Implemented gsasl authenticator.
2835 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2837 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2838 `pkg-config foo` for cflags/libs.
2840 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2841 with rest of GSASL and with heimdal_gssapi.
2843 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2844 `pkg-config foo` for cflags/libs for the TLS implementation.
2846 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2847 properties get this fed in as external SSF. A number of robustness
2848 and debugging improvements to the cyrus_sasl authenticator.
2850 PP/08 cyrus_sasl server now expands the server_realm option.
2852 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2853 Patch by Jeremy Harris.
2855 PP/10 Added dbmjz lookup type.
2857 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2859 PP/12 MAIL args handles TAB as well as SP, for better interop with
2860 non-compliant senders.
2861 Analysis and variant patch by Todd Lyons.
2863 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2864 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2865 Patch from Dirk Mueller <dmueller@suse.com>
2867 PP/13 tls_peerdn now print-escaped for spool files.
2868 Observed some $tls_peerdn in wild which contained \n, which resulted
2869 in spool file corruption.
2871 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2872 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2873 or write after TLS renegotiation, which otherwise led to messages
2876 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2877 as a tracking header (ie: a signed header comes before the signature).
2878 Patch from Wolfgang Breyha.
2880 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2881 comma-sep list; embedded commas doubled.
2883 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2885 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2887 Report and patch from Dmitry Banschikov.
2889 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2890 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2891 failures. We appear to now support TLS1.1+ with Exim.
2893 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2894 lets Exim select keys and certificates based upon TLS SNI from client.
2895 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2896 before an outbound SMTP session. New log_selector, +tls_sni.
2898 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2899 NULL dereference. Report and patch from Alun Jones.
2901 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2902 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2903 Not seeing resolver debug output on NetBSD, but suspect this is a
2904 resolver implementation change.
2906 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2907 Left warnings. Added "eximon gdb" invocation mode.
2909 PP/21 Defaulting "accept_8bitmime" to true, not false.
2911 PP/22 Added -bw for inetd wait mode support.
2913 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2914 locate the relevant includes and libraries. Made this the default.
2916 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2917 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2919 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2920 This may cause build issues on older platforms.
2922 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2923 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2924 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2925 Added SNI support via GnuTLS too.
2926 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2928 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2930 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2931 Applied second patch from Janne, implementing suggestion to default
2932 multiple-strings-in-record handling to match SPF spec.
2934 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2936 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2937 read-only, out of scope).
2938 Patch from Wolfgang Breyha, report from Stuart Northfield.
2940 PP/29 Fix three issues highlighted by clang analyser static analysis.
2941 Only crash-plausible issue would require the Cambridge-specific
2942 iplookup router and a misconfiguration.
2943 Report from Marcin Mirosław.
2945 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
2947 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
2948 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
2949 As part of this, removing so much warning spew let me fix some minor
2950 real issues in debug logging.
2952 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
2953 assignment on my part. Fixed.
2955 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
2956 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
2957 Janne Snabb (who went above and beyond: thank you).
2959 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
2960 string otherwise requires a connection and a bunch more work and it's
2961 relatively easy to get wrong. Should also expose TLS library linkage
2964 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
2965 64-bit ${eval} (JH/03).
2967 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
2968 GNU libc to support some of the 64-bit stuff, should not lead to
2969 conflicts. Defined before os.h is pulled in, so if a given platform
2970 needs to override this, it can.
2972 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
2973 protection layer was required, which is not implemented.
2974 Bugzilla 1254, patch from Wolfgang Breyha.
2976 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
2977 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
2978 tls_dhparam take prime identifiers. Also unbreak combination of
2979 OpenSSL+DH_params+TLSSNI.
2981 PP/39 Disable SSLv2 by default in OpenSSL support.
2987 PP/01 Solaris build fix for Oracle's LDAP libraries.
2988 Bugzilla 1109, patch from Stephen Usher.
2990 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
2992 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
2995 TF/02 Fix a couple more cases where we did not log the error message
2996 when unlink() failed. See also change 4.74-TF/03.
2998 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
2999 lock up or crash if it happened to be inside a call to libc when it
3000 got a SIGUSR1 from exiwhat.
3002 The SIGUSR1 handler appends the current process status to the process
3003 log which is later printed by exiwhat. It used to use the general
3004 purpose logging code to do this, but several functions it calls are
3005 not safe for signals.
3007 The new output code in the SIGUSR1 handler is specific to the process
3008 log, and simple enough that it's easy to inspect for signal safety.
3009 Removing some special cases also simplifies the general logging code.
3010 Removing the spurious timestamps from the process log simplifies
3013 TF/04 Improved ratelimit ACL condition.
3015 The /noupdate option has been deprecated in favour of /readonly which
3016 has clearer semantics. The /leaky, /strict, and /readonly update modes
3017 are mutually exclusive. The update mode is no longer included in the
3018 database key; it just determines when the database is updated. (This
3019 means that when you upgrade Exim will forget old rate measurements.)
3021 Exim now checks that the per_* options are used with an update mode that
3022 makes sense for the current ACL. For example, when Exim is processing a
3023 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3024 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3025 must specify per_mail/readonly. If you omit the update mode it defaults to
3026 /leaky where that makes sense (as before) or /readonly where required.
3028 The /noupdate option is now undocumented but still supported for
3029 backwards compatibility. It is equivalent to /readonly except that in
3030 ACLs where /readonly is required you may specify /leaky/noupdate or
3031 /strict/noupdate which are treated the same as /readonly.
3033 A useful new feature is the /count= option. This is a generalization
3034 of the per_byte option, so that you can measure the throughput of other
3035 aggregate values. For example, the per_byte option is now equivalent
3036 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3038 The per_rcpt option has been generalized using the /count= mechanism
3039 (though it's more complicated than the per_byte equivalence). When it is
3040 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3041 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3042 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3043 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3044 Note that using per_rcpt with a non-readonly update mode in more than
3045 one ACL will cause the recipients to be double-counted. (The per_mail
3046 and per_byte options don't have this problem.)
3048 The handling of very low rates has changed slightly. If the computed rate
3049 is less than the event's count (usually one) then this event is the first
3050 after a long gap. In this case the rate is set to the same as this event's
3051 count, so that the first message of a spam run is counted properly.
3053 The major new feature is a mechanism for counting the rate of unique
3054 events. The new per_addr option counts the number of different
3055 recipients that someone has sent messages to in the last time period. It
3056 behaves like per_rcpt if all the recipient addresses are different, but
3057 duplicate recipient addresses do not increase the measured rate. Like
3058 the /count= option this is a general mechanism, so the per_addr option
3059 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3060 example, measure the rate that a client uses different sender addresses
3061 with the options per_mail/unique=$sender_address. There are further
3062 details in the main documentation.
3064 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3066 TF/06 Removed a few PCRE remnants.
3068 TF/07 Automatically extract Exim's version number from tags in the git
3069 repository when doing development or release builds.
3071 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3072 Bugzilla 879. Patch from Paul Fisher.
3074 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3075 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3078 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3080 PP/05 Variable $av_failed, true if the AV scanner deferred.
3081 Bugzilla 1078. Patch from John Horne.
3083 PP/06 Stop make process more reliably on build failure.
3084 Bugzilla 1087. Patch from Heiko Schlittermann.
3086 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3087 Bugzilla 1089. Patch from Heiko Schlittermann.
3089 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3090 Bugzilla 1131. Patch from Holger Weiß.
3092 PP/09 Handle IPv6 addresses with SPF.
3093 Bugzilla 860. Patch from Wolfgang Breyha.
3095 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3097 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3100 PP/11 match_* no longer expand right-hand-side by default.
3101 New compile-time build option, EXPAND_LISTMATCH_RHS.
3102 New expansion conditions, "inlist", "inlisti".
3104 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3106 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3108 PP/14 fix log_write() format string regression from TF/03.
3109 Bugzilla 1152. Patch from Dmitry Isaikin.
3115 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3117 PP/02 Harmonised TLS library version reporting; only show if debugging.
3118 Layout now matches that introduced for other libraries in 4.74 PP/03.
3120 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3122 PP/04 New "dns_use_edns0" global option.
3124 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3127 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3128 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3130 TK/01 Updated PolarSSL code to 0.14.2.
3131 Bugzilla 1097. Patch from Andreas Metzler.
3133 PP/07 Catch divide-by-zero in ${eval:...}.
3134 Fixes bugzilla 1102.
3136 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3139 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3140 format-string attack -- SECURITY: remote arbitrary code execution.
3142 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3143 time unintentionally subject to list matching rules, letting the header
3144 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3145 arbitrary string expansion). This allowed for information disclosure.
3147 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3148 INT_MIN/-1 -- value coerced to INT_MAX.
3154 NM/01 Workaround for PCRE version dependency in version reporting
3157 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3158 This fixes portability to compilers other than gcc, notably
3159 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3161 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3162 makefiles for portability to HP-UX and POSIX correctness.
3164 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3165 Also via indented variable definition in the Makefile.
3166 (Debugging by Oliver Heesakkers).
3168 PP/02 Restore caching of spamd results with expanded spamd_address.
3169 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3171 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3172 Improves build reliability. Fix from: Frank Elsner
3174 NM/02 Fix wide character breakage in the rfc2047 coding
3175 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3177 NM/03 Allow underscore in dnslist lookups
3178 Fixes bug 1026. Patch from Graeme Fowler
3180 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3181 Code patches from Adam Ciarcinski of NetBSD.
3183 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3186 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3187 is logged, to avoid truncation. Patch from John Horne.
3189 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3190 Patch from Jakob Hirsch.
3192 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3193 SQL string expansion failure details.
3194 Patch from Andrey Oktyabrski.
3196 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3197 Patch from Simon Arlott.
3199 PP/09 New lookups functionality failed to compile on old gcc which rejects
3200 extern declarations in function scope.
3201 Patch from Oliver Fleischmann
3203 PP/10 Use sig_atomic_t for flags set from signal handlers.
3204 Check getgroups() return and improve debugging.
3205 Fixed developed for diagnosis in bug 927 (which turned out to be
3208 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3209 Patch from Mark Zealey.
3211 PP/12 Bugzilla 1056: Improved spamd server selection.
3212 Patch from Mark Zealey.
3214 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3215 Based on patch from Heiko Schlittermann.
3217 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3218 Patch from Uwe Doering, sign-off by Michael Haardt.
3220 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3221 variable declaration deep within a block. Bug and patch from
3224 PP/15 lookups-Makefile IRIX compatibility coercion.
3226 PP/16 Make DISABLE_DKIM build knob functional.
3228 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3229 Patch by Simon Arlott
3231 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3232 variable argument macros. Our copy now differs from upstream.
3238 TF/01 Failure to get a lock on a hints database can have serious
3239 consequences so log it to the panic log.
3241 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3242 controlled using the smtp_confirmation log selector.
3244 TF/03 Include the error message when we fail to unlink a spool file.
3246 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3247 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3248 for maintaining out-of-tree patches for some time.
3250 PP/01 Bugzilla 139: Documentation and portability issues.
3251 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3252 Handle per-OS dynamic-module compilation flags.
3254 PP/02 Let /dev/null have normal permissions.
3255 The 4.73 fixes were a little too stringent and complained about the
3256 permissions on /dev/null. Exempt it from some checks.
3257 Reported by Andreas M. Kirchwitz.
3259 PP/03 Report version information for many libraries, including
3260 Exim version information for dynamically loaded libraries. Created
3261 version.h, now support a version extension string for distributors
3262 who patch heavily. Dynamic module ABI change.
3264 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3265 privilege escalation vulnerability whereby the Exim run-time user
3266 can cause root to append content of the attacker's choosing to
3269 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3272 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3273 If dropping privileges for untrusted macros, we disabled normal logging
3274 on the basis that it would fail; for the Exim run-time user, this is not
3275 the case, and it resulted in successful deliveries going unlogged.
3276 Fixed. Reported by Andreas Metzler.
3282 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3283 only prepend for the Resent-* case. Fixes regression introduced in
3284 Exim 4.70 by NM/22 for Bugzilla 607.
3286 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3287 increasing numbers of administrators be bitten by this.
3289 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3291 PP/03 Bugzilla 994: added openssl_options main configuration option.
3293 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3295 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3297 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3299 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3300 without a peer certificate, leading to a segfault because of an
3301 assumption that peers always have certificates. Be a little more
3302 paranoid. Problem reported by Martin Tscholak.
3304 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3305 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3306 NB: ClamAV planning to remove STREAM in "middle of 2010".
3307 CL also introduces -bmalware, various -d+acl logging additions and
3308 more caution in buffer sizes.
3310 PP/09 Implemented reverse_ip expansion operator.
3312 PP/10 Bugzilla 937: provide a "debug" ACL control.
3314 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3316 PP/12 Bugzilla 973: Implement --version.
3318 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3320 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3322 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3324 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3325 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3326 ignore trailing whitespace.
3328 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3330 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3333 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3334 Notification from Dr Andrew Aitchison.
3336 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3337 ExtendedDetectionInfo response format.
3338 Notification from John Horne.
3340 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3343 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3344 XSL and documented dependency on system catalogs, with examples of how
3347 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3350 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3351 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3352 configuration file which is writeable by the Exim user or group.
3354 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3355 of configuration files to cover files specified with the -C option if
3356 they are going to be used with root privileges, not just the default
3359 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3360 option (effectively making it always true).
3362 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3363 files to be used while preserving root privileges.
3365 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3366 that rogue child processes cannot use them.
3368 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3369 run-time user, instead of root.
3371 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3372 Exim run-time user without dropping privileges.
3374 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3375 result string, instead of calling string_vformat() twice with the same
3378 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3379 for other users. Others should always drop root privileges if they use
3380 -C on the command line, even for a whitelisted configure file.
3382 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3384 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3390 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3391 $data_path, and $header_path variables; fixed documentation bugs and
3394 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3395 exipick to access non-standard spools, including the "frozen" queue
3398 NM/01 Bugzilla 965: Support mysql stored procedures.
3399 Patch from Alain Williams
3401 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3403 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3404 Patch from Andreas Metzler
3406 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3407 Patch from Kirill Miazine
3409 NM/05 Bugzilla 671: Added umask to procmail example.
3411 JJ/03 installed exipick 20100323.0, fixing doc bug
3413 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3414 directory. Notification and patch from Dan Rosenberg.
3416 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3418 TK/02 Improve log output when DKIM signing operation fails.
3420 MH/01 Treat the transport option dkim_domain as a colon separated
3421 list, not as a single string, and sign the message with each element,
3422 omitting multiple occurences of the same signer.
3424 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3425 Bugzilla 985, 986. Patch by Simon Arlott
3427 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3428 Patch by Simon Arlott
3430 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3431 MBX locking. Notification from Dan Rosenberg.
3437 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3439 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3441 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3443 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3445 NM/04 Bugzilla 915: Flag broken perl installation during build.
3451 TK/01 Added patch by Johannes Berg that expands the main option
3452 "spamd_address" if it starts with a dollar sign.
3454 TK/02 Write list of recipients to X-Envelope-Sender header when building
3455 the mbox-format spool file for content scanning (suggested by Jakob
3458 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3459 (http://www.dcc-servers.net/) support via dccifd. Activated by
3460 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3462 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3463 by Mark Daniel Reidel <mr@df.eu>.
3465 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3466 When building exim an external PCRE library is now needed -
3467 PCRE is a system library on the majority of modern systems.
3468 See entry on PCRE_LIBS in EDITME file.
3470 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3471 conversation. Added nologin parameter to request.
3472 Patch contributed by Kirill Miazine.
3474 TF/01 Do not log submission mode rewrites if they do not change the address.
3476 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3478 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3479 log files in place. Contributed by Roberto Lima.
3481 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3483 TF/03 Bugzilla 615: When checking the local_parts router precondition
3484 after a local_part_suffix or local_part_prefix option, Exim now
3485 does not use the address's named list lookup cache, since this
3486 contains cached lookups for the whole local part.
3488 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3489 Robert Millan. Documentation is in experimental-spec.txt.
3491 TF/04 Bugzilla 668: Fix parallel build (make -j).
3493 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3495 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3496 Patch provided by Jan Srzednicki.
3498 TF/05 Leading white space used to be stripped from $spam_report which
3499 wrecked the formatting. Now it is preserved.
3501 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3502 that they are available at delivery time.
3504 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3506 TF/08 TLS error reporting now respects the incoming_interface and
3507 incoming_port log selectors.
3509 TF/09 Produce a more useful error message if an SMTP transport's hosts
3510 setting expands to an empty string.
3512 NM/06 Bugzilla 744: EXPN did not work under TLS.
3513 Patch provided by Phil Pennock.
3515 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3516 Patch provided by Richard Godbee.
3518 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3519 acl_smtp_notquit, added index entry.
3521 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3522 Patch provided by Eugene Bujak.
3524 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3525 accept(). Patch provided by Maxim Dounin.
3527 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3528 Patch provided by Phil Pennock.
3530 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3532 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3533 Patch provided by Brad "anomie" Jorsch.
3535 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3536 Patch provided by Dean Brooks.
3538 TK/05 Add native DKIM support (does not depend on external libraries).
3540 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3541 Patch provided by Graeme Fowler.
3543 NM/16 Bugzilla 851: Documentation example syntax fix.
3545 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3547 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3550 NM/19 Bugzilla 745: TLS version reporting.
3551 Patch provided by Phil Pennock.
3553 NM/20 Bugzilla 167: bool: condition support.
3554 Patch provided by Phil Pennock.
3556 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3557 clients. Patch provided by Phil Pennock.
3559 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3560 Patch provided by Brad "anomie" Jorsch.
3562 NM/23 Bugzilla 687: Fix misparses in eximstats.
3563 Patch provided by Heiko Schlittermann.
3565 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3566 Patch provided by Heiko Schlittermann.
3568 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3569 plus update to original patch.
3571 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3573 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3574 Patch provided by David Brownlee.
3576 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3578 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3580 NM/30 Bugzilla 888: TLS documentation bugfixes.
3582 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3584 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3585 Unlike the original bugzilla I have changed all shell scripts in src tree.
3587 NM/33 Bugzilla 898: Transport filter timeout fix.
3588 Patch by Todd Rinaldo.
3590 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3591 Patch by Serge Demonchaux.
3593 NM/35 Bugzilla 39: Base64 decode bug fixes.
3594 Patch by Jakob Hirsch.
3596 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3598 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3600 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3602 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3608 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3609 ALT-N's libdkim that I have put here:
3610 http://duncanthrax.net/exim-experimental/
3612 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3613 were called 'true' and it seems that C99 defines that as a reserved
3614 keyword to be used with 'bool' variable types. That means you could
3615 not include C99-style headers which use bools without triggering
3616 build errors in sieve.c.
3618 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3619 as mailq or other aliases. Changed the --help handling significantly
3620 to do whats expected. exim_usage() emits usage/help information.
3622 SC/01 Added the -bylocaldomain option to eximstats.
3624 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3626 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3628 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3634 PH/01 Another patch from the Sieve maintainer.
3636 PH/02 When an IPv6 address is converted to a string for single-key lookup
3637 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3638 dots are used instead of colons so that keys in lsearch files need not
3639 contain colons. This was done some time before quoting was made available
3640 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3641 (notated using the quote facility) so as to distinguish them from IPv4
3642 keys. This meant that lookups for IP addresses in host lists did not work
3643 for iplsearch lookups.
3645 This has been fixed by arranging for IPv6 addresses to be expressed with
3646 colons if the lookup type is iplsearch. This is not incompatible, because
3647 previously such lookups could never work.
3649 The situation is now rather anomalous, since one *can* have colons in
3650 ordinary lsearch keys. However, making the change in all cases is
3651 incompatible and would probably break a number of configurations.
3653 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3656 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3657 conversion specification without a maximum field width, thereby enabling
3658 a rogue spamd server to cause a buffer overflow. While nobody in their
3659 right mind would setup Exim to query an untrusted spamd server, an
3660 attacker that gains access to a server running spamd could potentially
3661 exploit this vulnerability to run arbitrary code as the Exim user.
3663 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3664 $primary_hostname instead of what libspf2 thinks the hosts name is.
3666 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3667 a directory entry by the name of the lookup key. Previously, if a
3668 symlink pointed to a non-existing file or a file in a directory that
3669 Exim lacked permissions to read, a lookup for a key matching that
3670 symlink would fail. Now it is enough that a matching directory entry
3671 exists, symlink or not. (Bugzilla 503.)
3673 PH/03 The body_linecount and body_zerocount variables are now exported in the
3676 PH/04 Added the $dnslist_matched variable.
3678 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3679 This means they are set thereafter only if the connection becomes
3682 PH/06 Added the client_condition to authenticators so that some can be skipped
3683 by clients under certain conditions.
3685 PH/07 The error message for a badly-placed control=no_multiline_responses left
3686 "_responses" off the end of the name.
3688 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3690 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3691 (without spaces) instead of just copying the configuration text.
3693 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3695 PH/11 Added $max_received_linelength.
3697 PH/12 Added +ignore_defer and +include_defer to host lists.
3699 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3700 way in which PCRE > 7.0 is built.
3702 PH/14 Implemented queue_only_load_latch.
3704 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3705 MAIL command. The effect was to mangle the value on 64-bit systems.
3707 PH/16 Another patch from the Sieve maintainer.
3709 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3711 PH/18 If a system quota error occurred while trying to create the file for
3712 a maildir delivery, the message "Mailbox is full" was not appended to the
3713 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3714 only to a quota excession during the actual writing of the file.
3716 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3717 characters?) which causes problems in log lines. The DN values are now
3718 passed through string_printing() before being added to log lines.
3720 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3721 and InterBase are left for another time.)
3723 PH/21 Added message_body_newlines option.
3725 PH/22 Guard against possible overflow in moan_check_errorcopy().
3727 PH/23 POSIX allows open() to be a macro; guard against that.
3729 PH/24 If the recipient of an error message contained an @ in the local part
3730 (suitably quoted, of course), incorrect values were put in $domain and
3731 $local_part during the evaluation of errors_copy.
3737 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3738 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3741 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3742 issue a MAIL command.
3744 PH/02 In an ACL statement such as
3746 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3748 if a client was not listed at all, or was listed with a value other than
3749 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3750 the condition was not true (as it should be), so access was not denied.
3751 The bug was that the ! inversion was incorrectly passed on to the second
3752 item. This has been fixed.
3754 PH/03 Added additional dnslists conditions == and =& which are different from
3755 = and & when the dns lookup returns more than one IP address.
3757 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3758 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3760 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3761 FSYNC, which compiles an option called disable_fsync that allows for
3762 bypassing fsync(). The documentation is heavily laced with warnings.
3764 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3766 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3767 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3768 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3769 including adding "make clean"; (3) Added -fPIC when compiling the test
3770 dynamically loaded module, to get rid of a warning.
3772 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3773 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3774 = 0s. The bug is otherwise harmless.
3776 PH/07 There was a bug in the dovecot authenticator such that the value of
3777 $auth1 could be overwritten, and so not correctly preserved, after a
3778 successful authentication. This usually meant that the value preserved by
3779 the server_setid option was incorrect.
3781 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3783 PH/09 Installed PCRE release 7.0.
3785 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3786 run for batched SMTP input. It is now run at the start of every message
3787 in the batch. While fixing this I discovered that the process information
3788 (output by running exiwhat) was not always getting set for -bs and -bS
3789 input. This is fixed, and it now also says "batched" for BSMTP.
3791 PH/11 Added control=no_pipelining.
3793 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3794 patch, slightly modified), and move the expansion of helo_data till after
3795 the connection is made in the smtp transport (so it can use these
3798 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3800 PH/14 Added log_selector = +pid.
3802 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3804 PH/16 Add ${if forany and ${if forall.
3806 PH/17 Added dsn_from option to vary the From: line in DSNs.
3808 PH/18 Flush SMTP output before performing a callout, unless control =
3809 no_callout_flush is set.
3811 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3812 was true (the default) a successful delivery failed to delete the retry
3813 item, thus causing premature timeout of the address. The bug is now
3816 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3818 PH/21 Long custom messages for fakedefer and fakereject are now split up
3819 into multiline responses in the same way that messages for "deny" and
3820 other ACL rejections are.
3822 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3823 with slight modification.
3825 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3826 draft, changing the syntax and factoring some duplicate code".
3828 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3829 for deliveries of the second and subsequent messages over the same SMTP
3832 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3833 ${reduce, with only minor "tidies".
3835 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3837 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3838 expansion side effects.
3840 PH/27 When a message times out after an over-quota error from an Exim-imposed
3841 quota, the bounce message says "mailbox is full". This message was not
3842 being given when it was a system quota that was exceeded. It now should
3845 MH/03 Made $recipients available in local_scan(). local_scan() already has
3846 better access to the recipient list through recipients_list[], but
3847 $recipients can be useful in postmaster-provided expansion strings.
3849 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3850 in the case of a MAIL command with additional options following the
3851 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3852 were accidentally chopped off.
3854 PH/29 SMTP synchronization checks are implemented when a command is read -
3855 there is a check that no more input is waiting when there shouldn't be
3856 any. However, for some commands, a delay in an ACL can mean that it is
3857 some time before the response is written. In this time, more input might
3858 arrive, invalidly. So now there are extra checks after an ACL has run for
3859 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3860 pipelining has not been advertised.
3862 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3864 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3865 returned in the SMTP response when smtp_return_error_details was false.
3866 This has been fixed.
3868 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3869 instead of the C I/O that was originally supplied, because problems were
3870 reported on Solaris.
3872 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3873 Exim which did not show up earlier: it was assuming that a call to
3874 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3875 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3876 was a macro that became an assignment, so it seemed to work. This has
3877 changed to a proper function call with a void return, hence the compile
3878 error. Exim's code has been fixed.
3880 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3883 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3885 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3887 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3888 because it turns out to be too big or there is a timeout) there is no
3889 "Completed" line in the log. When some messages of this type were
3890 selected by exigrep, they were listed as "not completed". Others were
3891 picked up by some special patterns. I have improved the selection
3892 criteria to be more general.
3894 PH/38 The host_find_failed option in the manualroute router can now be set
3895 to "ignore", to completely ignore a host whose IP address cannot be
3896 found. If all hosts are ignored, the behaviour is controlled by the new
3897 host_all_ignored option.
3899 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3900 homing or because of multiple MX records with /mx) generated more than
3901 one IP address, and the following item turned out to be the local host,
3902 all the secondary addresses of the first item were incorrectly removed
3903 from the list, along with the local host and any following hosts (which
3904 is what is supposed to happen).
3906 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3907 whoever called Exim into the -H file. In the case of the daemon it was
3908 behaving confusingly. When first started, it used values for whoever
3909 started the daemon, but after a SIGHUP it used the Exim user (because it
3910 calls itself on a restart). I have changed the code so that it now always
3913 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3914 message are rejected with the same error (e.g. no authentication or bad
3915 sender address), and a DATA command is nevertheless sent (as can happen
3916 with PIPELINING or a stupid MUA), the error message that was given to the
3917 RCPT commands is included in the rejection of the DATA command. This is
3918 intended to be helpful for MUAs that show only the final error to their
3921 PH/42 Another patch from the Sieve maintainer.
3923 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3924 Eximstats - Fixed some broken HTML links and added missing column headers
3926 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3927 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3929 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3931 PH/43 Yet another patch from the Sieve maintainer.
3933 PH/44 I found a way to check for a TCP/IP connection going away before sending
3934 the response to the final '.' that terminates a message, but only in the
3935 case where the client has not sent further data following the '.'
3936 (unfortunately, this is allowed). However, in many cases there won't be
3937 any further data because there won't be any more messages to send. A call
3938 to select() can be used: if it shows that the input is "ready", there is
3939 either input waiting, or the socket has been closed. An attempt to read
3940 the next input character can distinguish the two cases. Previously, Exim
3941 would have sent an OK response which the client would never have see.
3942 This could lead to message repetition. This fix should cure that, at
3943 least in a lot of common cases.
3945 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
3946 advertised in response to EHLO.
3952 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
3953 fixed by 4.65/MH/01 (is this a record?) are fixed:
3955 (i) An empty string was always treated as zero by the numeric comparison
3956 operators. This behaviour has been restored.
3958 (ii) It is documented that the numeric comparison operators always treat
3959 their arguments as decimal numbers. This was broken in that numbers
3960 starting with 0 were being interpreted as octal.
3962 While fixing these problems I realized that there was another issue that
3963 hadn't been noticed. Values of message_size_limit (both the global option
3964 and the transport option) were treated as octal if they started with 0.
3965 The documentation was vague. These values are now always treated as
3966 decimal, and I will make that clear in the documentation.
3972 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
3973 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
3976 MH/01 Don't check that the operands of numeric comparison operators are
3977 integers when their expansion is in "skipping" mode (fixes bug
3978 introduced by 4.64-PH/07).
3980 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
3981 child addresses, Exim now panics and dies. Previously, because the count
3982 is held in a short int, deliveries were likely to be lost. As such a
3983 large number of recipients for a single message is ridiculous
3984 (performance will be very, very poor), I have chosen to impose a limit
3985 rather than extend the field.
3991 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
3992 leftover -K file (the existence of which was triggered by #402).
3993 While we were at it, introduced process PID as part of the -K
3994 filename. This should rule out race conditions when creating
3997 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
3998 processing considerably. Previous code took too long for large mails,
3999 triggering a timeout which in turn triggers #401.
4001 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4002 in the DK code in transports.c. sendfile() is not really portable,
4003 hence the _LINUX specificness.
4005 TF/01 In the add_headers option to the mail command in an Exim filter,
4006 there was a bug that Exim would claim a syntax error in any
4007 header after the first one which had an odd number of characters
4010 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4011 callout verification, Exim cached a "reject" for the entire domain. This
4012 is correct for most verifications, but it is not correct for a recipient
4013 verification with use_sender or use_postmaster set, because in that case
4014 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4015 case of MAIL FROM:<> rejection from other early rejections (e.g.
4016 rejection of HELO). When verifying a recipient using a non-null MAIL
4017 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4018 Whatever the result of the callout, the value of the domain cache is
4019 left unchanged (for any other kind of callout, getting as far as trying
4020 RCPT means that the domain itself is ok).
4022 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4025 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4026 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4027 session, the daemon ignores ECONNECT errors and logs others; it now
4028 ignores EPIPE as well.
4030 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4031 (quoted-printable decoding).
4033 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4034 later the small subsequent patch to fix an introduced bug.
4036 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4038 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4040 PH/08 An error is now given if message_size_limit is specified negative.
4042 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4043 to be given (somewhat) arbitrary names.
4045 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4048 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4049 miscellaneous code fixes
4051 PH/10 Added the log_reject_target ACL modifier to specify where to log
4054 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4055 hostname. This is wrong, because it relates to the incoming message (and
4056 probably the interface on which it is arriving) and not to the outgoing
4057 callout (which could be using a different interface). This has been
4058 changed to use the value of the helo_data option from the smtp transport
4059 instead - this is what is used when a message is actually being sent. If
4060 there is no remote transport (possible with a router that sets up host
4061 addresses), $smtp_active_hostname is used.
4063 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4064 tweaks were necessary in order to get it to work (see also 21 below):
4065 (a) The code assumed that strncpy() returns a negative number on buffer
4066 overflow, which isn't the case. Replaced with Exim's string_format()
4068 (b) There were several signed/unsigned issues. I just did the minimum
4069 hacking in of casts. There is scope for a larger refactoring.
4070 (c) The code used strcasecmp() which is not a standard C function.
4071 Replaced with Exim's strcmpic() function.
4072 (d) The code set only $1; it now sets $auth1 as well.
4073 (e) A simple test gave the error "authentication client didn't specify
4074 service in request". It would seem that Dovecot has changed its
4075 interface. Fortunately there's a specification; I followed it and
4076 changed what the client sends and it appears to be working now.
4078 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4081 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4082 address A is aliased to B and C, where B exists and C does not. Without
4083 -v the output is "A verified" because verification stops after a
4084 successful redirection if more than one address is generated. However,
4085 with -v the child addresses are also verified. Exim was outputting "A
4086 failed to verify" and then showing the successful verification for C,
4087 with its parentage. It now outputs "B failed to verify", showing B's
4088 parentage before showing the successful verification of C.
4090 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4091 look up a TXT record in a specific list after matching in a combined
4094 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4095 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4096 they consult the DNS. I had assumed they would set it the way they
4097 wanted; and indeed my experiments on Linux seem to show that in some
4098 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4099 To be on the safe side, however, I have now made the interface to
4100 host_find_byname() similar to host_find_bydns(), with an argument
4101 containing the DNS resolver options. The host_find_byname() function now
4102 sets these options at its start, just as host_find_bydns() does. The smtp
4103 transport options dns_qualify_single and dns_search_parents are passed to
4104 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4105 of host_find_byname() use the default settings of RES_DEFNAMES
4106 (qualify_single) but not RES_DNSRCH (search_parents).
4108 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4109 spool_read_header() do less string testing, by means of a preliminary
4110 switch on the second character of optional "-foo" lines. (This is
4111 overdue, caused by the large number of possibilities that now exist.
4112 Originally there were few.) While I was there, I also converted the
4113 str(n)cmp tests so they don't re-test the leading "-" and the first
4114 character, in the hope this might squeeze out yet more improvement.
4116 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4117 flag allowing group syntax was set by the header_syntax check but not
4118 turned off, possible causing trouble later; (2) The flag was not being
4119 set at all for the header_verify test, causing "group"-style headers to
4120 be rejected. I have now set it in this case, and also caused header_
4121 verify to ignore an empty address taken from a group. While doing this, I
4122 came across some other cases where the code for allowing group syntax
4123 while scanning a header line wasn't quite right (mostly, not resetting
4124 the flag correctly in the right place). These bugs could have caused
4125 trouble for malformed header lines. I hope it is now all correct.
4127 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4128 with the "reply" argument non-NULL. The code, however (which originally
4129 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4130 but it didn't always do it. This confused somebody who was copying the
4131 code for some other use. I have removed all the tests.
4133 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4134 feature that was used to support insecure browsers during the U.S. crypto
4135 embargo. It requires special client support, and Exim is probably the
4136 only MTA that supported it -- and would never use it because real RSA is
4137 always available. This code has been removed, because it had the bad
4138 effect of slowing Exim down by computing (never used) parameters for the
4139 RSA_EXPORT functionality.
4141 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4142 authenticator to fail if there's a tab character in the incoming data
4143 (there should never be unless someone is messing about, as it's supposed
4144 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4145 if the connection is using TLS or if the remote IP is the same as the
4146 local IP, and the "valid-client-cert option" if a client certificate has
4149 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4150 authenticators. This can be used for authorization after authentication
4151 succeeds. (In the case of plaintext, it servers for both authentication
4154 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4155 if any retry times were supplied.
4157 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4158 connection, where there is no client IP address to check. In this
4159 situation, the verify now always succeeds.
4161 PH/25 Applied John Jetmore's -Mset patch.
4163 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4165 PH/27 In a string expansion for a processed (not raw) header when multiple
4166 headers of the same name were present, leading whitespace was being
4167 removed from all of them, but trailing whitespace was being removed only
4168 from the last one. Now trailing whitespace is removed from each header
4169 before concatenation. Completely empty headers in a concatenation (as
4170 before) are ignored.
4172 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4173 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4175 PH/29 [Removed. This was a change that I later backed out, and forgot to
4176 correct the ChangeLog entry (that I had efficiently created) before
4177 committing the later change.]
4179 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4180 address errors (4xx response to RCPT) over the same connection as other
4181 messages routed to the same hosts. Such deliveries are always "forced",
4182 so retry times are not inspected. This resulted in far too many retries
4183 for the affected addresses. The effect occurred only when there were more
4184 hosts than the hosts_max_try setting in the smtp transport when it had
4185 the 4xx errors. Those hosts that it had tried were not added to the list
4186 of hosts for which the message was waiting, so if all were tried, there
4187 was no problem. Two fixes have been applied:
4189 (i) If there are any address or message errors in an SMTP delivery, none
4190 of the hosts (tried or untried) are now added to the list of hosts
4191 for which the message is waiting, so the message should not be a
4192 candidate for sending over the same connection that was used for a
4193 successful delivery of some other message. This seems entirely
4194 reasonable: after all the message is NOT "waiting for some host".
4195 This is so "obvious" that I'm not sure why it wasn't done
4196 previously. Hope I haven't missed anything, but it can't do any
4197 harm, as the worst effect is to miss an optimization.
4199 (ii) If, despite (i), such a delivery is accidentally attempted, the
4200 routing retry time is respected, so at least it doesn't keep
4201 hammering the server.
4203 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4204 in ${readsocket because some servers need this prod.
4206 PH/32 Added some extra debug output when updating a wait-xxx database.
4208 PH/33 The hint "could be header name not terminated by colon", which has been
4209 given for certain expansion errors for a long time, was not being given
4210 for the ${if def:h_colon_omitted{... case.
4212 PH/34 The spec says: "With one important exception, whenever a domain list is
4213 being scanned, $domain contains the subject domain." There was at least
4214 one case where this was not true.
4216 PH/35 The error "getsockname() failed: connection reset by peer" was being
4217 written to the panic log as well as the main log, but it isn't really
4218 panic-worthy as it just means the connection died rather early on. I have
4219 removed the panic log writing for the ECONNRESET error when getsockname()
4222 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4223 runs only) independently of the message's sender address. This meant
4224 that, if the 4xx error was in fact related to the sender, a different
4225 message to the same recipient with a different sender could confuse
4226 things. In particular, this can happen when sending to a greylisting
4227 server, but other circumstances could also provoke similar problems.
4228 I have changed the default so that the retry time for these errors is now
4229 based a combination of the sender and recipient addresses. This change
4230 can be overridden by setting address_retry_include_sender=false in the
4233 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4234 remote server are returned as part of bounce messages. This was not
4235 happening for LMTP over a pipe (the lmtp transport), but now it is the
4236 same for both kinds of LMTP.
4238 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4239 in header lines that were in fact CNAMEs. This is no longer the case.
4241 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4242 and queue runs started by the daemon processed all messages. This has
4243 been fixed so that -R and -S can now usefully be given with -q<time>.
4245 PH/40 Import PCRE release 6.7 (fixes some bugs).
4247 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4249 PH/42 Give an error if -q is specified more than once.
4251 PH/43 Renamed the variables $interface_address and $interface_port as
4252 $received_ip_address and $received_port, to make it clear that these
4253 values apply to message reception, and not to the outgoing interface when
4254 a message is delivered. (The old names remain recognized, of course.)
4256 PH/44 There was no timeout on the connect() call when using a Unix domain
4257 socket in the ${readsocket expansion. There now is.
4259 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4260 be meaningful with "accept".
4262 SC/01 Eximstats V1.43
4263 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4265 SC/02 Eximstats V1.44
4266 Use a glob alias rather than an array ref in the generated
4267 parser. This improves both readability and performance.
4269 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4270 Collect SpamAssassin and rejection statistics.
4271 Don't display local sender or destination tables unless
4272 there is data to show.
4273 Added average volumes into the top table text output.
4275 SC/04 Eximstats V1.46
4276 Collect data on the number of addresses (recipients)
4277 as well as the number of messages.
4279 SC/05 Eximstats V1.47
4280 Added 'Message too big' to the list of mail rejection
4281 reasons (thanks to Marco Gaiarin).
4283 SC/06 Eximstats V1.48
4284 Mainlog lines which have GMT offsets and are too short to
4285 have a flag are now skipped.
4287 SC/07 Eximstats V1.49 (Alain Williams)
4288 Added the -emptyok flag.
4290 SC/08 Eximstats V1.50
4291 Fixes for obtaining the IP address from reject messages.
4293 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4294 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4295 whitespace changes from 4.64-PH/27
4297 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4300 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4301 are found, allow negative numbers in numeric criteria)
4303 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4305 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4306 to match changes made in 4.64-PH/43
4308 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4310 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4311 "domain" is actually an IP address, and force a failure. This locks out
4312 those revolvers/nameservers that support "A-for-A" lookups, in
4313 contravention of the specifications.
4315 PH/48 When a host name was looked up from an IP address, and the subsequent
4316 forward lookup of the name timed out, the host name was left in
4317 $sender_host_name, contrary to the specification.
4319 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4320 restricted to single-key lookups, Exim was not diagnosing an error if
4321 * or *@ was used with a query-style lookup.
4323 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4325 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4326 long ago, but noone interested enough thought of it. Let's just say that
4327 the "1.1" means that there are some new functions that weren't there at
4328 some point in the past.
4330 PH/51 Error processing for expansion failure of helo_data from an smtp
4331 transport during callout processing was broken.
4333 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4334 tested/used via the -bh/-bhc/-bs options.
4336 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4337 bug, fixed in subsequent PCRE releases).
4339 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4340 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4342 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4348 SC/01 Use a glob alias rather than an array ref in eximstats generated
4349 parser. This improves both readability and performance.
4351 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4352 Don't display local sender or destination tables in eximstats unless
4353 there is data to show.
4354 Added average volumes into the eximstats top table text output.
4356 SC/03 Collect data on the number of addresses (recipients) as well
4357 as the number of messages in eximstats.
4359 TF/01 Correct an error in the documentation for the redirect router. Exim
4360 does (usually) call initgroups() when daemonizing.
4362 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4363 with consistent privilege compared to when running as a daemon.
4365 TF/03 Note in the spec that $authenticated_id is not set for local
4366 submissions from trusted users.
4368 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4369 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4371 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4372 by adding some example configuration directives to the default
4373 configuration file. A little bit of work is required to uncomment the
4374 directives and define how usernames and passwords are checked, but
4375 there is now a framework to start from.
4377 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4378 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4379 without this. I don't know how relevant this is to other LDAP libraries.
4381 PH/02 Add the verb name to the "unknown ACL verb" error.
4383 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4385 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4387 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4388 directory not expanded when it should be if an expanded home directory
4389 was set for the address (which is overridden by the transport).
4391 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4394 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4395 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4396 because it is too late at that time, and has no effect.
4398 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4399 security issue with \' (bugzilla #107). I could not use the
4400 PQescapeStringConn() function, because it needs a PGconn value as one of
4403 PH/08 When testing addresses using -bt, indicate those final addresses that
4404 are duplicates that would not cause an additional delivery. At least one
4405 person was confused, thinking that -bt output corresponded to deliveries.
4406 (Suppressing duplicates isn't a good idea as you lose the information
4407 about possibly different redirections that led to the duplicates.)
4409 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4410 systems where poll() doesn't work, in particular OS X.
4412 PH/10 Added more information to debugging output for retry time not reached.
4414 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4415 operations in malware.c.
4417 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4420 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4421 syslog_duplication set false, log lines that would normally be written
4422 both the the main log and to the reject log were not written to syslog at
4425 PH/14 In the default configuration, change the use of "message" in ACL warn
4426 statements to "add_header".
4428 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4429 not followed by a command (e.g. "seen endif").
4431 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4432 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4435 PH/17 Added extra conditions to the default value of delay_warning_condition
4439 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4440 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4441 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4444 The Auto-Submitted: and various List- headers are standardised, whereas I
4445 don't think Precedence: ever was.
4447 PH/18 Refactored debugging code in route_finduser() to show more information,
4448 in particular, the error code if getpwnam() issues one.
4450 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4451 This is apparently needed in addition to the PH/07 change above to avoid
4452 any possible encoding problems.
4454 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4455 but not after initializing Perl.
4457 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4458 output them only if debugging. By default they are written stderr,
4459 apparently, which is not desirable.
4461 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4464 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4467 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4469 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4470 authenticated or an ident call has been made. Suppress the default
4471 values for $authenticated_id and $authenticated_sender (but permit -oMai
4472 and -oMas) when testing with -bh.
4474 PH/24 Re-jigged the order of the tests in the default configuration so that the
4475 tests for valid domains and recipients precede the DNS black list and CSA
4476 tests, on the grounds that those ones are more expensive.
4478 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4479 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4480 This bug exists in every version of Exim that I still have, right back to
4483 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4484 However, an attempt to turn on case-sensitivity in a regex key by
4485 including (?-i) didn't work because the subject string was already
4486 lowercased, and the effects were non-intuitive. It turns out that a
4487 one-line patch can be used to allow (?-i) to work as expected.
4493 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4494 other effects) broke the use of negated acl sub-conditions.
4496 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4499 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4500 "Deny" causes Exim to reject the incoming connection with a 554 error.
4501 Unfortunately, if there is a major crisis, such as a disk failure,
4502 tcp-wrappers gives "deny", whereas what one would like would be some
4503 kind of temporary error. A kludge has been added to help with this.
4504 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4505 554 error is used if errno is still zero or contains ENOENT (which occurs
4506 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4509 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4511 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4512 errors. Otherwise a message that provokes a temporary error (when other
4513 messages do not) can cause a whole host to time out.
4515 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4516 addresses were routed directly to files or pipes from a redirect router.
4517 File deliveries just didn't batch; pipe deliveries might have suffered
4520 PH/06 A failure to get a lock for a hints database would erroneously always say
4521 "Failed to get write lock", even when it was really a read lock.
4523 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4524 of 0600. This has been changed to use the value of the lockfile_mode
4525 option (which defaults to 0600).
4527 PH/08 Applied small patch from the Sieve maintainer.
4529 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4530 folder from quota calculations, a direct delivery into this folder messed
4531 up the contents of the maildirsize file. This was because the regex was
4532 used only to exclude .Trash (or whatever) when the size of the mailbox
4533 was calculated. There was no check that a delivery was happening into an
4534 excluded directory. This bug has been fixed by ignoring all quota
4535 processing for deliveries into excluded directories.
4537 PH/10 Added the maildirfolder_create_regex option to appendfile.
4543 PH/01 The code for finding all the local interface addresses on a FreeBSD
4544 system running IPv6 was broken. This may well have applied to all BSD
4545 systems, as well as to others that have similar system calls. The broken
4546 code found IPv4 interfaces correctly, but gave incorrect values for the
4547 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4548 that it would not match correctly against @[] and not recognize the IPv6
4551 PH/02 The ipliteral router was not recognizing addresses of the form user@
4552 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4554 PH/03 Added disable_ipv6.
4556 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4557 decoded form, because it is most often used to construct To: headers
4558 lines in autoreplies, and the decoded form may well be syntactically
4559 invalid. However, $reply_address has leading white space removed, and all
4560 newlines turned into spaces so that the autoreply transport does not
4563 PH/05 If group was specified without a user on a router, and no group or user
4564 was specified on a transport, the group from the router was ignored.
4566 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4567 for visible compile-time settings that can be used to change these
4568 numbers, for those that want even more. Backwards compatibility with old
4569 spool files has been maintained. However, going back to a previous Exim
4570 release will lost any variables that are in spool files.
4572 PH/07 Two small changes when running in the test harness: increase delay when
4573 passing a TCP/IP connection to a new process, in case the original
4574 process has to generate a bounce, and remove special handling of
4575 127.0.0.2 (sic), which is no longer necessary.
4577 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4578 be the same on different OS.
4580 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4583 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4584 whether --show-vars was specified or not
4586 JJ/02 exipick: Added support for new ACL variable spool format introduced
4589 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4590 syntactically invalid From: or Reply-to: line, and a filter used this to
4591 generate an autoreply, and therefore failed to obtain an address for the
4592 autoreply, Exim could try to deliver to a non-existent relative file
4593 name, causing unrelated and misleading errors. What now happens is that
4594 it logs this as a hard delivery error, but does not attempt to create a
4597 PH/11 The exinext utility has a -C option for testing purposes, but although
4598 the given file was scanned by exinext itself; it wasn't being passed on
4599 when Exim was called.
4601 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4602 an end-of-file indication when reading a command response.
4604 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4605 compiled. In many other places in Exim, IPv6 addresses are always
4606 recognized, so I have changed this. It also means that IPv4 domain
4607 literals of the form [IPV4:n.n.n.n] are now always recognized.
4609 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4610 used if the router is not running as root, for example, when verifying at
4611 ACL time, or when using -bh. The debugging output from this situation was
4612 non-existent - all you got was a failure to exec. I have made two
4615 (a) Failures to set uid/gid, the current directory, or a process leader
4616 in a subprocess such as that created by queryprogram now generate
4617 suitable debugging output when -d is set.
4619 (b) The queryprogram router detects when it is not running as root,
4620 outputs suitable debugging information if -d is set, and then runs
4621 the subprocess without attempting to change uid/gid.
4623 PH/15 Minor change to Makefile for building test_host (undocumented testing
4626 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4627 additional section of a DNS packet that returns MX or SRV records.
4628 Instead, it always explicitly searches for A/AAAA records. This avoids
4629 major problems that occur when a DNS server includes only records of one
4630 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4631 fixed another bug: if SRV records were looked up and the corresponding
4632 address records were *not* found in the additional section, the port
4633 values from the SRV records were lost.
4635 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4636 using the correct key (the original address) when searching the retry
4637 rules in order to find which one to use for generating the retry hint.
4639 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4640 adding the default one. Similarly, if it contains a Reply-To: header, the
4641 errors_reply_to option, if set, is not used.
4643 PH/19 When calculating a retry time, Exim used to measure the "time since
4644 failure" by looking at the "first failed" field in the retry record. Now
4645 it does not use this if it is later than than the arrival time of the
4646 message. Instead it uses the arrival time. This makes for better
4647 behaviour in cases where some deliveries succeed, thus re-setting the
4648 "first failed" field. An example is a quota failure for a huge message
4649 when small messages continue to be delivered. Without this change, the
4650 "time since failure" will always be short, possible causing more frequent
4651 delivery attempts for the huge message than are intended.
4652 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4654 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4655 $1, $2, $3) because the numerical variables can be reset during some
4656 expansion items (e.g. "match"), thereby losing the authentication data.
4658 PH/21 Make -bV show the size of off_t variables so that the test suite can
4659 decide whether to run tests for quotas > 2G.
4661 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4662 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4663 filecount value is greater than 2G or if a quota value is greater than 2G
4664 on a system where the size of off_t is not greater than 4, a panic error
4667 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4668 never match. The debug and -bh output now contains an explicit error
4669 message indicating a malformed IPv4 address or mask.
4671 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4672 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4673 PH/23 above applies.
4675 PH/25 Do not write to syslog when running in the test harness. The only
4676 occasion when this arises is a failure to open the main or panic logs
4677 (for which there is an explicit test).
4679 PH/26 Added the /no_tell option to "control=freeze".
4681 PH/27 If a host name lookup failed very early in a connection, for example, if
4682 the IP address matched host_lookup and the reverse lookup yielded a name
4683 that did not have a forward lookup, an error message of the form "no IP
4684 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4685 could be logged. Now it outputs the IP address instead of "NULL".
4687 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4688 allows the sender and the authenticated sender to be set when
4689 submitting a message from within Exim. Since child_open_exim() is
4690 documented for local_scan(), the new function should be too.
4692 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4693 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4694 results in an empty string is now treated as unset.
4696 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4698 PH/31 Added sender_verify_fail logging option.
4700 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4701 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4702 tidied the source and removed it altogether.
4704 PH/33 When a queue run was abandoned because the load average was too high, a
4705 log line was always written; now it is written only if the queue_run log
4706 selector is set. In addition, the log line for abandonment now contains
4707 information about the queue run such as the pid. This is always present
4708 in "start" and "stop" lines but was omitted from the "abandon" line.
4710 PH/34 Omit spaces between a header name and the colon in the error message that
4711 is given when verify = headers_syntax fails (if there are lots of them,
4712 the message gets confusing).
4714 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4715 names, as there are now some PTR records that contain slashes. This check
4716 is only to protect against broken name servers that fall over on strange
4717 characters, so the fact that it applies to all lookups doesn't matter.
4719 PH/36 Now that the new test suite is complete, we can remove some of the
4720 special code in Exim that was needed for the old test suite. For example,
4721 sorting DNS records because real resolvers return them in an arbitrary
4722 order. The new test suite's fake resolver always returns records in the
4725 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4726 bounces) except when queue_only is set, to avoid logging races between
4727 the different processes.
4729 PH/38 Panic-die if .include specifies a non-absolute path.
4731 PH/39 A tweak to the "H" retry rule from its user.
4733 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4734 a label. They prevented compilation on older perls.
4736 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4737 a warning to be raised on newish perls.
4739 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4740 on queue. Changes to match documented behaviour of showing count of
4741 messages matching specified criteria.
4743 PH/40 Changed the default ident timeout from 30s to 5s.
4745 PH/41 Added support for the use of login_cap features, on those BSD systems
4746 that have them, for controlling the resources used by pipe deliveries.
4748 PH/42 The content-scanning code uses fopen() to create files in which to put
4749 message data. Previously it was not paying any attention to the mode of
4750 the files. Exim runs with umask(0) because the rest of the code creates
4751 files with open(), and sets the required mode explicitly. Thus, these
4752 files were ending up world-writeable. This was not a big issue, because,
4753 being within the spool directory, they were not world-accessible. I have
4754 created a function called modefopen, which takes an additional mode
4755 argument. It sets umask(777), creates the file, chmods it to the required
4756 mode, then resets the umask. All the relevant calls to fopen() in the
4757 content scanning code have been changed to use this function.
4759 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4760 to 24 hours. This avoids potential overflow problems when processing G
4761 and H retry rules. I suspect nobody ever tinkers with this value.
4763 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4765 PH/45 When the plaintext authenticator is running as a client, the server's
4766 challenges are checked to ensure they are valid base64 strings. By
4767 default, the authentication attempt is cancelled if an invalid string is
4768 received. Setting client_ignore_invalid_base64 true ignores these errors.
4769 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4770 they are received. Thus, the responses can be made to depend on the
4771 challenges. If an invalid string is ignored, an empty string is placed in
4774 PH/46 Messages that are created by the autoreply transport now contains a
4775 References: header, in accordance with RFCs 2822 and 3834.
4777 PH/47 Added authenticated_sender_force to the smtp transport.
4779 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4781 PH/49 Installed latest patch from the Sieve maintainer.
4783 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4784 was also set, the appendfile transport was unnecessarily scanning a
4785 directory of message files (e.g. for maildir delivery) to find the count
4786 of files (along with the size), even though it did not need this
4787 information. It now does the scan only if it needs to find either the
4788 size of the count of files.
4790 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4792 PH/52 Two bugs concerned with error handling when the smtp transport is
4795 (i) Exim was not creating retry information for temporary errors given
4796 for individual recipients after the DATA command when the smtp transport
4797 was used in LMTP mode. This meant that they could be retried too
4798 frequently, and not timed out correctly.
4800 (ii) Exim was setting the flag that allows error details to be returned
4801 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4802 recipients that were returned after the DATA command.
4804 PH/53 This is related to PH/52, but is more general: for any failing address,
4805 when detailed error information was permitted to be returned to the
4806 sender, but the error was temporary, then after the final timeout, only
4807 "retry timeout exceeded" was returned. Now it returns the full error as
4808 well as "retry timeout exceeded".
4810 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4811 do this, and (what is worse) MTAs that accept it.
4813 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4814 will now be deprecated.
4816 PH/56 New os.c-cygwin from the Cygwin maintainer.
4818 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4819 formats (previously only available in exim formats via -bpr, -bpru,
4820 and -bpra. Now also available in native and exiqgrep formats)
4822 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4823 with very large, slow to parse queues
4825 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4827 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4829 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4830 responses to authentication challenges, though it was showing the
4831 challenges; (ii) I've removed the CR characters from the debug output for
4834 PH/58 Allow for the insertion of a newline as well as a space when a string
4835 is turned into more than one encoded-word during RFC 2047 encoding. The
4836 Sieve code now uses this.
4838 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4839 data_4xx, lost_connection, tls_required.
4841 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4842 message was being sent as an SMTP response.
4844 PH/61 Add -l and -k options to exicyclog.
4846 PH/62 When verifying, if an address was redirected to one new address, so that
4847 verification continued, and the new address failed or deferred after
4848 having set something in $address_data, the value of $address_data was not
4849 passed back to the ACL. This was different to the case when no
4850 redirection occurred. The value is now passed back in both cases.
4852 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4853 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4854 use login_cap.h, so on its own it isn't the distinguishing feature. The
4855 new name refers directly to the setclassresources() function.
4857 PH/65 Added configuration files for NetBSD3.
4859 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4861 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4862 is preferred over IPv4.
4864 PH/68 The bounce_return_message and bounce_return_body options were not being
4865 honoured for bounces generated during the reception of non-SMTP messages.
4866 In particular, this applied to messages rejected by the ACL. This bug has
4867 been fixed. However, if bounce_return_message is true and bounce_return_
4868 body is false, the headers that are returned for a non-SMTP message
4869 include only those that have been read before the error was detected.
4870 (In the case of an ACL rejection, they have all been read.)
4872 PH/69 The HTML version of the specification is now built in a directory called
4873 spec_html instead of spec.html, because the latter looks like a path with
4874 a MIME-type, and this confuses some software.
4876 PH/70 Catch two compiler warnings in sieve.c.
4878 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4879 function verify_get_ident() calls ip_connect() to connect a socket, but
4880 if the "connect()" function timed out, ip_connect() used to close the
4881 socket. However, verify_get_ident() also closes the socket later, and in
4882 between Exim writes to the log, which may get opened at this point. When
4883 the socket was closed in ip_connect(), the log could get the same file
4884 descriptor number as the socket. This naturally causes chaos. The fix is
4885 not to close the socket in ip_connect(); the socket should be closed by
4886 the function that creates it. There was only one place in the code where
4887 this was missing, in the iplookup router, which I don't think anybody now
4888 uses, but I've fixed it anyway.
4890 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4891 well as to direct DNS lookups. Otherwise the handling of names in host
4892 lists is inconsistent and therefore confusing.
4898 PH/01 Two changes to the default runtime configuration:
4900 (1) Move the checks for relay_from_hosts and authenticated clients from
4901 after to before the (commented out) DNS black list checks.
4903 (2) Add control=submission to the relay_from_hosts and authenticated
4904 clients checks, on the grounds that messages accepted by these
4905 statements are most likely to be submissions.
4907 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4909 (1) Generate an error if the third argument for the ${prvs expansion is
4912 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4915 (3) Reset the variables that are obtained from the first argument of
4916 ${prvscheck and used in the second argument before leaving the code,
4917 because their memory is reclaimed, so using them afterwards may do
4920 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4921 one (it's much easier than Tom thought :-).
4923 (5) Because of (4), we can now allow for the use of $prvscheck_result
4924 inside the third argument.
4926 PH/03 For some reason, the default setting of PATH when running a command from
4927 a pipe transport was just "/usr/bin". I have changed it to
4930 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4931 anything to be listed in the output from -bV.
4933 PH/05 When a filter generated an autoreply, the entire To: header line was
4934 quoted in the delivery log line, like this:
4936 => >A.N.Other <ano@some.domain> <original@ddress> ...
4938 This has been changed so that it extracts the operative address. There
4939 may be more than one such address. If so, they are comma-separated, like
4942 => >ano@some.domain,ona@other.domain <original@ddress> ...
4944 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
4945 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
4946 not being looked up in the rDNS to get a host name, Exim was showing the
4947 IP address twice in Received: lines, even though the IP addresses were
4948 identical. For example:
4950 Received: from [1.2.3.4] (helo=[1.2.3.4])
4952 However, if the real host name was known, it was omitting the HELO data
4953 if it matched the actual IP address. This has been tidied up so that it
4954 doesn't show the same IP address twice.
4956 PH/07 When both +timestamp and +memory debugging was on, the value given by
4957 $tod_xxx expansions could be wrong, because the tod_stamp() function was
4958 called by the debug printing, thereby overwriting the timestamp buffer.
4959 Debugging no longer uses the tod_stamp() function when +timestamp is set.
4961 PH/08 When the original message was included in an autoreply transport, it
4962 always said "this is a copy of the message, including all the headers",
4963 even if body_only or headers_only was set. It now gives an appropriate
4966 PH/09 Applied a patch from the Sieve maintainer which:
4968 o fixes some comments
4969 o adds the (disabled) notify extension core
4970 o adds some debug output for the result of if/elsif tests
4971 o points to the current vacation draft in the documentation
4972 and documents the missing references header update
4976 o fixes a bug in processing the envelope test (when testing
4977 multiple envelope elements, the last element determined the
4980 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
4981 Electronic Mail") by including:
4983 Auto-submitted: auto-generated
4985 in the messages that it generates (bounce messages and others, such as
4986 warnings). In the case of bounce messages for non-SMTP messages, there was
4987 also a typo: it was using "Auto_submitted" (underscore instead of
4988 hyphen). Since every message generated by Exim is necessarily in response
4989 to another message, thes have all been changed to:
4991 Auto-Submitted: auto-replied
4993 in accordance with these statements in the RFC:
4995 The auto-replied keyword:
4997 - SHOULD be used on messages sent in direct response to another
4998 message by an automatic process,
5000 - MUST NOT be used on manually-generated messages,
5002 - MAY be used on Delivery Status Notifications (DSNs) and Message
5003 Disposition Notifications (MDNs),
5005 - MUST NOT be used on messages generated by automatic or periodic
5006 processes, except for messages which are automatic responses to
5009 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5010 to the default Received: header definition.
5012 PH/12 Added log selector acl_warn_skipped (default on).
5014 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5015 variables because (a) they are in the wrong storage pool and (b) even if
5016 they were copied, it wouldn't work properly because of the caching.
5018 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5019 checking when decoding. Apparently there are clients that generate
5020 overlong encoded strings. Why am I not surprised?
5022 PH/15 If the first argument of "${if match_address" was not empty, but did not
5023 contain an "@" character, Exim crashed. Now it writes a panic log message
5024 and treats the condition as false.
5026 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5028 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5029 extension "envelope-auth". The code is finished and in agreement with
5030 other implementations, but there is no documentation so far and in fact,
5031 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5032 not changing the active code.
5034 Print executed "if" and "elsif" statements when debugging is used. This
5035 helps a great deal to understand what a filter does.
5037 Document more things not specified clearly in RFC3028. I had all this
5038 sorted out, when out of a sudden new issues came to my mind. Oops."
5040 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5043 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5044 canonical form (as documented). However, after a host name lookup from
5045 the IP address, check_host() was doing a simple string comparison with
5046 addresses acquired from the DNS when checking that the found name did
5047 have the original IP as one of its addresses. Since any found IPv6
5048 addresses are likely to be in abbreviated form, the comparison could
5049 fail. Luckily, there already exists a function for doing the comparison
5050 by converting both addresses to binary, so now that is used instead of
5051 the text comparison.
5053 PH/20 There was another similar case to PH/19, when a complete host name was
5054 given in a host list; looking up its IP address could give an abbreviated
5055 form, whereas the current host's name might or might not be abbreviated.
5056 The same fix has been applied.
5062 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5063 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5066 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5067 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5069 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5071 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5072 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5073 Also, for "Auto-Submitted", treat anything other than "no" as
5074 non-personal, in accordance with RFC 3834. (Previously it treated
5075 anything starting "auto-" as non-personal.)
5077 TF/01 The control=submission/name=... option had a problem with syntax
5078 errors if the name included a slash character. The /name= option
5079 now slurps the rest of the string, so it can include any characters
5080 but it must come last in the list of options (after /sender_retain
5083 PH/05 Some modifications to the interface to the fake nameserver for the new
5091 TK/01 Added the "success_on_redirect" address verification option. See
5092 NewStuff for rationale and an example.
5094 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5096 PH/02 Patch to exigrep to allow it to work on syslog lines.
5098 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5099 fread() to skip over the body file's header line, because in Cygwin the
5100 header line is locked and is inaccessible.
5102 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5103 co-exist for some time) to make it clear that it is the Exim ID that is
5104 referenced, not the Message-ID: header line.
5106 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5107 string_format() function, because snprintf() does not exist on all
5110 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5111 ${stat: expansion item.
5113 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5114 protocol synchronization error", to keep the pedants happy.
5116 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5117 well as for IRIX systems, when gcc is being used. See the host.c source
5120 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5122 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5125 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5126 transport and to the smtp transport in LMTP mode.
5128 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5130 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5131 run a filter in a subprocess. This could lead to confusion in subsequent
5132 lookups in the parent process. There should also be a search_tidyup() at
5133 the end of the subprocess.
5135 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5136 only if the host matched helo_try_verify_hosts, which caused the
5137 verification to occur when the EHLO/HELO command was issued. The ACL just
5138 tested the remembered result. Now, if a previous verification attempt has
5139 not happened, "verify = helo" does it there and then.
5141 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5143 TK/03 Fix log output including CR from clamd.
5145 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5146 exist provoked a memory error which could cause a segfault.
5148 PH/15 Installed PCRE 6.2
5150 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5152 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5153 of the problem. Specifically, suggested +O2 rather than +O1 for the
5156 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5158 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5159 record, the port was not correctly logged when the outgoing_port log
5160 selector was set (it logged the transort's default port).
5162 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5163 fallback_hosts, and "hosts" in the smtp transport.
5165 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5166 host errors such as "Connection refused".
5168 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5169 authentication with radiusclient 0.4.9:
5171 - Error returned from rc_read_config was caught wrongly
5172 - Username/password not passed on to radius server due to wrong length.
5174 The presumption is that some radiusclient API changes for 4.51/PH/17
5175 were not taken care of correctly. The code is still untested by me (my
5176 Linux distribution still has 0.3.2 of radiusclient), but it was
5177 contributed by a Radius user.
5179 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5180 expanding the "port" option of the smtp transport.
5182 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5183 while reading a MIME header. Thanks to Tom Hughes for a patch.
5185 PH/24 Include config.h inside local_scan.h so that configuration settings are
5188 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5189 that in an ACL for RCPT (for example), you can examine exactly what was
5192 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5193 commands, but it was not correctly comparing the address with the actual
5194 client host address. Thus, it would show the EHLO address in Received:
5195 header lines when this was not necessary.
5197 PH/27 Added the % operator to ${eval:}.
5199 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5200 it should be ignoring failures (because with -C, for example, it has lost
5201 privilege). It wasn't ignoring creation failures other than "already
5204 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5207 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5208 because an input line was too long, either on its own, or by virtue of
5209 too many continuations, the temporary file was not being removed, and the
5210 return code was incorrect.
5212 PH/31 Missing "BOOL" in function definition in filtertest.c.
5214 PH/32 Applied Sieve patches from the maintainer.
5216 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5218 PH/33 Added "verify = not_blind".
5220 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5221 Local/Makefile (with some defaults set). These are used in built scripts
5222 such as exicyclog, but they have never been used in the exim_install
5223 script (though there are many overriding facilities there). I have
5224 arranged that the exim_install script now takes note of these two
5227 PH/35 Installed configuration files for Dragonfly.
5229 PH/36 When a locally submitted message by a trusted user did not contain a
5230 From: header, and the sender address was obtained from -f or from an SMTP
5231 MAIL command, and the trusted user did not use -F to supply a sender
5232 name, $originator_name was incorrectly used when constructing a From:
5233 header. Furthermore, $originator_name was used for submission mode
5234 messages from external hosts without From: headers in a similar way,
5235 which is clearly wrong.
5237 PH/37 Added control=suppress_local_fixups.
5239 PH/38 When log_selector = +received_sender was set, and the addition of the
5240 sender made the log line's construction buffer exactly full, or one byte
5241 less than full, an overflow happened when the terminating "\n" was
5244 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5245 when the result of a list match is failure because a DNS lookup failed.
5247 PH/40 RM_COMMAND is now used in the building process.
5249 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5250 the "build-* directories that it finds.
5252 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5253 address in a domain literal was a prefix of an interface address.
5255 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5256 when verifying a sender address, unless rewrite_headers is false.
5258 PH/44 Wrote a long comment about why errors_to addresses are verified as
5259 recipients, not senders.
5261 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5262 the ratelimit ACL was added.
5264 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5266 PH/46 Added extra information about PostgreSQL errors to the error string.
5268 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5269 avoiding the need to install special zones in a real server. This is
5270 backwards compatible; if it can't find the fake resolver, it drops back.
5271 Thus, both old and new test suites can be run.
5273 TF/02 Added util/ratelimit.pl
5275 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5276 clock is set back in time.
5278 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5279 Candler <B.Candler@pobox.com>.
5281 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5282 OK for addresses that are the result of redirections.
5284 PH/48 A number of further additions for the benefit of the new test suite,
5285 including a fake gethostbyname() that interfaces to the fake DNS resolver
5288 TF/06 The fix for widen_domains has also been applied to qualify_single and
5289 search_parents which are the other dnslookup options that can cause
5292 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5295 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5297 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5298 service" if it encountered an MX record with an empty target hostname.
5299 The message is now "an MX or SRV record indicated no SMTP service".
5301 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5302 if the DNS of the sending site is misconfigured. This is quite a
5303 common situation. This change restores the behaviour of treating a
5304 helo verification defer as a failure.
5306 PH/51 If self=fail was set on a router, the bounce message did not include the
5307 actual error message.
5313 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5315 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5316 command itself ended in error, the underlying message about the transport
5317 filter timeout was being overwritten with the pipe command error. Now the
5318 underlying error message should be appended to the second error message.
5320 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5322 PH/02 Reduce the amount of output that "make" produces by default. Full output
5323 can still be requested.
5325 PH/03 The warning log line about a condition test deferring for a "warn" verb
5326 was being output only once per connection, rather than after each
5327 occurrence (because it was using the same function as for successful
5328 "warn" verbs). This seems wrong, so I have changed it.
5330 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5331 it should not have, which might have caused a crash in the right
5332 circumstances, but probably never did.
5334 PH/04 Installed a modified version of Tony Finch's patch to make submission
5335 mode fix the return path as well as the Sender: header line, and to
5336 add a /name= option so that you can make the user's friendly name appear
5339 TF/03 Added the control = fakedefer ACL modifier.
5341 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5342 Mark Lowes for thorough testing.
5344 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5346 TK/03 Merged latest SRS patch from Miles Wilton.
5348 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5349 with the definition in sysexits.h (which is #included earlier).
5350 Fortunately, Exim does not actually use EX_OK. The code used to try to
5351 preserve the sysexits.h value, by assuming that macro definitions were
5352 scanned for macro replacements. I have been disabused of this notion,
5353 so now the code just undefines EX_OK before #including unistd.h.
5355 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5356 in the smtp transport. When a block could not be written in a single
5357 write() function, the timeout was being re-applied to each part-write.
5358 This seems wrong - if the receiver was accepting one byte at a time it
5359 would take for ever. The timeout is now adjusted when this happens. It
5360 doesn't have to be particularly precise.
5362 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5363 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5365 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5366 without a domain if the check to <postmaster@domain> fails.
5368 SC/01 Eximstats: added -xls and the ability to specify output files
5369 (patch written by Frank Heydlauf).
5371 SC/02 Eximstats: use FileHandles for outputting results.
5373 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5375 SC/04 Eximstats: fixed display of large numbers with -nvr option
5377 SC/05 Eximstats: fixed merging of reports with empty tables.
5379 SC/06 Eximstats: added the -include_original_destination flag
5381 SC/07 Eximstats: removed tabs and trailing whitespace.
5383 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5385 TK/06 MBOX spool code: Add real "From " MBOX separator line
5386 so the .eml file is really in mbox format (even though
5387 most programs do not really care). Patch from Alex Miller.
5389 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5390 The latter is generated from $received_to and is only set if the
5391 message has one envelope recipient. SA can use these headers,
5392 obviously out-of-the-box. Patch from Alex Miller.
5394 PH/08 The ${def test on a variable was returning false if the variable's
5395 value was "0", contrary to what the specification has always said!
5396 The result should be true unless the variable is empty.
5398 PH/09 The syntax error of a character other than { following "${if
5399 def:variable_name" (after optional whitespace) was not being diagnosed.
5400 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5401 accidental colon was present, for example, could give incorrect results.
5403 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5404 result is used (not including appendfile, where other changes are about
5407 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5408 This involved changing a lot of size variables from int to off_t. It
5409 should work with maildirs and everything.
5411 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5412 spamd dying while we are connected to it.
5414 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5417 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5418 function for BDB 4.3.
5420 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5422 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5423 the sources for PCRE have been re-arranged and more files are now
5426 PH/15 The code I had for printing potentially long long variables in PH/11
5427 above was not the best (it lost precision). The length of off_t variables
5428 is now inspected at build time, and an appropriate printing format (%ld
5429 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5430 to be "long long int" or "long int". This is needed for the internal
5431 formatting function string_vformat().
5433 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5434 the configuration file to be ":syslog", then the script "guesses" where
5435 the logs files are, rather than using the compiled in default. In our
5436 case the guess is not the same as the compiled default, so the script
5437 suddenly stopped working when I started to use syslog. The patch checks
5438 to see if log_file_path is "". If so, it attempts to read it from exim
5439 with no configuration file to get the compiled in version, before it
5440 falls back to the previous guessing code."
5442 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5443 implementing BATV in an Exim configuration. See NewStuff for the gory
5446 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5447 Makefile that are specific to HP-UX.
5449 PH/18 If the "use_postmaster" option was set for a recipient callout together
5450 with the "random" option, the postmaster address was used as the MAIL
5451 FROM address for the random test, but not for the subsequent recipient
5452 test. It is now used for both.
5454 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5455 patch removes a few documentation additions to RFC 3028, because the
5456 latest draft now contains them. It adds the new en;ascii-case comparator
5457 and a new error check for 8bit text in MIME parts. Comparator and
5458 require names are now matched exactly. I enabled the subaddress
5459 extension, but it is not well tested yet (read: it works for me)."
5461 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5462 rework some of the code of TK/09 above to avoid the hardwired use of
5463 "%lld" and "long long". Replaced the call to snprintf() with a call to
5466 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5467 records point to non-existent hosts", "retry timeout exceeded", and
5468 "retry time not reached for any host after a long failure period".
5470 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5471 experimental DomainKeys support:
5473 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5474 (2) On an error such as an illegally used "control", the wrong name for
5475 the control was given.
5477 These problems did NOT occur unless DomainKeys support was compiled.
5479 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5481 PH/24 Added ${if match_ip condition.
5483 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5484 databases so that it will be absolutely obvious if a crash occurs in the
5485 DB library. This is a regular occurrence (often caused by mis-matched
5488 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5489 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5490 picked up on a user's system that detects such things. There doesn't seem
5491 to be a gcc warning option for this - only an attribute that has to be
5492 put on the function's prototype. It seems that in Fedora Core 4 they have
5493 set this on a number of new functions. No doubt there will be more in due
5496 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5497 specify a transport. However, if an address that was verified by such a
5498 router was the subject of a callout, Exim crashed because it tried to
5499 read the rcpt_include_affixes from the non-existent transport. Now it
5500 just assumes that the setting of that option is false. This bug was
5501 introduced by 4.51/PH/31.
5503 PH/28 Changed -d+all to exclude +memory, because that information is very
5504 rarely of interest, but it makes the output a lot bigger. People tend to
5505 do -d+all out of habit.
5507 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5508 code in os-type was giving problems when libc.so lives in lib64, like on
5511 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5512 aren't the modern standard, and it seems that some systems' include files
5513 don't always have them. Exim was already checking for some of the newer
5514 ones like T_AAAA, and defining it itself. I've added checks for all the
5515 record types that Exim uses.
5517 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5518 not automatically generating a new one, as it is supposed to. This
5519 prevented TLS from working. If the file did exist, but contained invalid
5520 data, a new version was generated, as expected. It was only the case of a
5521 non-existent file that was broken.
5523 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5524 with a change in libdomainkeys > 0.64.
5526 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5527 from DNS. If the selector record carries the flag, it now has
5528 precedence over the domain-wide flag.
5530 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5532 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5533 the use of a port name that isn't defined in /etc/services) occurred, the
5534 message was deferred as in a normal delivery, and thus remained on the
5535 spool, instead of being failed because of the mua_wrapper setting. This
5536 is now fixed, and I tidied up some of the mua_wrapper messages at the
5539 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5540 the messages in a hash of arrays rather than using individual hashes.
5541 This is a bit cleaner and results in dramatic memory savings, albeit
5542 at a slight CPU cost.
5544 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5545 as requested by Marc Sherman.
5547 SC/10 Eximstats: added histograms for user specified patterns as requested
5550 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5552 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5553 fopen() in the content-scanning modules that did not already have it.
5559 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5560 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5562 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5564 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5566 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5567 to test Sieve filters that use "vacation".
5569 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5570 that changes the way the GnuTLS parameters are stored in the cache file.
5571 The new format can be generated externally. For backward compatibility,
5572 if the data in the cache doesn't make sense, Exim assumes it has read an
5573 old-format file, and it generates new data and writes a new file. This
5574 means that you can't go back to an older release without removing the
5577 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5578 work if there are any delivery delays because "one_time" forces the
5579 parent to be marked "delivered", so its unseen clone is never tried
5580 again. For this reason, Exim now forbids the simultaneous setting of
5583 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5584 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5585 entry if you want to know the details. The fix, however, neglected to
5586 consider the case when local delivery batching is involved. The test for
5587 "previously delivered" was not happening when checking to see if an
5588 address could be batched with a previous (undelivered) one; under
5589 certain circumstances this could lead to multiple deliveries to the same
5592 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5593 in its include files, and this causes problems building Exim.
5595 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5596 header_syntax) but Exim was just ignoring anything given after a slash.
5597 In particular, this caused confusion with an attempt to use "verify =
5598 reverse_host_lookup/defer_ok". An error is now given when options are
5599 supplied for verify items that do not have them. (Maybe reverse_host_
5600 lookup should have a defer_ok option, but that's a different point.)
5602 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5603 defined by RFC 821) to 2048, because there were problems with some AUTH
5604 commands, and RFC 1869 says the size should be increased for extended
5605 SMTP commands that take arguments.
5607 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5610 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5611 "unknown" error; now it says that the functionality isn't in the binary.
5613 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5614 an address' error message when a string expansion fails (syntax or
5615 whatever). Otherwise the password may appear in the log. Following change
5616 PH/42 below, there is no longer a chance of it appearing in a bounce
5619 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5621 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5622 of its addresses was ever tried. (Bugzilla bug #2.)
5624 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5625 the result incorrectly in the debug output. (It correctly added a newline
5626 to what was transported.)
5628 TF/01 Added $received_time.
5630 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5631 commented out examples of how to interface to a virus scanner and to
5632 SpamAssassin. Also added commented examples of av_scanner and
5633 spamd_address settings.
5635 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5636 and controls are allowed in which ACLs. There were a couple of minor
5637 errors. Some of the entries in the conditions table (which is a table of
5638 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5639 negation of where the condition IS allowed.
5641 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5643 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5644 header file does not have a version number, so I've had to invent a new
5645 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5646 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5647 radiusclient), but it was contributed by a Radius user.
5649 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5650 files or directories, for OpenSSL.
5652 PH/19 When an Exim process that is running as root has to create an Exim log
5653 file, it does so in a subprocess that runs as exim:exim so as to get the
5654 ownership right at creation (otherwise, other Exim processes might see
5655 the file with the wrong ownership). There was no test for failure of this
5656 fork() call, which would lead to the process getting stuck as it waited
5657 for a non-existent subprocess. Forks do occasionally fail when resources
5658 run out. I reviewed all the other calls to fork(); they all seem to check
5661 PH/20 When checking for unexpected SMTP input at connect time (before writing
5662 the banner), Exim was not dealing correctly with a non-positive return
5663 from the read() function. If the client had disconnected by this time,
5664 the result was a log entry for a synchronization error with an empty
5665 string after "input=" when read() returned zero. If read() returned -1
5666 (an event I could not check), uninitialized data bytes were printed.
5667 There were reports of junk text (parts of files, etc) appearing after
5670 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5672 PH/22 Added support for macro redefinition, and (re)definition in between
5673 driver and ACL definitions.
5675 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5676 forgetting to use the resulting value; it was using the unexpanded value.
5678 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5679 hadn't been configured. The fix is from Juergen Kreileder, who
5680 understands it better than I do:
5682 "Here's what I see happening with three configured cyrus_sasl
5683 authenticators configured (plain, login, cram-md5):
5685 On startup auth_cyrus_sasl_init() gets called for each of these.
5686 This means three calls to sasl_listmech() without a specified mech_list.
5687 => SASL tests which mechs of all available mechs actually work
5688 => three warnings about OTP not working
5689 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5691 With the patch, sasl_listmech() also gets called three times. But now
5692 SASL's mech_list option is set to the server_mech specified in the the
5693 authenticator. Or in other words, the answer from sasl_listmech()
5694 gets limited to just the mech you're testing for (which is different
5696 => the return list contains just 'plain' or 'login', 'cram-md5' or
5697 nothing depending on the value of ob->server_mech.
5699 I've just tested the patch: Authentication still works fine,
5700 unavailable mechs specified in the exim configuration are still
5701 caught, and the auth.log warnings about OTP are gone."
5703 PH/25 When debugging is enabled, the contents of the command line are added
5704 to the debugging output, even when log_selector=+arguments is not
5707 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5708 answer is "GNU", and only if the return is "GNU/something" is the answer
5711 PH/27 $acl_verify_message is now set immediately after the failure of a
5712 verification in an ACL, and so is available in subsequent modifiers. In
5713 particular, the message can be preserved by coding like this:
5715 warn !verify = sender
5716 set acl_m0 = $acl_verify_message
5718 Previously, $acl_verify_message was set only while expanding "message"
5719 and "log_message" when a very denied access.
5721 PH/28 Modified OS/os.c-Linux with
5723 -#ifndef OS_LOAD_AVERAGE
5724 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5726 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5727 nomenclature these days.)
5729 PH/29 Installed patch from the Sieve maintainer that adds the options
5730 sieve_useraddress and sieve_subaddress to the redirect router.
5732 PH/30 In these circumstances:
5733 . Two addresses routed to the same list of hosts;
5734 . First host does not offer TLS;
5735 . First host accepts first address;
5736 . First host gives temporary error to second address;
5737 . Second host offers TLS and a TLS session is established;
5738 . Second host accepts second address.
5739 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5740 and peerdn, if requested) that were in fact used only for the second
5743 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5744 attention to any local part prefix or suffix that was matched by the
5745 router that accepted the address. It now behaves in the same way as it
5746 does for delivery: the affixes are removed from the local part unless
5747 rcpt_include_affixes is set on the transport.
5749 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5750 timeout during the DATA phase of an incoming message.
5752 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5753 applied a patch sanctioned by the Sieve maintainer.
5755 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5756 the uid or gid is negative. A case of a negative gid caused this to be
5757 noticed. The fix allows for either to be negative.
5759 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5760 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5763 PH/36 The change PH/12 above was broken. Fixed it.
5765 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5766 the grounds that routing the same address twice would always produce the
5767 same answer. This might have been true once, but it is certainly no
5768 longer true now. Routing a child address may depend on the previous
5769 routing that produced that child. Some complicated redirection strategies
5770 went wrong when messages had multiple recipients, and made Exim's
5771 behaviour dependent on the order in which the addresses were given.
5773 I have moved the duplicate checking until after the routing is complete.
5774 Exim scans the addresses that are assigned to local and remote
5775 transports, and removes any duplicates. This means that more work will be
5776 done, as duplicates will always all be routed, but duplicates are
5777 presumably rare, so I don't expect this is of any significance.
5779 For deliveries to pipes, files, and autoreplies, the duplicate checking
5780 still happens during the routing process, since they are not going to be
5783 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5784 It corrects a timeout issue with spamd. This is Ian's comment: "The
5785 background is that sometimes spamd either never reads data from a
5786 connection it has accepted, or it never writes response data. The exiscan
5787 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5788 blindly assumes that writes won't block so it may never time out."
5790 PH/39 Allow G after quota size as well as K and M.
5792 PH/40 The value set for $authenticated_id in an authenticator may not contain
5793 binary zeroes or newlines because the value is written to log lines and
5794 to spool files. There was no check on this. Now the value is run through
5795 the string_printing() function so that such characters are converted to
5796 printable escape sequences.
5798 PH/41 $message_linecount is a new variable that contains the total number of
5799 lines in the message. Compare $body_linecount, which is the count for the
5802 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5803 bounce and delay warning messages, except in certain special cases, which
5806 (a) An SMTP error message from a remote host;
5807 (b) A message specified in a :fail: redirection;
5808 (c) A message specified in a "fail" command in a system filter;
5809 (d) A message specified in a FAIL return from the queryprogram router;
5810 (e) A message specified by the cannot_route_message router option.
5812 In these cases only, Exim does include the error details in bounce and
5813 warning messages. There are also a few cases where bland messages such
5814 as "unrouteable address" or "local delivery error" are given.
5816 PH/43 $value is now also set for the "else" part of a ${run expansion.
5818 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5819 being worked on, but at least Exim now implements the latest version to
5822 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5823 process to complete was treated as a delivery failure, a timeout while
5824 writing the message to the pipe was logged, but erroneously treated as a
5825 successful delivery. Such timeouts include transport filter timeouts. For
5826 consistency with the overall process timeout, these timeouts are now
5827 treated as errors, giving rise to delivery failures by default. However,
5828 there is now a new Boolean option for the pipe transport called
5829 timeout_defer, which, if set TRUE, converts the failures into defers for
5830 both kinds of timeout. A transport filter timeout is now identified in
5833 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5834 systems where "make" and "gmake" are different, calling "gmake" at top
5835 level broke things. I've arranged for the value of $(MAKE) to be passed
5836 from the Makefile to this script so that it can call the same version of
5840 A note about Exim versions 4.44 and 4.50
5841 ----------------------------------------
5843 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5844 changes of various kinds. As a consequence, a big documentation update was
5845 needed. This delayed the release for rather longer than seemed good, especially
5846 in the light of a couple of (minor) security issues. Therefore, the changes
5847 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5848 release. So 4.44 and 4.50 are in effect two different branches that both start
5851 I have left the 4.50 change log unchanged; it contains all the changes since
5852 4.43. The change log for 4.44 is below; many of its items are identical to
5853 those for 4.50. This seems to be the most sensible way to preserve the
5854 historical information.
5860 1. Minor wording change to the doc/README.SIEVE file.
5862 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5863 computation of the current number of files was incorrect.
5865 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5866 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5869 4. Give more explanation in the error message when the command for a transport
5870 filter fails to execute.
5872 5. There are several places where Exim runs a non-Exim command in a
5873 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5874 was being done only for the command run by the queryprogram router. It is
5875 now done for all such subprocesses. The other cases are: ${run, transport
5876 filters, and the commands run by the lmtp and pipe transports.
5878 6. Added CONFIGURE_GROUP build-time option.
5880 7. Some older OS have a limit of 256 on the maximum number of file
5881 descriptors. Exim was using setrlimit() to set 1000 as a large value
5882 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5883 systems. I've change it so that if it can't get 1000, it tries for 256.
5885 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5886 was an oversight, and furthermore, ever since the addition of extra
5887 controls (e.g. 4.43/32), the checks on when to allow different forms of
5888 "control" were broken. There should now be diagnostics for all cases when a
5889 control that does not make sense is encountered.
5891 9. Added the /retain_sender option to "control=submission".
5893 10. $recipients is now available in the predata ACL (oversight).
5895 11. Tidy the search cache before the fork to do a delivery from a message
5896 received from the command line. Otherwise the child will trigger a lookup
5897 failure and thereby defer the delivery if it tries to use (for example) a
5898 cached ldap connection that the parent has called unbind on.
5900 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5901 of $address_data from the recipient verification was clobbered by the
5902 sender verification.
5904 13. The value of address_data from a sender verification is now available in
5905 $sender_address_data in subsequent conditions in the ACL statement.
5907 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5909 15. Added a new option "connect=<time>" to callout options, to set a different
5912 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5913 was its contents. (It was OK if the option was not defined at all.)
5915 17. A "Completed" log line is now written for messages that are removed from
5916 the spool by the -Mrm option.
5918 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5919 information about exactly what failed.
5921 19. Added -dd to debug only the daemon process.
5923 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5924 handles timeouts, both on the server side and network timeouts. Renamed the
5925 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5927 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5928 It is now set to "smtps".
5930 22. $host_address is now set to the target address during the checking of
5931 ignore_target_hosts.
5933 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5934 being passed; this would have caused $sender_host_name to have been used if
5935 matching the list had actually called for a host name (not very likely,
5936 since this list is usually IP addresses). A host name is now passed as
5939 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5940 code that specifies a non-restarting handler (typically sigaction() in
5941 modern systems) in an attempt to fix a rare and obscure crash bug.
5943 25. Narrowed the window for a race in the daemon that could cause it to ignore
5944 SIGCHLD signals. This is not a major problem, because they are used only to
5945 wake it up if nothing else does.
5947 26. A malformed maildirsize file could cause Exim to calculate negative values
5948 for the mailbox size or file count. Odd effects could occur as a result.
5949 The maildirsize information is now recalculated if the size or filecount
5952 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5953 support for a long time. Removed HAVE_SYS_VFS_H.
5955 28. Installed the latest version of exipick from John Jetmore.
5957 29. In an address list, if the pattern was not a regular expression, an empty
5958 subject address (from a bounce message) matched only if the pattern was an
5959 empty string. Non-empty patterns were not even tested. This was the wrong
5960 because it is perfectly reasonable to use an empty address as part of a
5961 database query. An empty address is now tested by patterns that are
5962 lookups. However, all the other forms of pattern expect the subject to
5963 contain a local part and a domain, and therefore, for them, an empty
5964 address still always fails if the pattern is not itself empty.
5966 30. Exim went into a mad DNS loop when attempting to do a callout where the
5967 host was specified on an smtp transport, and looking it up yielded more
5968 than one IP address.
5970 31. Re-factored the code for checking spool and log partition space into a
5971 function that finds that data and another that does the check. The former
5972 is then used to implement four new variables: $spool_space, $log_space,
5973 $spool_inodes, and $log_inodes.
5975 32. The RFC2047 encoding function was originally intended for short strings
5976 such as real names; it was not keeping to the 75-character limit for
5977 encoded words that the RFC imposes. It now respects the limit, and
5978 generates multiple encoded words if necessary. To be on the safe side, I
5979 have increased the buffer size for the ${rfc2047: expansion operator from
5982 33. It is now permitted to omit both strings after an "if" condition; if the
5983 condition is true, the result is "true". As before, when the second string
5984 is omitted, a false condition yields an empty string. This makes it less
5985 cumbersome to write custom ACL and router conditions.
5987 34. Failure to deliver a bounce message always caused it to be frozen, even if
5988 there was an errors_to setting on the router. The errors_to setting is now
5991 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5992 canonical form (fully expanded) before being placed in
5993 $sender_host_address.
5995 36. The table in the code that translates DNS record types into text (T_A to
5996 "A" for instance) was missing entries for NS and CNAME. It is just possible
5997 that this could have caused confusion if both these types were looked up
5998 for the same domain, because the text type is used as part of Exim's
5999 per-process caching. But the chance of anyone hitting this buglet seems
6002 37. The dnsdb lookup has been extended in a number of ways.
6004 (1) There is a new type, "zns", which walks up the domain tree until it
6005 finds some nameserver records. It should be used with care.
6007 (2) There is a new type, "mxh", which is like "mx" except that it returns
6008 just the host names, not the priorities.
6010 (3) It is now possible to give a list of domains (or IP addresses) to be
6011 looked up. The behaviour when one of the lookups defers can be
6012 controlled by a keyword.
6014 (4) It is now possible to specify the separator character for use when
6015 multiple records are returned.
6017 38. The dnslists ACL condition has been extended: it is now possible to supply
6018 a list of IP addresses and/or domains to be looked up in a particular DNS
6021 39. Added log_selector=+queue_time_overall.
6023 40. When running the queue in the test harness, wait just a tad after forking a
6024 delivery process, to get repeatability of debugging output.
6026 41. Include certificate and key file names in error message when GnuTLS fails
6027 to set them up, because the GnuTLS error message doesn't include the name
6028 of the failing file when there is a problem reading it.
6030 42. Allow both -bf and -bF in the same test run.
6032 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6034 44. The "Exiscan patch" is now merged into the mainline Exim source.
6036 45. Sometimes the final signoff response after QUIT could fail to get
6037 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6038 < 0 before doing a fflush(). This bug looks as though it goes back to the
6039 introduction of TLS in release 3.20, but "sometimes" must have been rare
6040 because the tests only now provoked it.
6042 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6043 (this can affect the format of dates).
6045 47. exim_tidydb, when checking for the continued existence of a message for
6046 which it has found a message-specific retry record, was not finding
6047 messages that were in split spool directories. Consequently, it was
6048 deleting retry records that should have stayed in existence.
6050 48. Steve fixed some bugs in eximstats.
6052 49. The SPA authentication driver was not abandoning authentication and moving
6053 on to the next authenticator when an expansion was forced to fail,
6054 contradicting the general specification for all authenticators. Instead it
6055 was generating a temporary error. It now behaves as specified.
6057 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6058 (the order specifies the preference for clients). The order is now AES256,
6059 AES128, 3DES, ARCFOUR128.
6061 51. Small patch to Sieve code - explicitly set From: when generating an
6064 52. Exim crashed if a remote delivery caused a very long error message to be
6065 recorded - for instance if somebody sent an entire SpamAssassin report back
6066 as a large number of 550 error lines. This bug was coincidentally fixed by
6067 increasing the size of one of Exim's internal buffers (big_buffer) that
6068 happened as part of the Exiscan merge. However, to be on the safe side, I
6069 have made the code more robust (and fixed the comments that describe what
6072 53. Now that there can be additional text after "Completed" in log lines (if
6073 the queue_time_overall log selector is set), a one-byte patch to exigrep
6074 was needed to allow it to recognize "Completed" as not the last thing in
6077 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6078 patch that reportedly fixes this has been added. I am not expert enough to
6079 create a test for it. This is what the patch creator wrote:
6081 "I found a little strange behaviour of ldap code when working with
6082 Windows 2003 AD Domain, where users was placed in more than one
6083 Organization Units. When I tried to give exim partial DN, the exit code
6084 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6085 But simultaneously result of request was absolutely normal ldap result,
6086 so I produce this patch..."
6088 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6089 REFERENCE, so I have modified the code to exclude the patch when that macro
6092 55. Some experimental protocols are using DNS PTR records for new purposes. The
6093 keys for these records are domain names, not reversed IP addresses. The
6094 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6095 leaves it alone. Component reversal etc. now happens only for IP addresses.
6098 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6100 57. Double the size of the debug message buffer (to 2048) so that more of very
6101 long debug lines gets shown.
6103 58. The exicyclog utility now does better if the number of log files to keep
6104 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6106 59. Two changes related to the smtp_active_hostname option:
6108 (1) $smtp_active_hostname is now available as a variable.
6109 (2) The default for smtp_banner uses $smtp_active_hostname instead
6110 of $primary_hostname.
6112 60. The host_aton() function is supposed to be passed a string that is known
6113 to be a valid IP address. However, in the case of IPv6 addresses, it was
6114 not checking this. This is a hostage to fortune. Exim now panics and dies
6115 if the condition is not met. A case was found where this could be provoked
6116 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6117 components; fortuitously, this particular loophole had already been fixed
6118 by change 4.50/55 above.
6120 If there are any other similar loopholes, the new check in host_aton()
6121 itself should stop them being exploited. The report I received stated that
6122 data on the command line could provoke the exploit when Exim was running as
6123 exim, but did not say which command line option was involved. All I could
6124 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6125 running as the user.
6128 61. There was a buffer overflow vulnerability in the SPA authentication code
6129 (which came originally from the Samba project). I have added a test to the
6130 spa_base64_to_bits() function which I hope fixes it.
6133 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6134 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6136 63. The daemon start-up calls getloadavg() while still root for those OS that
6137 need the first call to be done as root, but it missed one case: when
6138 deliver_queue_load_max is set with deliver_drop_privilege. This is
6139 necessary for the benefit of the queue runner, because there is no re-exec
6140 when deliver_drop_privilege is set.
6142 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6143 This has been fixed.
6145 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6146 was in use, was not putting the data itself into the right store pool;
6147 consequently, it could be overwritten for a subsequent message in the same
6148 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6151 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6153 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6154 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6155 the function were treating the return as a boolean value, which happened to
6156 work because 0=false and not-0=true, but is not correct code.
6158 68. The host_aton() function was not handling scoped IPv6 addresses (those
6159 with, for example, "%eth0" on the end) correctly.
6161 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6162 negated items (that is, ~something) in unsigned ints. Some compilers
6163 apparently mutter when there is no cast.
6165 70. If an address verification called from an ACL failed, and did not produce a
6166 user-specific message (i.e. there was only a "system" message), nothing was
6167 put in $acl_verify_message. In this situation, it now puts the system
6170 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6171 session; change 4.31/43 added the unwanted input to the log line - except
6172 that it did not do this in the start of session case. It now does.
6174 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6175 This is wrong and can cause the other end to generate a synchronization
6176 error if it is another Exim or anything else that does the synchronization
6177 check. A QUIT command is no longer sent after a timeout.
6179 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6180 during host lookups.
6182 74. The defer_ok option of callout verification was not working if it was used
6183 when verifying addresses in header lines, that is, for this case:
6185 verify = header_sender/callout=defer_ok
6187 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6188 those file descriptors could be used for SMTP connections. If anything
6189 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6190 could be sent to the SMTP client, causing chaos. The daemon now opens
6191 stdin, stdout, and stderr to /dev/null when it puts itself into the
6194 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6195 log by default. The user can override this with suitable Perl magic.
6197 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6198 the log message when discard triggers, was not working for the DATA ACL or
6199 for the non-SMTP ACL.
6201 78. Error message wording change in sieve.c.
6203 79. If smtp_accept_max_per_host was set, the number of connections could be
6204 restricted to fewer than expected, because the daemon was trying to set up
6205 a new connection before checking whether the processes handling previous
6206 connections had finished. The check for completed processes is now done
6207 earlier. On busy systems, this bug wouldn't be noticed because something
6208 else would have woken the daemon, and it would have reaped the completed
6211 80. If a message was submitted locally by a user whose login name contained one
6212 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6213 It caused a spool format error. I have fixed the spool reading code. A
6214 related problem was that the "from" clause in the Received: line became
6215 illegal because of the space(s). It is now covered by ${quote_local_part.
6217 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6220 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6222 83. Patch from Sieve maintainer to fix unterminated string problem in
6223 "vacation" handling.
6225 84. Some minor changes to the Linux configuration files to help with other
6226 OS variants using glibc.
6228 85. One more patch for Sieve to update vacation handling to latest spec.
6231 ----------------------------------------------------
6232 See the note above about the 4.44 and 4.50 releases.
6233 ----------------------------------------------------
6239 1. Change 4.43/35 introduced a bug that caused file counts to be
6240 incorrectly computed when quota_filecount was set in an appendfile
6243 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6244 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6247 3. Give more explanation in the error message when the command for a transport
6248 filter fails to execute.
6250 4. There are several places where Exim runs a non-Exim command in a
6251 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6252 was being done only for the command run by the queryprogram router. It is
6253 now done for all such subprocesses. The other cases are: ${run, transport
6254 filters, and the commands run by the lmtp and pipe transports.
6256 5. Some older OS have a limit of 256 on the maximum number of file
6257 descriptors. Exim was using setrlimit() to set 1000 as a large value
6258 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6259 systems. I've change it so that if it can't get 1000, it tries for 256.
6261 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6262 was an oversight, and furthermore, ever since the addition of extra
6263 controls (e.g. 4.43/32), the checks on when to allow different forms of
6264 "control" were broken. There should now be diagnostics for all cases when a
6265 control that does not make sense is encountered.
6267 7. $recipients is now available in the predata ACL (oversight).
6269 8. Tidy the search cache before the fork to do a delivery from a message
6270 received from the command line. Otherwise the child will trigger a lookup
6271 failure and thereby defer the delivery if it tries to use (for example) a
6272 cached ldap connection that the parent has called unbind on.
6274 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6275 of $address_data from the recipient verification was clobbered by the
6276 sender verification.
6278 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6279 was its contents. (It was OK if the option was not defined at all.)
6281 11. A "Completed" log line is now written for messages that are removed from
6282 the spool by the -Mrm option.
6284 12. $host_address is now set to the target address during the checking of
6285 ignore_target_hosts.
6287 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6288 being passed; this would have caused $sender_host_name to have been used if
6289 matching the list had actually called for a host name (not very likely,
6290 since this list is usually IP addresses). A host name is now passed as
6293 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6294 code that specifies a non-restarting handler (typically sigaction() in
6295 modern systems) in an attempt to fix a rare and obscure crash bug.
6297 15. Narrowed the window for a race in the daemon that could cause it to ignore
6298 SIGCHLD signals. This is not a major problem, because they are used only to
6299 wake it up if nothing else does.
6301 16. A malformed maildirsize file could cause Exim to calculate negative values
6302 for the mailbox size or file count. Odd effects could occur as a result.
6303 The maildirsize information is now recalculated if the size or filecount
6306 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6307 support for a long time. Removed HAVE_SYS_VFS_H.
6309 18. Updated exipick to current release from John Jetmore.
6311 19. Allow an empty sender to be matched against a lookup in an address list.
6312 Previously the only cases considered were a regular expression, or an
6315 20. Exim went into a mad DNS lookup loop when doing a callout where the
6316 host was specified on the transport, if the DNS lookup yielded more than
6319 21. The RFC2047 encoding function was originally intended for short strings
6320 such as real names; it was not keeping to the 75-character limit for
6321 encoded words that the RFC imposes. It now respects the limit, and
6322 generates multiple encoded words if necessary. To be on the safe side, I
6323 have increased the buffer size for the ${rfc2047: expansion operator from
6326 22. Failure to deliver a bounce message always caused it to be frozen, even if
6327 there was an errors_to setting on the router. The errors_to setting is now
6330 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6331 canonical form (fully expanded) before being placed in
6332 $sender_host_address.
6334 24. Updated eximstats to version 1.33
6336 25. Include certificate and key file names in error message when GnuTLS fails
6337 to set them up, because the GnuTLS error message doesn't include the name
6338 of the failing file when there is a problem reading it.
6340 26. Expand error message when OpenSSL has problems setting up cert/key files.
6343 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6344 (this can affect the format of dates).
6346 28. exim_tidydb, when checking for the continued existence of a message for
6347 which it has found a message-specific retry record, was not finding
6348 messages that were in split spool directories. Consequently, it was
6349 deleting retry records that should have stayed in existence.
6351 29. eximstats updated to version 1.35
6352 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6353 1.35 - bugfix such that pie charts by volume are generated correctly
6355 30. The SPA authentication driver was not abandoning authentication and moving
6356 on to the next authenticator when an expansion was forced to fail,
6357 contradicting the general specification for all authenticators. Instead it
6358 was generating a temporary error. It now behaves as specified.
6360 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6361 (the order specifies the preference for clients). The order is now AES256,
6362 AES128, 3DES, ARCFOUR128.
6364 31. Small patch to Sieve code - explicitly set From: when generating an
6367 32. Exim crashed if a remote delivery caused a very long error message to be
6368 recorded - for instance if somebody sent an entire SpamAssassin report back
6369 as a large number of 550 error lines. This bug was coincidentally fixed by
6370 increasing the size of one of Exim's internal buffers (big_buffer) that
6371 happened as part of the Exiscan merge. However, to be on the safe side, I
6372 have made the code more robust (and fixed the comments that describe what
6375 33. Some experimental protocols are using DNS PTR records for new purposes. The
6376 keys for these records are domain names, not reversed IP addresses. The
6377 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6378 leaves it alone. Component reversal etc. now happens only for IP addresses.
6381 34. The host_aton() function is supposed to be passed a string that is known
6382 to be a valid IP address. However, in the case of IPv6 addresses, it was
6383 not checking this. This is a hostage to fortune. Exim now panics and dies
6384 if the condition is not met. A case was found where this could be provoked
6385 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6386 components; fortuitously, this particular loophole had already been fixed
6387 by change 4.50/55 or 4.44/33 above.
6389 If there are any other similar loopholes, the new check in host_aton()
6390 itself should stop them being exploited. The report I received stated that
6391 data on the command line could provoke the exploit when Exim was running as
6392 exim, but did not say which command line option was involved. All I could
6393 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6394 running as the user.
6397 35. There was a buffer overflow vulnerability in the SPA authentication code
6398 (which came originally from the Samba project). I have added a test to the
6399 spa_base64_to_bits() function which I hope fixes it.
6402 36. The daemon start-up calls getloadavg() while still root for those OS that
6403 need the first call to be done as root, but it missed one case: when
6404 deliver_queue_load_max is set with deliver_drop_privilege. This is
6405 necessary for the benefit of the queue runner, because there is no re-exec
6406 when deliver_drop_privilege is set.
6408 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6409 was in use, was not putting the data itself into the right store pool;
6410 consequently, it could be overwritten for a subsequent message in the same
6411 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6414 38. Sometimes the final signoff response after QUIT could fail to get
6415 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6416 < 0 before doing a fflush(). This bug looks as though it goes back to the
6417 introduction of TLS in release 3.20, but "sometimes" must have been rare
6418 because the tests only now provoked it.
6424 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6425 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6426 Now it returns an integer. A number of calls were still expecting a T/F
6427 return. Fortuitously, in all cases, the tests worked in OK situations,
6428 which is the norm. However, things would have gone wrong on any write
6429 failures on the smtp file descriptor. This function is used when sending
6430 messages over SMTP and also when doing verify callouts.
6432 2. When Exim is called to do synchronous delivery of a locally submitted
6433 message (the -odf or -odi options), it no longer closes stderr before doing
6436 3. Implemented the mua_wrapper option.
6438 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6440 5. Implemented the functions header_remove(), header_testname(),
6441 header_add_at_position(), and receive_remove_recipient(), and exported them
6444 6. If an ACL "warn" statement specified the addition of headers, Exim already
6445 inserted X-ACL-Warn: at the start if there was no header name. However, it
6446 was not making this test for the second and subsequent header lines if
6447 there were newlines in the string. This meant that an invalid header could
6448 be inserted if Exim was badly configured.
6450 7. Allow an ACL "warn" statement to add header lines at the start or after all
6451 the Received: headers, as well as at the end.
6453 8. Added the rcpt_4xx retry error code.
6455 9. Added postmaster_mailfrom=xxx to callout verification option.
6457 10. Added mailfrom=xxxx to the callout verification option, for verify=
6460 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6461 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6463 12. Inserted some casts to stop certain compilers warning when using pointer
6464 differences as field lengths or precisions in printf-type calls (mostly
6465 affecting debugging statements).
6467 13. Added optional readline() support for -be (dynamically loaded).
6469 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6470 same clock tick as a message's arrival, so that its received time was the
6471 same as the "first fail" time on the retry record, and that message
6472 remained on the queue past the ultimate address timeout, every queue runner
6473 would try a delivery (because it was past the ultimate address timeout) but
6474 after another failure, the ultimate address timeout, which should have then
6475 bounced the address, did not kick in. This was a "< instead of <=" error;
6476 in most cases the first failure would have been in the next clock tick
6477 after the received time, and all would be well.
6479 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6480 being recognized when the domain list was tested by the match_domain
6481 condition in an expansion string.
6483 16. Added the ${str2b64: operator.
6485 17. Exim was always calling setrlimit() to set a large limit for the number of
6486 processes, without checking whether the existing limit was already
6487 adequate. (It did check for the limit on file descriptors.) Furthermore,
6488 errors from getrlimit() and setrlimit() were being ignored. Now they are
6489 logged to the main and panic logs, but Exim does carry on, to try to do its
6490 job under whatever limits there are.
6492 18. Imported PCRE 5.0.
6494 19. Trivial typo in log message " temporarily refused connection" (the leading
6497 20. If the log selector return_path_on_delivery was set and an address was
6498 redirected to /dev/null, the delivery process crashed because it assumed
6499 that a return path would always be set for a "successful" delivery. In this
6500 case, the whole delivery is bypassed as an optimization, and therefore no
6503 21. Internal re-arrangement: the function for sending a challenge and reading
6504 a response while authentication was assuming a zero-terminated challenge
6505 string. It's now changed to take a pointer and a length, to allow for
6506 binary data in such strings.
6508 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6510 23. Exim was not respecting finduser_retries when seeking the login of the
6511 uid under which it was called; it was always trying 10 times. (The default
6512 setting of finduser_retries is zero.) Also, it was sleeping after the final
6513 failure, which is pointless.
6515 24. Implemented tls_on_connect_ports.
6517 25. Implemented acl_smtp_predata.
6519 26. If the domain in control=submission is set empty, Exim assumes that the
6520 authenticated id is a complete email address when it generates From: or
6521 Sender: header lines.
6523 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6524 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6525 chown and chgrp in /bin and hostname in /usr/bin.
6527 28. Exim was keeping the "process log" file open after each use, just as it
6528 does for the main log. This opens the possibility of it remaining open for
6529 long periods when the USR1 signal hits a daemon. Occasional processlog
6530 errors were reported, that could have been caused by this. Anyway, it seems
6531 much more sensible not to leave this file open at all, so that is what now
6534 29. The long-running daemon process does not normally write to the log once it
6535 has entered its main loop, and it closes the log before doing so. This is
6536 so that log files can straightforwardly be renamed and moved. However,
6537 there are a couple of unusual error situations where the daemon does write
6538 log entries, and I had neglected to close the log afterwards.
6540 30. The text of an SMTP error response that was received during a remote
6541 delivery was being truncated at 512 bytes. This is too short for some of
6542 the long messages that one sometimes sees. I've increased the limit to
6545 31. It is now possible to make retry rules that apply only when a message has a
6546 specific sender, in particular, an empty sender.
6548 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6549 it possible to be selective about when SMTP synchronization is enforced.
6551 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6553 32. Implemented hosts_connection_nolog.
6555 33. Added an ACL for QUIT.
6557 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6560 35. Added mailbox_size and mailbox_filecount to appendfile.
6562 36. Added control = no_multiline_responses to ACLs.
6564 37. There was a bug in the logic of the code that waits for the clock to tick
6565 in the case where the clock went backwards by a substantial amount such
6566 that the microsecond fraction of "now" was more than the microsecond
6567 fraction of "then" (but the whole seconds number was less).
6569 38. Added support for the libradius Radius client library this is found on
6570 FreeBSD (previously only the radiusclient library was supported).
6576 1. When certain lookups returned multiple values in the form name=value, the
6577 quoting of the values was not always being done properly. Specifically:
6578 (a) If the value started with a double quote, but contained no whitespace,
6580 (b) If the value contained whitespace other than a space character (i.e.
6581 tabs or newlines or carriage returns) it was not quoted.
6582 This fix has been applied to the mysql and pgsql lookups by writing a
6583 separate quoting function and calling it from the lookup code. The fix
6584 should probably also be applied to nisplus, ibase and oracle lookups, but
6585 since I cannot test any of those, I have not disturbed their existing code.
6587 2. A hit in the callout cache for a specific address caused a log line with no
6588 reason for rejecting RCPT. Now it says "Previous (cached) callout
6589 verification failure".
6591 3. There was an off-by-one bug in the queryprogram router. An over-long
6592 return line was truncated at 256 instead of 255 characters, thereby
6593 overflowing its buffer with the terminating zero. As well as fixing this, I
6594 have increased the buffer size to 1024 (and made a note to document this).
6596 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6597 when Exim is waiting for an SMTP response from a remote server, Exim
6598 restarts its select() call on the socket, thereby resetting its timeout.
6599 This is not a problem when such interrupts are rare. Somebody set up a cron
6600 job to run exiwhat every 2 minutes, which is less than the normal select()
6601 timeout (5 or 10 minutes). This meant that the select() timeout never
6602 kicked in because it was always reset. I have fixed this by comparing the
6603 time when an interrupt arrives with the time at the start of the first call
6604 to select(). If more time than the timeout has elapsed, the interrupt is
6605 treated as a timeout.
6607 5. Some internal re-factoring in preparation for the addition of Sieve
6608 extensions (by MH). In particular, the "personal" test is moved to a
6609 separate function, and given an option for scanning Cc: and Bcc: (which is
6610 not set for Exim filters).
6612 6. When Exim created an email address using the login of the caller as the
6613 local part (e.g. when creating a From: or Sender: header line), it was not
6614 quoting the local part when it contained special characters such as @.
6616 7. Installed new OpenBSD configuration files.
6618 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6619 try to make them clearer.
6621 9. Callout options, other than the timeout value, were being ignored when
6622 verifying sender addresses in header lines. For example, when using
6624 verify = header_sender/callout=no_cache
6626 the cache was (incorrectly) being used.
6628 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6629 only the Cygwin environment.
6631 11. When return_path_on_delivery was set as a log selector, if different remote
6632 addresses in the same message used different return paths and parallel
6633 remote delivery occurred, the wrong values would sometimes be logged.
6634 (Whenever a remote delivery process finished, the return path value from
6635 the most recently started remote delivery process was logged.)
6637 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6638 lines when AUTH and/or TLS are in use. This is the "received protocol"
6639 field. Exim used to use "asmtp" for authenticated SMTP, without any
6640 indication (in the protocol name) for TLS use. Now it follows the RFC and
6641 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6642 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6643 names appear in log lines as well as in Received: header lines.
6645 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6646 extensions, and comparison tests, and to fix some bugs.
6648 14. Changes to the "personal" filter test:
6650 (1) The test was buggy in that it was just doing the equivalent of
6651 "contains" tests on header lines. For example, if a user's address was
6652 anne@some.where, the "personal" test would incorrectly be true for
6654 To: susanne@some.where
6656 This test is now done by extracting each address from the header in turn,
6657 and checking the entire address. Other tests that are part of "personal"
6658 are now done using regular expressions (for example, to check local parts
6659 of addresses in From: header lines).
6661 (2) The list of non-personal local parts in From: addresses has been
6662 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6663 taken from the Sieve specification recommendations.
6665 (3) If the message contains any header line starting with "List-" it is
6666 treated as non-personal.
6668 (4) The test for "circular" in the Subject: header line has been removed
6669 because it now seems ill-conceived.
6671 15. Minor typos in src/EDITME comments corrected.
6673 16. Installed latest exipick from John Jetmore.
6675 17. If headers_add on a router specified a text string that was too long for
6676 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6677 of string_sprintf() is now avoided.
6679 18. $message_body_size was not set (it was always zero) when running the DATA
6680 ACL and the local_scan() function.
6682 19. For the "mail" command in an Exim filter, no default was being set for
6683 the once_repeat time, causing a random time value to be used if "once" was
6684 specified. (If the value happened to be <= 0, no repeat happened.) The
6685 default is now 0s, meaning "never repeat". The "vacation" command was OK
6686 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6687 (I found it when inspecting the code).
6689 20. There is now an overall timeout for performing a callout verification. It
6690 defaults to 4 times the callout timeout, which applies to individual SMTP
6691 commands during the callout. The overall timeout applies when there is more
6692 than one host that can be tried. The timeout is checked before trying the
6693 next host. This prevents very long delays if there are a large number of
6694 hosts and all are timing out (e.g. when the network connections are timing
6695 out). The value of the overall timeout can be changed by specifying an
6696 additional sub-option for "callout", called "maxwait". For example:
6698 verify = sender/callout=5s,maxwait=20s
6700 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6701 to the end before writing, but this should make it even safer).
6703 22. Exim was forgetting that it had advertised PIPELINING for the second and
6704 subsequent messages on an SMTP connection. It was also not resetting its
6705 memory on STARTTLS and an internal HELO.
6707 23. When Exim logs an SMTP synchronization error within a session, it now
6708 records whether PIPELINING has been advertised or not.
6710 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6711 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6712 rather than long int.
6714 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6716 26. Added the never_mail option to autoreply.
6722 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6723 crash if the getsockname() call failed; this can happen if a connection is
6724 closed very soon after it is established. The problem was simply in the
6725 order in which certain operations were done, causing Exim to try to write
6726 to the SMTP stream before it had set up the file descriptor. The bug has
6727 been fixed by making things happen in the correct order.
6733 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6734 before the connection was closed, thus losing the rejection response.
6736 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6737 some early Solaris releases, but causes trouble in current releases where
6738 socklen_t is defined.
6740 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6743 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6746 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6747 that releasing the top of it at the end releases what was used for sub-
6748 expansions (unless the block got too big). However, discard this block if
6749 the first thing is a variable or header, so that we can use its block when
6750 it is dynamic (useful for very large $message_headers, for example).
6752 6. Lookups now cache *every* query, not just the most recent. A new, separate
6753 store pool is used for this. It can be recovered when all lookup caches are
6754 flushed. Lookups now release memory at the end of their result strings.
6755 This has involved some general refactoring of the lookup sources.
6757 7. Some code has been added to the store_xxx() functions to reduce the amount
6758 of flapping under certain conditions.
6760 8. log_incoming_interface used to affect only the <= reception log lines. Now
6761 it causes the local interface and port to be added to several more SMTP log
6762 lines, for example "SMTP connection from", and rejection lines.
6764 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6766 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6768 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6769 could be overwritten at the end of the current message (or the start of a
6770 new message if it was set in a HELO ACL). The value is now preserved for
6771 the duration of the SMTP connection.
6773 12. If a transport had a headers_rewrite setting, and a matching header line
6774 contained an unqualified address, that address was qualified, even if it
6775 did not match any rewriting rules. The underlying bug was that the values
6776 of the flags that permit the existence of unqualified sender and recipient
6777 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6778 non-local messages, and by -bnq for local messages) were not being
6779 preserved with the message after it was received.
6781 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6782 "next input=" as part of the text comprising the host identity instead of
6783 the correct text. The code was using the same buffer for two different
6784 strings. However, depending on which order the printing function evaluated
6785 its arguments, the bug did not always show up. Under Linux, for example, my
6786 test suite worked just fine.
6788 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6789 doesn't work with some older versions of Perl. It has been changed to "my",
6790 which in any case is probably the better facility to use.
6792 15. A really picky compiler found some instances of statements for creating
6793 error messages that either had too many or two few arguments for the format
6796 16. The size of the buffer for calls to the DNS resolver has been increased
6797 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6798 for addresses that have a lot of PTR records. This alleviates a problem; it
6799 does not fully solve it.
6801 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6802 buffer now truncates the list and logs the incident, which is the same
6803 action as happens when Exim is looking up a host name and its aliases.
6804 Previously in this situation something unpredictable would happen;
6805 sometimes it was "internal error: store_reset failed".
6807 18. If a server dropped the connection unexpectedly when an Exim client was
6808 using GnuTLS and trying to read a response, the client delivery process
6809 crashed while trying to generate an error log message.
6811 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6812 string, for example:
6814 warn message = H1: something\nH2: something
6816 the text was added as a single header line from Exim's point of view
6817 though it ended up OK in the delivered message. However, searching for the
6818 second and subsequent header lines using $h_h2: did not work. This has been
6819 fixed. Similarly, if a system filter added multiple headers in this way,
6820 the routers could not see them.
6822 20. Expanded the error message when iplsearch is called with an invalid key to
6823 suggest using net-iplsearch in a host list.
6825 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6826 ACLs are no longer actually imposed (and a message to that effect is
6829 22. If a "gecos" field in a passwd entry contained escaped characters, in
6830 particular, if it contained a \" sequence, Exim got it wrong when building
6831 a From: or a Sender: header from that name. A second bug also caused
6832 incorrect handling when an unquoted " was present following a character
6833 that needed quoting.
6835 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6836 was not being matched caselessly.
6838 24. Arranged for all hyphens in the exim.8 source to be escaped with
6841 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6842 properly. Recipient callout cache records were still being keyed to include
6843 the sender, even when use_sender was set false. This led to far more
6844 callouts that were necessary. The sender is no longer included in the key
6845 when use_sender is false.
6847 26. Added "control = submission" modifier to ACLs.
6849 27. Added the ${base62d: operator to decode base 62 numbers.
6851 28. dnsdb lookups can now access SRV records.
6853 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6854 the configuration file.
6856 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6857 (-v) mode. This makes the output for a verbose queue run more intelligible.
6859 31. Added a use_postmaster feature to recipient callouts.
6861 32. Added the $body_zerocount variable, containing the number of binary zero
6862 bytes in the message body.
6864 33. The time of last modification of the "new" subdirectory is now used as the
6865 "mailbox time last read" when there is a quota error for a maildir
6868 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6870 35. Added +ignore_unknown as a special item in host lists.
6872 36. Code for decoding IPv6 addresses in host lists is now included, even if
6873 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6874 address was recognized as an IP address, but was then not correctly decoded
6875 into binary, causing unexpected and incorrect effects when compared with
6882 1. Very minor rewording of debugging text in manualroute to say "list of
6883 hosts" instead of "hostlist".
6885 2. If verify=header_syntax was set, and a header line with an unqualified
6886 address (no domain) and a large number of spaces between the end of the
6887 name and the colon was received, the reception process suffered a buffer
6888 overflow, and (when I tested it) crashed. This was caused by some obsolete
6889 code that should have been removed. The fix is to remove it!
6891 3. When running in the test harness, delay a bit after writing a bounce
6892 message to get a bit more predictability in the log output.
6894 4. Added a call to search_tidyup() just before forking a reception process. In
6895 theory, someone could use a lookup in the expansion of smtp_accept_max_
6896 per_host which, without the tidyup, could leave open a database connection.
6898 5. Added the variables $recipient_data and $sender_data which get set from a
6899 lookup success in an ACL "recipients" or "senders" condition, or a router
6900 "senders" option, similar to $domain_data and $local_part_data.
6902 6. Moved the writing of debug_print from before to after the "senders" test
6905 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6906 problems for message scanning, either using a data ACL, or using
6907 local_scan() because the Received: header was not generated till after they
6908 were called (in order to set the time as the time of reception completion).
6909 I have revised the way this works. The header is now generated after the
6910 body is received, but before the ACL or local_scan() are called. After they
6911 are run, the timestamp in the header is updated.
6917 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6918 before starting a queue runner without re-exec. This happened only when
6919 deliver_drop_privilege was set or when the Exim user was set to root. The
6920 effect of the bug was that timeouts during subsequent deliveries caused
6921 crashes instead of being properly handled. The handler is now left at its
6922 default (and expected) setting.
6924 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6925 message, again when deliver_drop_privilege is set or Exim is run as root.
6926 The bug described in (1) was not present in this case, but the tidying up
6927 of the other signals was missing. I have made the two cases consistent.
6929 3. The ignore_target_hosts setting on a manualroute router was being ignored
6930 for hosts that were looked up using the /MX notation.
6932 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6935 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6936 operated on the sender address. After changing the $sender_address to <>
6937 for the sender address verify, Exim was re-instated it as the original
6938 (before rewriting) address, but remembering that it had rewritten it, so it
6939 wasn't rewriting it again. This bug also had the effect of breaking the
6940 sender address verification caching when the sender address was rewritten.
6942 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6943 This has been changed so that if the ip literal address matches
6944 ignore_target_hosts, the router declines.
6946 7. Added expansion conditions match_domain, match_address, and match_local_
6947 part (NOT match_host).
6949 8. The placeholder for the Received: header didn't have a length field set.
6951 9. Added code to Exim itself and to exim_lock to test for a specific race
6952 condition that could lead to file corruption when using MBX delivery. The
6953 issue is with the lockfile that is created in /tmp. If this file is removed
6954 after a process has opened it but before that process has acquired a lock,
6955 there is the potential for a second process to recreate the file and also
6956 acquire a lock. This could lead to two Exim processes writing to the file
6957 at the same time. The added code performs the same test as UW imapd; it
6958 checks after acquiring the lock that its file descriptor still refers to
6959 the same named file.
6961 10. The buffer for building added header lines was of fixed size, 8192 bytes.
6962 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
6965 11. Added the smtp_active_hostname option. If used, this will typically be made
6966 to depend on the incoming interface address. Because $interface_address is
6967 not set up until the daemon has forked a reception process, error responses
6968 that can happen earlier (such as "too many connections") no longer contain
6971 12. If an expansion in a condition on a "warn" statement fails because a lookup
6972 defers, the "warn" statement is abandoned, and the next ACL statement is
6973 processed. Previously this caused the whole ACL to be aborted.
6975 13. Added the iplsearch lookup type.
6977 14. Added ident_timeout as a log selector.
6979 15. Added tls_certificate_verified as a log selector.
6981 16. Added a global option tls_require_ciphers (compare the smtp transport
6982 option of the same name). This controls incoming TLS connections.
6984 17. I finally figured out how to make tls_require_ciphers do a similar thing
6985 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
6986 before starting the TLS session.
6988 18. Tabs are now shown as \t in -bP output.
6990 19. If the log selector return_path_on_delivery was set, Exim crashed when
6991 bouncing a message because it had too many Received: header lines.
6993 20. If two routers both had headers_remove settings, and the first one included
6994 a superfluous trailing colon, the final name in the first list and the
6995 first name in the second list were incorrectly joined into one item (with a
6996 colon in the middle).
7002 1. Added -C and -D options to the exinext utility, mainly to make it easier
7003 to include in the automated testing, but these could be helpful when
7004 multiple configurations are in use.
7006 2. The exinext utility was not formatting the output nicely when there was
7007 an alternate port involved in the retry record key, nor when there was a
7008 message id as well (for retries that were specific to a specific message
7009 and a specific host). It was also confused by IPv6 addresses, because of
7010 the additional colons they contain. I have fixed the IPv4 problem, and
7011 patched it up to do a reasonable job for IPv6.
7013 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7014 delivery, the log line now contains "pipelined" if PIPELINING was used.
7016 4. An SMTP transport process used to panic and die if the bind() call to set
7017 an explicit outgoing interface failed. This has been changed; it is now
7018 treated in the same way as a connect() failure.
7020 5. A reference to $sender_host_name in the part of a conditional expansion
7021 that was being skipped was still causing a DNS lookup. This no longer
7024 6. The def: expansion condition was not recognizing references to header lines
7025 that used bh_ and bheader_.
7027 7. Added the _cache feature to named lists.
7029 8. The code for checking quota_filecount in the appendfile transport was
7030 allowing one more file than it should have been.
7032 9. For compatibility with Sendmail, the command line option
7040 and sets the incoming protocol and host name (for trusted callers). The
7041 host name and its colon can be omitted when only the protocol is to be set.
7042 Note the Exim already has two private options, -pd and -ps, that refer to
7043 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7044 "s", but I don't think that's a major issue.
7046 10. A number of refactoring changes to the code, none of which should affect
7049 (a) The number of logging options was getting close to filling up the
7050 32-bit word that was used as a bit map. I have split them into two classes:
7051 those that are passed in the argument to log_write(), and those that are
7052 only ever tested independently outside of that function. These are now in
7053 separate 32-bit words, so there is plenty of room for expansion again.
7054 There is no change in the user interface or the logging behaviour.
7056 (b) When building, for example, log lines, the code previously used a
7057 macro that called string_cat() twice, in order to add two strings. This is
7058 not really sufficiently general. Furthermore, there was one instance where
7059 it was actually wrong because one of the argument was used twice, and in
7060 one call a function was used. (As it happened, calling the function twice
7061 did not affect the overall behaviour.) The macro has been replaced by a
7062 function that can join an arbitrary number of extra strings onto a growing
7065 (c) The code for expansion conditions now uses a table and a binary chop
7066 instead of a serial search (which was left over from when there were very
7067 few conditions). Also, it now recognizes conditions like "pam" even when
7068 the relevant support is not compiled in: a suitably worded error message is
7069 given if an attempt is made to use such a condition.
7071 11. Added ${time_interval:xxxxx}.
7073 12. A bug was causing one of the ddress fields not to be passed back correctly
7074 from remote delivery subprocesses. The field in question was not being
7075 subsequently used, so this caused to problems in practice.
7077 13. Added new log selectors queue_time and deliver_time.
7079 14. Might have fixed a bug in maildirsizefile handling that threw up
7080 "unexpected character" debug warnings, and recalculated the data
7081 unnecessarily. In any case, I expanded the warning message to give more
7084 15. Added the message "Restricted characters in address" to the statements in
7085 the default ACL that block characters like @ and % in local parts.
7087 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7088 Three changes have been made:
7090 (a) There was a serious bug; a negative response to MAIL caused the whole
7091 recipient domain to be cached as invalid, thereby blocking all messages
7092 to all local parts at the same domain, from all senders. This bug has
7093 been fixed. The domain is no longer cached after a negative response to
7094 MAIL if the sender used is not empty.
7096 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7099 (c) A new callout option, "use_sender" has been added for people who want
7100 the modified behaviour.
7106 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7109 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7110 indeed breaks things for older releases.
7112 3. Added additional logging to the case where there is a problem reading data
7113 from a filter that is running in a subprocess using a pipe, in order to
7114 try to track down a specific problem.
7116 4. Testing facility fudge: when running in the test harness and attempting
7117 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7118 getting "No route to host". Convert this to a timeout.
7120 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7123 6. Some OS don't have socklen_t but use size_t instead. This affects the
7124 fifth argument of getsockopt() amongst other things. This is now
7125 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7126 can be set for individual OS. I have set it for SunOS5, OSF1, and
7127 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7128 some earlier ones do not.
7130 7. Change 4.30/15 was not doing the test caselessly.
7132 8. The standard form for an IPv6 address literal was being rejected by address
7133 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7134 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7135 this, as well as the form without the "IPv6" on the front (but only when
7136 address literals are enabled, of course).
7138 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7140 10. Exim crashed if a message with an empty sender address specified by -f
7141 encountered a router with an errors_to setting. This could be provoked only
7142 by a command such as
7146 where an empty string was supplied; "<>" did not hit this bug.
7148 11. Installed PCRE release 4.5.
7150 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7151 remained set. It is now erased.
7153 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7154 times from message ids (which are base 36 rather than the normal 62).
7156 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7157 were being counted as actual protocol errors, and logged if the log
7158 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7159 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7160 and DATA following a set of rejected RCPTs do not count as protocol errors.
7161 In other words, Exim assumes they were pipelined, though this may not
7162 actually be the case. Of course, in all cases the client gets an
7163 appropriate error code.
7165 15. If a lookup fails in an ACL condition, a message about the failure may
7166 be available; it is used if testing the ACL cannot continue, because most
7167 such messages specify what the cause of the deferral is. However, some
7168 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7169 that caused an old message to be retained and used if a later statement
7170 caused a defer, replacing the real cause of the deferral.
7172 16. If an IP address had so many PTR records that the DNS lookup buffer
7173 was not large enough to hold them, Exim could crash while trying to process
7174 the truncated data. It now detects and logs this case.
7176 17. Further to 4.21/58, another change has been made: if (and only if) the
7177 first line of a message (the first header line) ends with CRLF, a bare LF
7178 in a subsequent header line has a space inserted after it, so as not to
7179 terminate the header.
7181 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7182 unnecessarily, used atoi() instead of strtol(), and didn't check the
7183 termination when getting file sizes from file names by regex.
7185 19. Completely re-implemented the support for maildirsize files, in the light
7186 of a number of problems with the previous contributed implementation
7187 (4.30/29). In particular:
7189 . If the quota is zero, the maildirsize file is maintained, but no quota is
7192 . If the maildir directory does not exist, it is created before any attempt
7193 to write a maildirsize file.
7195 . The quota value in the file is just a cache; if the quota is changed in
7196 the transport, the new value overrides.
7198 . A regular expression is available for excluding directories from the
7201 20. The autoreply transport checks the characters in options that define the
7202 message's headers; it allows continued headers, but it was checking with
7203 isspace() after an embedded newline instead of explicitly looking for a
7206 21. If all the "regular" hosts to which an address was routed had passed their
7207 expiry times, and had not reached their retry times, the address was
7208 bounced, even if fallback hosts were defined. Now Exim should go on to try
7211 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7212 equivalent code in the SMTP transport. Some hosts send humungous responses
7213 to HELO/EHLO, more than 1024 it seems.
7215 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7216 gives compiler warnings in some environments. I've now done it "properly",
7219 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7220 (because of problems with the built-in one) was declared to return uschar *
7221 instead of char *, causing compiler failure.
7223 25. Fixed a file descriptor leak when processing alias/forward files.
7225 26. Fixed a minor format string issue in dbfn.c.
7227 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7229 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7230 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7233 29. When a sender address is verified, it is cached, to save repeating the test
7234 when there is more than one recipient in a message. However, when the
7235 verification involves a callout, it is possible for different callout
7236 options to be set for different recipients. It is too complicated to keep
7237 track of this in the cache, so now Exim always runs a verification when a
7238 callout is required, relying on the callout cache for the optimization.
7239 The overhead is duplication of the address routing, but this should not be
7242 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7243 to be verified with callout=postmaster, and the main callout worked but the
7244 postmaster check failed, the verification correctly failed. However, if a
7245 subsequent RCPT command asked for sender verification *without* the
7246 postmaster check, incorrect caching caused this verification also to fail,
7249 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7250 it was not caching the DNS options (qualify_single, search_parents) that
7251 were used when the lookup failed. A subsequent lookup with different
7252 options therefore always gave the same answer, though there were cases
7253 where it should not have. (Example: a "domains = !$mx_any" option on a
7254 dnslookup router: the "domains" option is always processed without any
7255 widening, but the router might have qualify_single set.) Now Exim uses the
7256 cached value only when the same options are set.
7258 32. Added John Jetmore's "exipick" utility to the distribution.
7260 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7261 than a timeout (e.g. a certificate is required, and is not provided), an
7262 Exim server now closes the connection immediately. Previously it waited for
7263 the client to close - but if the client is SSL, it seems that they each
7264 wait for each other, leading to a delay before one of them times out.
7266 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7267 maintained 0.8.x compatibility because I don't think many are using it, and
7268 it is clearly obsolete.
7270 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7273 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7274 client certificate was expired. A simple patch fixes this, though I don't
7275 understand the full logic of why the verify callback is called multiple
7278 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7279 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7280 which causes problems with some clients (such as the Certicom SSL Plus
7281 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7282 disables the coutermeasure allowing Eudora to connect."
7284 38. Exim was not checking that a write() to a log file succeeded. This could
7285 lead to Bad Things if a log got too big, in particular if it hit a file
7286 size limit. Exim now panics and dies if it cannot write to a log file, just
7287 as it does if it cannot open a log file.
7289 39. Modified OS/Makefile-Linux so that it now contains
7291 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7293 The two -D definitions ensure that Exim is compiled with large file
7294 support, which makes it possible to handle log files that are bigger than
7297 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7298 instance) a domain was checked against a named list that involved a lookup,
7299 causing $domain_data to be set, then another domain was checked against the
7300 same list, then the first domain was re-checked, the value of $domain_data
7301 after the final check could be wrong. In particular, if the second check
7302 failed, it could be set empty. This bug probably also applied to
7305 41. The strip_trailing_dot option was not being applied to the address given
7306 with the -f command-line option.
7308 42. The code for reading a message's header from the spool was incrementing
7309 $received_count, but never initializing it. This meant that the value was
7310 incorrect (doubled) while delivering a message in the same process in which
7311 it was received. In the most common configuration of Exim, this never
7312 happens - a fresh exec is done - but it can happen when
7313 deliver_drop_privilege is set.
7315 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7316 it now includes up to 150 characters of the unexpected data in the log
7319 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7320 and building data strings. The size of both of these buffers was 10 000
7321 bytes - far larger than anybody would *ever* want, thought I. Needless to
7322 say, somebody hit the limit. I have increased the maximum line length to
7323 20 000 and the maximum data length of concatenated lines to 100 000. I have
7324 also fixed two bugs, because there was no checking on these buffers. Tsk,
7325 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7326 buffer is too small.
7328 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7329 lsearch lookups. Now it does.
7331 46. When parsing a route_list item in a manualroute router, a fixed-length
7332 buffer was used for the list of hosts. I made this 1024 bytes long,
7333 thinking that nobody would ever have a list of hosts that long. Wrong.
7334 Somebody had a whole pile of complicated expansion conditions, and the
7335 string was silently truncated, leading to an expansion error. It turns out
7336 that it is easier to change to an unlimited length (owing to other changes
7337 that have happened since this code was originally written) than to build
7338 structure for giving a limitation error. The length of the item that
7339 expands into the list of hosts is now unlimited.
7341 47. The lsearch lookup could not handle data where the length of text line was
7342 more than 4095 characters. Such lines were truncated, leading to shortened
7343 data being returned. It should now handle lines of any length.
7345 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7346 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7349 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7350 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7351 no longer happens in comment lines. A list of replacements is now placed
7352 at the head of all of the source files, except those whose only change is
7353 to replace PERL_COMMAND in the very first #! line.
7355 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7356 messages on the queue, with a bottom-up merge sort, using code contributed
7357 by Michael Haardt. This should make operations like -bp somewhat faster on
7358 large queues. It won't affect queue runners, except when queue_run_in_order
7361 51. Installed eximstats 1.31 in the distribution.
7363 52. Added support for SRV lookups to the dnslookup router.
7365 53. If an ACL referred to $message_body or $message_body_end, the value was not
7366 reset for any messages that followed in the same SMTP session.
7368 54. The store-handling optimization for building very long strings was not
7369 differentiating between the different store pools. I don't think this
7370 actually made any difference in practice, but I've tidied it.
7372 55. While running the routers to verify a sender address, $sender_address
7373 was still set to the sender address. This is wrong, because when routing to
7374 send a bounce to the sender, it would be empty. Therefore, I have changed
7375 it so that, while verifying a sender address, $sender_address is set to <>.
7376 (There is no change to what happens when verifying a recipient address.)
7378 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7379 target A or AAAA records (if not already returned) without resetting the
7380 qualify_single or search_parents options of the DNS resolver. These are
7381 inappropriate in this case because the targets of MX and SRV records must
7382 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7383 target that, when qualified, matched something in the local domain. These
7384 two options are now turned off when doing these lookups.
7386 57. It seems that at least some releases of Reiserfs (which does not have the
7387 concept of a fixed number of inodes) returns zero and not -1 for the
7388 number of available inodes. This interacted badly with check_spool_inodes,
7389 which assumed that -1 was the "no such thing" setting. What I have done is
7390 to check that the total number of inodes is greater than zero before doing
7391 the test of how many are available.
7393 58. When a "warn" ACL statement has a log_message modifier, the message is
7394 remembered, and not repeated. This is to avoid a lot of repetition when a
7395 message has many recipients that cause the same warning to be written.
7396 However, Exim was preserving the list of already written lines for an
7397 entire SMTP session, which doesn't seem right. The memory is now reset if a
7398 new message is started.
7400 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7401 debugging output unless log_rewrite was also set.
7403 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7404 of (int)(handle) when we know that handle contains (void *)(-1).
7406 61. The Exim daemon panic-logs an error return when it closes the incoming
7407 connection. However "connection reset by peer" seems to be common, and
7408 isn't really an error worthy of noting specially, so that particular error
7411 62. When Exim is trying to find all the local interfaces, it used to panic and
7412 die if the ioctl to get the interface flags failed. However, it seems that
7413 on at least one OS (Solaris 9) it is possible to have an interface that is
7414 included in the list of interfaces, but for which you get a failure error
7415 for this call. This happens when the interface is not "plumbed" into a
7416 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7417 failure of the "get flags" call assumes that the interface is down.
7419 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7420 makes life easier for people who are doing arithmetic on fields extracted
7421 from dates, where you often get leading zeros that should not be
7422 interpreted as octal.
7424 64. Added qualify_domain to the redirect router, to override the global
7427 65. If a pathologically long header line contained very many addresses (the
7428 report of this problem mentioned 10 000) and each of them was rewritten,
7429 Exim could use up a very large amount of memory. (It kept on making new
7430 copies of the header line as it rewrote, and never released the old ones.)
7431 At the expense of a bit more processing, the header rewriting function has
7432 been changed so that it no longer eats memory in this way.
7434 66. The generation of the Received: header has been moved from the time that a
7435 message starts to be received, to the time that it finishes. The timestamp
7436 in the Received: header should now be very close to that of the <= log
7437 line. There are two side-effects of this change:
7439 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7440 logged header lines no longer include the local Received: line, because
7441 it has not yet been created. The same applies to a copy of the message
7442 that is returned to a non-SMTP sender when a message is rejected.
7444 (b) When a filter file is tested using -bf, no additional Received: header
7445 is added to the test message. After some thought, I decided that this
7448 This change does not affect the value of $received_for. It is still set
7449 after address rewriting, but before local_scan() is called.
7451 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7453 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7454 gave an unhelpful panic error message, and a defer error. I have managed to
7455 change this behaviour so that it now rejects any supplied certificate,
7456 which seems right, as the list of acceptable certificates is empty.
7458 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7459 gave an unhelpful defer error. I have not managed to make this reject any
7460 supplied certificates, but the error message it gives is "no certificate
7461 supplied", which is not helpful.
7463 70. exigrep's output now also includes lines that are not associated with any
7464 message, but which match the given pattern. Implemented by a patch from
7465 Martin Sluka, which also tidied up the Perl a bit.
7467 71. Recipient callout verification, like sender verification, was using <> in
7468 the MAIL FROM command. This isn't really the right thing, since the actual
7469 sender may affect whether the remote host accepts the recipient or not. I
7470 have changed it to use the actual sender in the callout; this means that
7471 the cache record is now keyed on a recipient/sender pair, not just the
7472 recipient address. There doesn't seem to be a real danger of callout loops,
7473 since a callout by the remote host to check the sender would use <>.
7474 [SEE ABOVE: changed after hitting problems.]
7476 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7477 temporary errors. However, in the case of such a code being given after
7478 the end of a data transmission (i.e. after ".") Exim was failing to write
7479 a retry record for the message. (Yes, there was some broken host that was
7480 actually sending 8xx at this point.)
7482 73. An unknown lookup type in a host list could cause Exim to panic-die when
7483 the list was checked. (An example that provoked this was putting <; in the
7484 middle of a list instead of at the start.) If this happened during a DATA
7485 ACL check, a -D file could be left lying around. This kind of configuration
7486 error no longer causes Exim to die; instead it causes a defer error. The
7487 incident is still logged to the main and panic logs.
7489 74. Buglet left over from Exim 3 conversion. The message "too many messages
7490 in one connection" was written to the rejectlog but not the mainlog, except
7491 when address rewriting (yes!) was being logged.
7493 75. Added write_rejectlog option.
7495 76. When a system filter was run not as root (that is, when system_filter_user
7496 was set), the values of the $n variables were not being returned to the
7497 main process; thus, they were not subsequently available in the $sn
7500 77. Added +return_path_on_delivery log selector.
7502 78. A connection timeout was being treated differently from recipients deferred
7503 when testing hosts_max_try with a message that was older than the host's
7504 retry timeout. (The host should not be counted, thus allowing all hosts to
7505 be tried at least once before bouncing.) This may have been the cause of an
7506 occasionally reported bug whereby a message would remain on the queue
7507 longer than the retry timeout, but would be bounced if a delivery was
7508 forced. I say "may" because I never totally pinned down the problem;
7509 setting up timeout/retry tests is difficult. See also the next item.
7511 79. The ultimate address timeout was not being applied to errors that involved
7512 a combination of host plus message (for example, a timeout on a MAIL
7513 command). When an address resolved to a number of possible hosts, and they
7514 were not all tried for each delivery (e.g. because of hosts_max_try), a
7515 message could remain on the queue longer than the retry timeout.
7517 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7520 81. Fixed an obscure SMTP outgoing bug which required at least the following
7521 conditions: (a) there was another message waiting for the same server;
7522 (b) the server returned 5xx to all RCPT commands in the first message so
7523 that the message was not completed; (c) the server dropped the connection
7524 or gave a negative response to the RSET that Exim sends to abort the
7525 transaction. The observed case was a dropped connection after DATA that had
7526 been sent in pipelining mode. That is, the server had advertised PIPELINING
7527 but was not implementing it correctly. The effect of the bug was incorrect
7528 behaviour, such as trying another host, and this could lead to a crash.
7534 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7535 and daemon.c were passed as pointers to ints; they should have been
7536 pointers to socklen_t variables (which are typically unsigned ints).
7538 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7541 3. Fixed a really odd bug that affected only the testing scheme; patching a
7542 certain fixed string in the binary changed the value of another string that
7543 happened to be identical to the end of the original first string.
7545 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7546 name", it returns that address as the IP address. On some operating
7547 systems (e.g. Solaris), it also passes back the IP address string as the
7548 "host name". However, on others (e.g. Linux), it passes back an empty
7549 string. Exim wasn't checking for this, and was changing the host name to an
7550 empty string, assuming it had been canonicalized.
7552 5. Although rare, it is permitted to have more than one PTR record for a given
7553 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7554 all the names associated with an address, because they do in Solaris.
7555 However, it seems that they do not in Linux for data that comes from the
7556 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7557 I found this out when I moved to a new Linux workstation and tried to run
7558 the Exim test suite.
7560 To get round this problem I have changed the code so that it now does its
7561 own call to the DNS to look up PTR records when searching for a host name.
7562 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7563 addresses that are only in /etc/hosts are still found.
7565 This behaviour is, however, controlled by an option called host_lookup_
7566 order, which defaults to "bydns:byaddr". If people want to use the other
7567 order, or indeed, just use one or the other means of lookup, they can
7568 specify it in this variable.
7570 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7571 some operating systems, this comes back from gethostbyaddr() as an empty
7572 string, and this is what Exim used to test for. However, it seems that in
7573 other systems, "." is yielded. Exim now tests for this case too.
7575 7. The values of check_spool_space and check_log_space are now held internally
7576 as a number of kilobytes instead of an absolute number of bytes. If a
7577 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7578 kilobyte. This means that much larger values can be stored.
7580 8. Exim monitor: an attempt to get the action menu when not actually pointing
7581 at a message produces an empty menu entitled "No message selected". This
7582 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7583 no entries in it ("Shell widget menu has zero width and/or height"). So I
7584 have added a single, blank menu entry in this case.
7586 9. Added ${quote_local_part.
7588 10. MIME decoding is now applied to the contents of Subject: header lines when
7591 11. Now that a reference to $sender_host_address automatically causes a reverse
7592 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7593 host lookup before query-style lookups in lists that might use this
7594 variable. This has therefore been abolished, and the "net-" prefix is no
7595 longer necessary for query-style lookups.
7597 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7598 have been a typo for LFLAGS, so it has been changed.
7600 13. The install script calls Exim with "-C /dev/null" in order to find the
7601 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7602 to be output. However, since Exim outputs its version number before the
7603 error, it didn't break the script. It just looked ugly. I fixed this by
7604 always allowing "-C /dev/null" if the caller is root.
7606 14. Ignore overlarge ACL variable number when reading spool file - insurance
7607 against a later release with more variables having written the file.
7609 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7610 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7611 this, as well as the form without the "IPv6" on the front.
7613 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7614 OS/Makefile-Darwin file.
7616 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7617 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7619 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7620 variable. However, some OS use ints for inodes. I've added cast to long int
7621 to get rid of the compiler warning.
7623 19. I had forgotten to lock out "/../" in configuration file names when
7624 ALT_CONFIG_PREFIX was set.
7626 20. Routers used for verification do not need to specify transports. However,
7627 if such a router generated a host list, and callout was configured, Exim
7628 crashed, because it could not find a port number from the (non-existent)
7629 transport. It now assumes port 25 in this circumstance.
7631 21. Added the -t option to exigrep.
7633 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7634 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7635 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7636 them is set, LOOKUP_LSEARCH is forced.
7638 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7640 24. Added sender and host information to the "rejected by local_scan()" log
7641 line; previously there was no indication of these.
7643 25. Added .include_if_exists.
7645 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7646 Linux. It turns out that not all file systems support this. Apparently some
7647 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7648 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7649 not supported on the file descriptor, is now ignored when Exim is trying to
7650 sync a directory. This applies only to Linux.
7652 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7654 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7655 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7658 29. Added support for maildirsize files from supplied patch (modified a bit).
7660 30. The use of :fail: followed by an empty string could lead Exim to respond to
7661 sender verification failures with (e.g.):
7663 550 Verification failed for <xxx>
7664 550 Sender verify failed
7666 where the first response line was missing the '-' that indicates it is not
7667 the final line of the response.
7669 31. The loop for finding the name of the user that called Exim had a hardwired
7670 limit of 10; it now uses the value of finduser_retries, which is used for
7671 all other user lookups.
7673 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7676 33. Exim was neglecting to zero errno before one call of strtol() when
7677 expanding a string and expecting an integer value. On some systems this
7678 resulted in spurious "integer overflow" errors. Also, it was casting the
7679 result into an int without checking.
7681 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7682 did not work. The code looks as if it has *never* worked, though it appears
7683 to have been documented since at least release 1.62. I have made it work.
7685 35. The "timeout_DNS" error in retry rules, also documented since at least
7686 1.62, also never worked. As it isn't clear exactly what this means, and
7687 clearly it isn't a major issue, I have abolished the feature by treating it
7688 as "timeout", and writing a warning to the main and panic logs.
7690 36. The display of retry rules for -brt wasn't always showing the error code
7693 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7694 timeout_connect_A, timeout_connect_MX.
7696 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7697 to the empty sender.
7699 39. The daemon was not analysing the content of -oX till after it had closed
7700 stderr and disconnected from the controlling terminal. This meant that any
7701 syntax errors were only noted on the panic log, and the return code from
7702 the command was 0. By re-arranging the code a little, I've made the
7703 decoding happen first, so such errors now appear on stderr, and the return
7704 code is 1. However, the actual setting up of the sockets still happens in
7705 the disconnected process, so errors there are still only recorded on the
7708 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7709 connections (as happens on some IP stacks) was logged at start up time as
7710 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7711 it from "IPv6 and IPv4", which means that two separate sockets are being
7714 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7715 says whether AF_INET or AF_INET6 was passed as an argument.
7717 42. Exiwhat output was messed up when time zones were included in log
7720 43. Exiwhat now gives more information about the daemon's listening ports,
7721 and whether -tls-on-connect was used.
7723 44. The "port" option of the smtp transport is now expanded.
7725 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7726 silently ignored. Now an error message is written to the main and panic
7729 46. There's a new ACL modifier called "logwrite" which writes to a log file
7730 as soon as it is encountered.
7732 47. Added $local_user_uid and $local_user_gid at routing time.
7734 48. Exim crashed when trying to verify a sender address that was being
7737 49. Exim was recognizing only a space character after ".include". It now also
7738 recognizes a tab character.
7740 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7741 extracting the relevant information from the specification. The man page no
7742 longer contains scrambled data for the -d option, and I've added a section
7743 at the front about calling Exim under different names.
7745 51. Added "extra_headers" argument to the "mail" command in filter files.
7747 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7750 53. Installed eximstats 1.29.
7752 54. Added transport_filter_timeout as a generic transport option.
7754 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7755 Cc: header lines. This was required by RFC 822, but it not required by RFC
7758 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7759 incoming messages that did not have them. Now it does so only if the
7760 message originates locally, that is, if there is no associated remote host
7761 address. When Resent- header lines are present, this applies to the Resent-
7762 lines rather than the non-Resent- lines.
7764 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7765 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7767 58. Messages for configuration errors now include the name of the main
7768 configuration file - useful now that there may be more than one file in a
7769 list (.included file names were always shown).
7771 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7772 for those rare installations that do not start the daemon as root or run it
7773 setuid root. I've cut out the call to initgroups() if the daemon is not
7776 60. The Exim user and group can now be bound into the binary as text strings
7777 that are looked up at the start of Exim's processing.
7779 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7781 62. Added $mailstore_basename variable.
7783 63. Installed patch to sieve.c from Michael Haardt.
7785 64. When Exim failed to open the panic log after failing to open the main log,
7786 the original message it was trying to log was written to stderr and debug
7787 output, but if they were not available (the usual case in production), it
7788 was lost. Now it is written to syslog before the two lines that record the
7789 failures to open the logs.
7791 65. Users' Exim filters run in subprocesses under the user's uid. It is
7792 possible for a "deliver" command or an alias in a "personal" command to
7793 provoke an address rewrite. If logging of address rewriting is configured,
7794 this fails because the process is not running as root or exim. There may be
7795 a better way of dealing with this, but for the moment (because 4.30 needs
7796 to be released), I have disabled address rewrite logging when running a
7797 filter in a non-root, non-exim process.
7803 1. The buildconfig auxiliary program wasn't quoting the value set for
7804 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7805 not defined. This bug was masked in 4.22 by the effect that was fixed in
7808 2. Some messages that were rejected after a message id was allocated were
7809 shown as "incomplete" by exigrep. It no longer does this for messages that
7810 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7812 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7813 have allow_domain_literals set, the ID did not get logged in the <= line.
7814 Domain literals are now always recognized in Message-ID: header lines.
7816 4. The first argument for a ${extract expansion item is the key name or field
7817 number. Leading and trailing spaces in this item were not being ignored,
7818 causing some misleading effects.
7820 5. When deliver_drop_privilege was set, single queue runner processes started
7821 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7822 same command in the process it spins off) were not dropping privilege.
7824 6. When the daemon running as "exim" started a queue runner, it always
7825 re-executed Exim in the spun-off process. This is a waste of effort when
7826 deliver_drop_privilege is set. The new process now just calls the
7827 queue-runner function directly.
7833 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7836 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7837 ignored. Though the use of -oP was forcing the writing of a pid file, it
7838 was always written to the default place.
7840 3. If the message "no IP address found for host xxxx" is generated during
7841 incoming verification, it is now followed by identification of the incoming
7842 connection (so you can more easily find what provoked it).
7844 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7846 5. Added some features to "harden" Exim a bit more against certain attacks:
7848 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7849 be put in Local/Makefile. This is like the never_users runtime option,
7850 but it cannot be overridden. The default setting is "root".
7852 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7853 prefix string with which any file named in a -C command line option
7856 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7857 is retained for -C and -D only if the caller of Exim is root. Without
7858 it, the exim user may also use -C and -D and retain privilege.
7860 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7861 command line option is disabled.
7863 6. Macro names set by the -D option must start with an upper case letter, just
7864 like macro names defined in the configuration file.
7866 7. Added "dereference=" facility to LDAP.
7868 8. Two instances of the typo "uknown" in the source files are fixed.
7870 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7871 the Configure-Makefile script screwed up while processing it.
7873 10. Incorporated PCRE 4.4.
7875 11. The SMTP synchronization check was not operating right at the start of an
7876 SMTP session. For example, it could not catch a HELO sent before the client
7877 waited for the greeting. There is now a check for outstanding input at the
7878 point when the greeting is written. Because of the duplex, asynchronous
7879 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7880 way, but not yet received, when the check is performed.
7882 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7883 on TCP/IP sockets, because this apparently causes some broken clients to
7886 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7887 unchanged) from the Cygwin maintainer.
7889 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7890 was supported instead of testing for mbx. Effectively a typo.
7892 15. The spa authenticator server code was not checking that the input it
7893 received was valid base64.
7895 16. The debug output line for the "set" modifier in ACLs was not showing the
7896 name of the variable that was being set.
7898 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7900 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7901 lookup on its own. Something else was needed to trigger the lookup. For
7902 example, a match in host_lookup or the need for a host name in a host list.
7903 Now, if $sender_host_name is referenced and the host name has not yet been
7904 looked up, a lookup is performed. If the lookup fails, the variable remains
7905 empty, and $host_lookup_failed is set to "1".
7907 19. Added "eqi" as a case-independent comparison operator.
7909 20. The saslauthd authentication condition could segfault if neither service
7910 nor realm was specified.
7912 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7913 error message that was logged was misleading, and incoming SMTP
7914 connections were dropped. The message is now more accurate, and temporary
7915 errors are given to SMTP connections.
7917 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7918 (except RSET) with a 421 error, until QUIT is received. However, it was
7919 failing to send a response to QUIT.
7921 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7922 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7923 ACL is now run after the helo_try_verify_hosts code.
7925 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7926 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7927 case-independent, so other case variants are also recognized). Apparently
7928 some systems use these upper case variants.
7930 25. If more than two messages were waiting for the same host, and a transport
7931 filter was specified for the transport, Exim sent two messages over the
7932 same TCP/IP connection, and then failed with "socket operation on non-
7933 socket" when it tried to send the third.
7935 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7937 27. The extern definition of crypt16() in expand.c was not being excluded when
7938 the OS had its own crypt16() function.
7940 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7941 as a preferred synonym for return_size_limit, both as an option and as an
7944 29. Added LIBS=-liconv to OS/Makefile-OSF1.
7946 30. Changed the default configuration ACL to relax the local part checking rule
7947 for addresses that are not in any local domains. For these addresses,
7948 slashes and pipe symbols are allowed within local parts, but the sequence
7949 /../ is explicitly forbidden.
7951 31. SPA server authentication was not clearing the challenge buffer before
7954 32. log_message in a "warn" ACL statement was writing to the reject log as
7955 well as to the main log, which contradicts the documentation and doesn't
7956 seem right (because no rejection is happening). So I have stopped it.
7958 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
7959 However, I am unable to do any testing of this.
7961 34. Fixed an infelicity in the appendfile transport. When checking directories
7962 for a mailbox, to see if any needed to be created, it was accidentally
7963 using path names with one or more superfluous leading slashes; tracing
7964 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
7966 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
7967 contents are added to the log line that is written for every discarded
7968 recipient. (Previously a log_message setting was ignored.)
7970 36. The ${quote: operator now quotes the string if it is empty.
7972 37. The install script runs exim in order to find its version number. If for
7973 some reason other than non-existence or emptiness, which it checks, it
7974 could not run './exim', it was installing it with an empty version number,
7975 i.e. as "exim-". This error state is now caught, and the installation is
7978 38. An argument was missing from the function that creates an error message
7979 when Exim fails to connect to the socket for saslauthd authentication.
7980 This could cause Exim to crash, or give a corrupted message.
7982 39. Added isip, isip4, and isip6 to ${if conditions.
7984 40. The ACL variables $acl_xx are now saved with the message, and can be
7985 accessed later in routers, transports, and filters.
7987 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
7988 strings in the file are not string-expanded.
7990 42. If a MAIL command specified a SIZE value that was too large to fit into an
7991 int variable, the check against message_size_limit failed. Such values are
7992 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
7993 day this will have to be increased, but I don't think I want to be around
7994 when emails are that large.
8001 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8002 iconv() is not standard in FreeBSD.
8004 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8005 IPv6 enabled. The observed symptom was a segmentation fault on return from
8006 the function os_common_find_running_interfaces() in src/os.c.
8008 3. In the check_special_case() function in daemon.c I had used "errno" as an
8009 argument name, which causes warnings on some systems. This was basically a
8010 typo, since it was named "eno" in the comments!
8012 4. The code that waits for the clock to tick (at a resolution of some fraction
8013 of a second) so as to ensure message-id uniqueness was always waiting for
8014 at least one whole tick, when it could have waited for less. [This is
8015 almost certainly not relevant at current processor speeds, where it is
8016 unlikely to ever wait at all. But we try to future-proof.]
8018 5. The function that sleeps for a time interval that includes fractions of a
8019 second contained a race. It did not block SIGALRM between setting the
8020 timer, and suspending (a couple of lines later). If the interval was short
8021 and the sigsuspend() was delayed until after it had expired, the suspension
8022 never ended. On busy systems this could lead to processes getting stuck for
8025 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8026 process, before it forks any delivery processes. The open lookup caching
8027 mechanism meant that the open file or database connection was passed into
8028 the delivery process. The problem was that delivery processes always tidy
8029 up cached lookup data. This could cause a problem for the next delivery
8030 process started by the queue runner, because the external queue runner
8031 process does not know about the closure. So the next delivery process
8032 still has data in the lookup cache. In the case of a file lookup, there was
8033 no problem because closing a file descriptor in a subprocess doesn't affect
8034 the parent. However, if the lookup was caching a connection to a database,
8035 the connection was closed, and the second delivery process was likely to
8036 see errors such as "PGSQL: query failed: server closed the connection
8037 unexpectedly". The problem has been fixed by closing all cached lookups
8038 in a queue runner before running a delivery process.
8040 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8041 seem to have the "const" qualifier which it has on other OS. I've
8044 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8045 *of the same type* (client or server) with the same public name that an
8046 error should be diagnosed.
8048 9. When Exim looked up a host name for an IP address, but failed to find the
8049 original IP address when looking up the host name (a safety check), it
8050 output the message "<ip address> does not match any IP for NULL", which was
8051 confusing, to say the least. The bug was that the host name should have
8052 appeared instead of "NULL".
8054 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8055 user that is built into the binary, and the -C or -D options is used, root
8056 privilege is dropped before the configuration file is read. In addition,
8057 logging is switched to stderr instead of the normal log files. If the
8058 configuration then re-defines the Exim user, the unprivileged environment
8059 is probably not what is expected, so Exim logs a panic warning message (but
8062 However, if deliver_drop_privilege is set, the unprivileged state may well
8063 be exactly what is intended, so the warning has been cut out in that case,
8064 and Exim is allowed to try to write to its normal log files.
8070 1. smtp_return_error_details was not giving details for temporary sender
8071 or receiver verification errors.
8073 2. Diagnose a configuration error if two authenticators have the same public
8076 3. Exim used not to create the message log file for a message until the first
8077 delivery attempt. This could be confusing when incoming messages were held
8078 for policy or load reasons. The message log file is now created at the time
8079 the message is received, and an initial "Received" line is written to it.
8081 4. The automatically generated man page for command line options had a minor
8082 bug that caused no ill effects; however, a more serious problem was that
8083 the procedure for building the man page automatically didn't always
8084 operate. Consequently, release 4.20 contains an out-of-date version. This
8085 shouldn't happen again.
8087 5. When building Exim with embedded Perl support, the script that builds the
8088 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8089 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8091 6. The freeze_tell option was not being used for messages that were frozen on
8092 arrival, either by an ACL or by local_scan().
8094 7. Added the smtp_incomplete_transaction log selector.
8096 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8097 was accepting AUTH without a new EHLO.
8099 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8100 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8103 10. Logging of TCP/IP connections (when configured) now happens in the main
8104 daemon process instead of the child process, so that the TCP/IP connection
8105 count is more accurate (but it can never be perfect).
8107 11. The use of "drop" in a nested ACL was not being handled correctly in the
8108 outer ACL. Now, if condition failure induced by the nested "drop" causes
8109 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8110 or "require"), the connection is dropped.
8112 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8113 that yield "discard" can now be used with an "accept" or a "discard" verb,
8114 but an error is generated for any others (because I can't see a useful way
8115 to define what should happen).
8117 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8118 are now processed in the same way as lines in the Exim configuration file.
8119 In particular, continuation lines are supported.
8121 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8123 15. Added -ti meaning -t -i.
8125 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8126 domains, and warn by logging if others are found.
8128 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8129 returned from GIFCONF when Exim is trying to find the list of interfaces on
8130 a host. The code in os.c has been modified to copy each ifreq to an aligned
8131 structure in all cases.
8133 Also, in some cases, the returned ifreq's were being copied to a 'struct
8134 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8135 means the last couple of bytes of an IPv6 address could be chopped if the
8136 ifreq contained only a normal sockaddr (14 bytes storage).
8138 18. Named domain lists were not supported in the hosts_treat_as_local option.
8139 An entry such as +xxxx was not recognized, and was treated as a literal
8142 19. Ensure that header lines added by a DATA ACL are included in the reject log
8143 if the ACL subsequently rejects the message.
8145 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8146 MD5 (which is deprecated).
8148 21. When testing a filter file using -bf, Exim was writing a message when it
8149 took the sender from a "From " line in the message, but it was not doing so
8150 when it took $return_path from a Return-Path: header line. It now does.
8152 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8153 with a valid header line field name (a series of printing characters
8154 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8156 23. Changed "disc" in the source to "disk" to conform to the documentation and
8157 the book and for uniformity.
8159 24. Ignore Sendmail's -Ooption=value command line item.
8161 25. When execve() failed while trying to run a command in a pipe transport,
8162 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8163 could be confused with a return value of 69 from the command itself. This
8164 has been changed to 127, the value the shell returns if it is asked to run
8165 a non-existent command. The wording for the related log line suggests a
8166 non-existent command as the problem.
8168 26. If received_header_text expands to an empty string, do not add a Received:
8169 header line to the message. (Well, it adds a token one on the spool, but
8170 marks it "old" so that it doesn't get used or transmitted.)
8172 27. Installed eximstats 1.28 (addition of -nt option).
8174 28. There was no check for failure on the call to getsockname() in the daemon
8175 code. This can fail if there is a shortage of resources on the system, with
8176 ENOMEM, for example. A temporary error is now given on failure.
8178 29. Contrary to the C standard, it seems that in some environments, the
8179 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8180 program. Exim now does this explicitly; it affects the formatting of
8181 timestamps using strftime().
8183 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8184 complaints. I've now initialized all the variables, to avoid this.
8186 32. Header lines added by a system filter were not being "seen" during
8187 transport-time rewrites.
8189 33. The info_callback() function passed to OpenSSL is set up with type void
8190 (*)(SSL *, int, int), as described somewhere. However, when calling the
8191 function (actually a macro) that sets it up, the type void(*)() is
8192 expected. I've put in a cast to prevent warnings from picky compilers.
8194 34. If a DNS black list lookup found a CNAME record, but there were no A
8195 records associated with the domain it pointed at, Exim crashed.
8197 35. If a DNS black list lookup returned more than one A record, Exim ignored
8198 all but the first. It now scans all returned addresses if a particular IP
8199 value is being sought. In this situation, the contents of the
8200 $dnslist_value variable are a list of all the addresses, separated by a
8203 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8204 to accept a host name and all its aliases if the forward lookup for any of
8205 them yielded the IP address of the incoming connection. Now it accepts only
8206 those names whose forward lookup yields the correct IP address. Any other
8207 names are discarded. This closes a loophole whereby a rogue DNS
8208 administrator could create reverse DNS records to break through a
8209 wildcarded host restriction in an ACL.
8211 37. If a user filter or a system filter that ran in a subprocess used any of
8212 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8213 the wrong values were passed to the pipe command ($thisaddress had the
8214 value of $0, $0 had the value of $1, etc). This bug was introduced by
8215 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8217 38. Improved the line breaking for long SMTP error messages from ACLs.
8218 Previously, if there was no break point between 40 and 75 characters, Exim
8219 left the rest of the message alone. Two changes have been made: (a) I've
8220 reduced the minimum length to 35 characters; (b) if it can't find a break
8221 point between 35 and 75 characters, it looks ahead and uses the first one
8222 that it finds. This may give the occasional overlong line, but at least the
8223 remaining text gets split now.
8225 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8226 file descriptors might be low, and that setting 1000 would always raise it.
8227 It turns out that in some environments, the limit is already over 1000 and
8228 that lowering it causes trouble. So now Exim takes care not to decrease it.
8230 40. When delivering a message, the value of $return_path is set to $sender_
8231 address at the start of routing (routers may change the value). By an
8232 oversight, this default was not being set up when an address was tested by
8233 -bt or -bv, which affected the outcome if any router or filter referred to
8236 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8237 log only when "message" or "log_message" are set. However, if one of the
8238 conditions was an address verification, or a call to a nested ACL, the
8239 messages generated by the underlying test were being passed through. This
8240 no longer happens. The underlying message is available in $acl_verify_
8241 message for both "message" and "log_message" expansions, so it can be
8242 passed through if needed.
8244 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8245 new expansion $bh_ to give the encoded byte string without charset
8246 translation. Translation happens only if iconv() is available; HAVE_ICONV
8247 indicates this at build time. HEADERS_CHARSET gives the charset to
8248 translate to; headers_charset can change it in the configuration, and
8249 "headers charset" can change it in an individual filter file.
8251 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8252 that creates RFC 2047 encoded "words" labels them as that charset instead
8253 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8254 expansion operator; (ii) when Exim creates a From: line for a local
8255 message; (iii) when a header line is rewritten to include a "phrase" part.
8257 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8258 buggy, causing it to skip the first lines of messages whose message ID
8259 ended in 'D'. This would not have bitten before Exim release 4.14, because
8260 message IDs were unlikely to end in 'D' before then. The effect was to have
8261 incorrect size information for certain domains.
8263 45. #include "config.h" was missing at the start of the crypt16.c module. This
8264 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8267 46. If there was a timeout during a "random" callout check, Exim treated it as
8268 a failure of the random address, and carried on sending RSET and the real
8269 address. If the delay was just some slowness somewhere, the response to the
8270 original RCPT would be taken as a response to RSET and so on, causing
8271 mayhem of various kinds.
8273 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8274 when I implemented it. It didn't allow for the fact that some option values
8275 may legitimately be negative (e.g. size_addition), and it didn't even do
8276 the right test for positive values.
8278 48. Domain names in DNS records are case-independent. Exim always looks them up
8279 in lower case. Some resolvers return domain names in exactly the case they
8280 appear in the zone file, that is, they may contain uppercase letters. Not
8281 all resolvers do this - some return always lower case. Exim was treating a
8282 change of case by a resolver as a change of domain, similar to a widening
8283 of a domain abbreviation. This triggered its re-routing code and so it was
8284 trying to route what was effectively the same domain again. This normally
8285 caused routing to fail (because the router wouldn't handle the domain
8286 twice). Now Exim checks for this case specially, and just changes the
8287 casing of the domain that it ultimately uses when it transmits the message
8290 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8293 50. If a filter generated a file delivery with a non-absolute name (possible if
8294 no home directory exists for the router), the forbid_file option was not
8297 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8298 the existing equality matching.
8300 52. Exim was using ints instead of ino_t variables in some places where it was
8301 dealing with inode numbers.
8303 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8304 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8305 called TMPDIR, and if it finds it is different, it changes its value.
8307 54. The smtp_printf() function is now made available to local_scan() so
8308 additional output lines can be written before returning. There is also an
8309 smtp_fflush() function to enable the detection of a dropped connection.
8310 The variables smtp_input and smtp_batched_input are exported to
8313 55. Changed the default runtime configuration: the message "Unknown user"
8314 has been removed from the ACL, and instead placed on the localuser router,
8315 using the cannot_route_message feature. This means that any verification
8316 failures that generate their own messages won't get overridden. Similarly,
8317 the "Unrouteable address" message that was in the ACL for unverifiable
8318 relay addresses has also been removed.
8320 56. Added hosts_avoid_esmtp to the smtp transport.
8322 57. The exicyclog script was not checking for the esoteric option
8323 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8324 will work only if exicyclog is run under the appropriate euid.
8326 58. Following a discussion on the list, the rules by which Exim recognises line
8327 endings on incoming messages have been changed. The -dropcr and drop_cr
8328 options are now no-ops, retained only for backwards compatibility. The
8329 following line terminators are recognized: LF CRLF CR. However, special
8330 processing applies to CR:
8332 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8333 nor a local message in the state where . is a terminator.
8335 (ii) If a bare CR is encountered in a header line, an extra space is added
8336 after the line terminator so as not to end the header. The reasoning
8337 behind this is that bare CRs in header lines are most likely either
8338 to be mistakes, or people trying to play silly games.
8340 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8341 was being incorrectly given as 18 bytes larger than it should have been.
8342 This is a VOB (very old bug).
8344 60. This may never have affected anything current, but just in case it has:
8345 When the local host is found other than at the start of a list of hosts,
8346 the local host, those with the same MX, and any that follow, are discarded.
8347 When the list in question was part of a longer list of hosts, the following
8348 hosts (not currently being processed) were also being discarded. This no
8349 longer happens. I'm not sure if this situation could ever has previously
8352 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8355 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8356 header. This is something that is recommended in a new Internet Draft, and
8357 is something that is documented as being done by Sendmail. There are two
8358 possible values. For messages generated by the autoreply transport, Exim
8361 Auto-Submitted: auto-replied
8363 whereas for all other generated messages (e.g. bounces) it adds
8365 Auto-Submitted: auto-generated
8367 63. The "personal" condition in filters now includes a test for the
8368 Auto-Submitted: header. If it contains the string "auto-" the message it
8369 not considered personal.
8371 64. Added rcpt_include_affixes as a generic transport option.
8373 65. Added queue_only_override (default true).
8375 66. Added the syslog_duplication option.
8377 67. If what should have been the first header line of a message consisted of
8378 a space followed by a colon, Exim was mis-interpreting it as a header line.
8379 It isn't of course - it is syntactically invalid and should therefore be
8380 treated as the start of the message body. The misbehaviour could have
8381 caused a number of strange effects, including loss of data in subsequent
8382 header lines, and spool format errors.
8384 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8385 client host had authenticated. This control can now be exercised by an ACL
8386 for more flexibility.
8388 69. By default, callouts do not happen when testing with -bh. There is now a
8389 variant, -bhc, which does actually run the callout code, including
8390 consulting and updating the callout cache.
8392 70. Added support for saslauthd authentication, courtesy of Alexander
8395 71. If statvfs() failed on the spool or log directories while checking their
8396 size for availability, Exim confusingly gave the error "space shortage".
8397 Furthermore, in debugging mode it crashed with a floating point exception.
8398 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8399 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8400 serious problem, Exim now writes to the main and panic logs when this
8401 happens, with details of the failure. It then refuses to accept the
8402 incoming message, giving the message "spool directory problem" or "log
8403 directory problem" with a 421 code for SMTP messages.
8405 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8406 0, 1, and 2 exist, because some OS complain for execs without them (see
8407 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8408 descriptors. However, the code omitted to check that the open succeeded,
8409 causing mysterious errors if for some reason the permissions on /dev/null
8410 got screwed. Now Exim writes a message to the main and panic logs, and
8411 bombs out if it can't open /dev/null.
8413 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8414 interact so that it is all more flexible. It is supposed to remain
8415 backwards compatible. Also added extra_local_interfaces.
8417 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8418 to bomb out with an assertion failure - to the client this appears as a
8419 connection drop. This problem occurs in the part of the code that was taken
8420 from the Samba project. Fortunately, the assertion is in a very simple
8421 function, so I have fixed this by reproducing the function inline in the
8422 one place where it is called, and arranging for authentication to fail
8423 instead of killing the process with assert().
8425 75. The SPA client code was not working when the server requested OEM rather
8426 than Unicode encoding.
8428 76. Added code to make require_files with a specific uid setting more usable in
8429 the case where statting the file as root fails - usually a non-root-mounted
8430 NFS file system. When this happens and the failure is EACCES, Exim now
8431 forks a subprocess and does the per-uid checking as the relevant uid.
8433 77. Added process_log_path.
8435 78. If log_file_path was not explicitly set, a setting of check_log_space or
8436 check_log_inodes was ignored.
8438 79. If a space check for the spool or log partitions fails, the incident is now
8439 logged. Of course, in the latter case the data may get lost...
8441 80. Added the %p formatting code to string_format() so that it can be used to
8442 print addresses in debug_print(). Adjusted all the address printing in the
8443 debugging in store.c to use %p rather than %d.
8445 81. There was a concern that a line of code in smtp_in.c could overflow a
8446 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8447 initially expressed, the concern was not well-founded, because trailing
8448 spaces are removed early. However, if the trailing spaces were followed by
8449 a NULL, they did not get removed, so the overflow was possible. Two fixes
8452 (a) I re-wrote the offending code in a cleaner fashion.
8453 (b) If an incoming SMTP command contains a NULL character, it is rejected
8456 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8457 runs initgroups(), so that if the Exim user is in any additional groups,
8458 they will be used during message reception.
8464 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob